Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

A healthy lifestyle is positively associated with mental health and well-being and core markers in ageing

A healthy lifestyle is positively associated with mental health and well-being and core markers... Background: Studies often evaluate mental health and well-being in association with individual health behaviours although evaluating multiple health behaviours that co-occur in real life may reveal important insights into the overall association. Also, the underlying pathways of how lifestyle might affect our health are still under debate. Here, we studied the mediation of different health behaviours or lifestyle factors on mental health and its effect on core mark - ers of ageing: telomere length ( TL) and mitochondrial DNA content (mtDNAc). Methods: In this study, 6054 adults from the 2018 Belgian Health Interview Survey (BHIS) were included. Mental health and well-being outcomes included psychological and severe psychological distress, vitality, life satisfaction, self-perceived health, depressive and generalised anxiety disorder and suicidal ideation. A lifestyle score integrating diet, physical activity, smoking status, alcohol consumption and BMI was created and validated. On a subset of 739 participants, leucocyte TL and mtDNAc were assessed using qPCR. Generalised linear mixed models were used while adjusting for a priori chosen covariates. Results: The average age (SD) of the study population was 49.9 (17.5) years, and 48.8% were men. A one-point increment in the lifestyle score was associated with lower odds (ranging from 0.56 to 0.74) for all studied mental health outcomes and with a 1.74% (95% CI: 0.11, 3.40%) longer TL and 4.07% (95% CI: 2.01, 6.17%) higher mtDNAc. Psychological distress and suicidal ideation were associated with a lower mtDNAc of − 4.62% (95% CI: − 8.85, − 0.20%) and − 7.83% (95% CI: − 14.77, − 0.34%), respectively. No associations were found between mental health and TL. Conclusions: In this large-scale study, we showed the positive association between a healthy lifestyle and both biological ageing and different dimensions of mental health and well-being. We also indicated that living a healthy lifestyle contributes to more favourable biological ageing. Keywords: Mental health, Lifestyle, Biological ageing, Mitochondrial DNA content, Telomere length Background According to the World Health Organization (WHO), a healthy lifestyle is defined as “a way of living that low - ers the risk of being seriously ill or dying early” [1]. Public health authorities emphasise the importance of *Correspondence: [email protected] a healthy lifestyle, but despite this, many individuals Sciensano, Risk and Health Impact Assessment, Juliette Wytsmanstraat 14, worldwide still live an unhealthy lifestyle [2]. In Europe, 1050 Brussels, Belgium 26% of adults smoke [3], nearly half (46%) never exercise Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Hautekiet et al. BMC Medicine (2022) 20:328 Page 2 of 13 [4], 8.4% drink alcohol on a daily basis [5] and over half frame of the BHIS was the Belgian National Register, and (51%) are overweight [5]. These unhealthy behaviours participants were selected based on a multistage strati- have been associated with adverse health outcomes like fied sampling design including a geographical stratifica - cardiovascular diseases [6–8], respiratory diseases [9], tion and a selection of municipalities within provinces, musculoskeletal diseases [10] and, to a lesser extent, of households within municipalities and of respondents mental disorders [11, 12]. within households [26]. The study population for this Even though the association between lifestyle and cross-sectional study included 6054 BHIS participants health outcomes has been extensively investigated, bio- (see flowchart in Additional file  1: Fig. S1) [27–31]. logical mechanisms explaining these observed asso- Minors (< 18 years) and participants not eligible to com- ciations are not yet fully understood. One potential plete the mental health modules (participants who par- mechanism that can be suggested is biological ageing. ticipated through a proxy respondent, i.e. a person of Both telomere length (TL) and mitochondrial DNA con- confidence filled out the survey) were excluded (n = 2172 tent (mtDNAc) are known biomarkers of ageing. Tel- and n = 846, respectively). Furthermore, of the 8593 eligi- omeres are the end caps of chromosomes and consist of ble participants, those with missing information to cre- multiple TTA GGG s equence repeats. They protect chro - ate the mental health indicators, the lifestyle score or the mosomes from degradation and shorten with every cell covariates used in this study were excluded (n = 1642, division because of the “end-replication problem” [13]. 788 and 109, respectively). Mitochondria are crucial to the cell as they are respon- For the first time in 2018, a subset of 1184 BHIS par - sible for apoptosis, the control of cytosolic calcium lev- ticipants contributed to the 2018 Belgian Health Exami- els and cell signalling [14]. Living a healthy lifestyle can nation Survey (BELHES). All BHIS participants were be linked with healthy ageing as both TL and mtDNAc invited to participate except for minors (< 18  years), have been associated with health behaviours like obesity BHIS participants who participated through a proxy [15], diet [16], smoking [17] and alcohol abuse [18]. Fur- respondent and residents of the German Community of thermore, as biomarkers of ageing, both TL and mtD- Belgium, the latter representing 1% of the Belgian popu- NAc have been associated with age-related diseases like lation. Participants were recruited on a voluntary basis Parkinson’s disease [19], coronary heart disease [20], until the regional quotas were reached (450, 300 and atherosclerosis [21] and early mortality [22]. Also, early 350 in respectively Flanders, Brussels Capital Region mortality and higher risks for the aforementioned age- and Wallonia). These participants underwent a health related diseases are observed in psychiatric illnesses, and examination, including anthropological measurements it is suggested that advanced biological ageing underlies and completed an additional questionnaire. Also, blood these observations [23]. and urine samples were collected. Of the 6054 included Multiple studies evaluated individual health behav- BHIS participants, 909 participated in the BELHES. Par- iours, but research on the combination of these health ticipants for whom we could not calculate both TL and behaviours is limited. As they often co-occur and may mtDNAc were excluded (n = 170). More specifically, cause synergistic effects, assessing them in combination participants were excluded because they did not provide with each other rather than independently might bet- a blood sample (n = 91) or because they did not provide ter reflect the real-life situation [24, 25]. Therefore, in a permission for DNA research (n = 32). Twenty samples general adult population, we combined five commonly were excluded from DNA extraction because either total studied health behaviours including diet, smoking sta- blood volume was too low (n = 7), samples were clothed tus, alcohol consumption, BMI and physical activity (n = 1) or tubes were broken due to freezing conditions into one healthy lifestyle score to evaluate its association (n = 12). Twenty-seven samples were excluded because with mental health and well-being and biological ageing. they did not meet the biomarker quality control criteria Furthermore, we evaluated the association between the (high technical variation in qPCR triplicates). This was markers of biological ageing and mental health and well- not met for 3 TL samples, 20 mtDNAc samples and 4 being. We hypothesise that individuals living a healthy samples for both biomarkers. For this subset, we ended lifestyle have a better mental health status, a longer TL up with a final number of 739 participants. Further in this and a higher mtDNAc and that these biomarkers are pos- paper, we refer to “the BHIS subset” for the BHIS partici- itively associated with mental health and well-being. pants (n = 6054) and the “BELHES subset” for the BEL- HES participants (n = 739). Methods As part of the BELHES, this project was approved by Study population the Medical Ethics Committee of the University Hospital In 2018, 11611 Belgian residents participated in the 2018 Ghent (registration number B670201834895). The pro - Belgian Health Interview Survey (BHIS). The sampling ject was carried out in line with the recommendations of Haut ekiet et al. BMC Medicine (2022) 20:328 Page 3 of 13 the Belgian Privacy Commission. All participants have Fourthly, depressive and generalised anxiety disorders signed a consent form that was approved by the Medical were defined using respectively the Patient Health Ques - Ethics Committee. tionnaire (PHQ-9) and the Generalised Anxiety Disorder Questionnaire (GAD-7). We identified individuals who Health interview survey suffer from major depressive syndrome or any other type The BHIS is a comprehensive survey which aims to gain of depressive syndrome according to the criteria of the insight into the health status of the Belgian popula- PHQ-9 [37]. A cut-off point of + 10 on the total sum of tion. The questions on the different dimensions of men - the GAD-7 score was used to indicate generalised anxi- tal health and well-being were based on international ety disorder [31]. Additionally, a dichotomous question standardised and validated questionnaires [32], and this on suicidal ideation was used: “Have you ever seriously resulted in eight mental health outcomes that were used thought of ending your life?”; “If yes, did you have such in this study. Detailed information on each indicator thoughts in the past 12  months?”. Finally, the BHIS also score and its use is addressed in Additional file  1: Table. includes personal, socio-economic and lifestyle informa- S1. Firstly, the General Health Questionnaire (GHQ-12) tion. The standardised Cronbach’s alpha coefficients for provides the prevalence of psychological and severe psy- the PHQ-9, GHQ-12, GAD-7 and questions on vitality of chological distress in the population [27]. On the total the SF-36 ranged between 0.80 and 0.90. GHQ score, cut-off points of + 2 and + 4 were used to identify respectively psychological and severe psycho- logical distress. Healthy lifestyle score Secondly, we used two indicators for the positive We developed a healthy lifestyle score based on five dif - dimensions of mental health: vitality and life satisfaction. ferent health behaviours: body mass index (BMI), smok- Four questions of the short form health survey (SF-36) ing status, physical activity, alcohol consumption and diet indicate the participant’s vital energy level [28, 33]. We (Table 1). These health behaviours were defined as much used a cut-off point to identify participants with an opti - as possible according to the existing guidelines for healthy mal vitality score, which is a score equal to or above one living issued by the Belgian Superior Health Council standard deviation above the mean, as used in previous [38] and the World Health Organisation [39–41]. Firstly, studies [34, 35]. Life satisfaction was measured by the BMI was calculated as a person’s self-reported weight in Cantril Scale, which ranges from 0 to 10 [29]. A cut-off kilogrammes divided by the square of the person’s self- point of + 6 was used to indicate participants with high reported height in metres (kg/m ). BMI was classified or medium life satisfaction versus low life satisfaction. into four categories: underweight (BMI < 18.5  kg/m ), Thirdly, the question “How is your health in general? normal weight (BMI 18.5–24.9 kg/m ), overweight (BMI 2 2 Is it very good, good, fair, bad or very bad?” was used 25.0–29.9  kg/m ) and obese (BMI ≥ 30.0  kg/m ). Due to to assess self-perceived health, also known as self-rated a J-shaped association of BMI with the overall mortality health. Based on WHO recommendations [36], the and multiple specific causes of death, obesity and under - answer categories were dichotomised into “good to very weight were both classified as least healthy [42]. BMI good self-perceived health” and “very bad to fair self-per- was scored as follows: obese and underweight = 0, over- ceived health”. weight = 1 and normal weight = 2. Table 1 Healthy lifestyle score, where each health behaviour is scored from the least healthy to the healthiest Hautekiet et al. BMC Medicine (2022) 20:328 Page 4 of 13 Secondly, smoking status was divided into four cat- (Qiagen, N.V.V Venlo, The Netherlands). The purity and egories. Participants were categorised as regular smok- quantity of the sample were measured with a NanoDrop ers if they smoked a minimum of 4  days per week or if spectrophotometer (ND-2000; Thermo Fisher Scientific, they quit smoking less than 1 month before participation Wilmington, DE, USA). DNA integrity was assessed by (= 0). Occasional smokers were defined as smoking more agarose gel electrophoresis. To ensure a uniform DNA than once per month up to 3  days per week (= 1). Par- input of 6  ng for each qPCR reaction, samples were ™ ® ticipants were classified as former smokers if they quit diluted and checked using the Quant-iT PicoGreen smoking at least 1  month before the questionnaire or if dsDNA Assay Kit (Life Technologies, Europe). they smoked less than once a month (= 2). The final cat - Relative TL and mtDNAc were measured in triplicate egory included people who never smoked (= 3). using a previously described quantitative real-time PCR Thirdly, physical activity was assessed by the question: (qPCR) assay with minor modifications [44, 45]. All reac- “What describes best your leisure time activities dur- tions were performed on a 7900HT Fast Real-Time PCR ing the last year?”. Four categories were established and System (Applied Biosystems, Foster City, CA, USA) in scored as follows: sedentary activities (= 0), light activi- a 384-well format. Used telomere, mtDNAc and single ties less than 4  h/week (= 1), light activities more than copy-gene reaction mixtures and PCR cycles are given 4  h/week or recreational sport less than 4  h/week (= 2) in Additional file  1: Text. S1. Reaction efficiency was and recreational sport more than 4 h or intense training assessed on each plate by using a 6-point serial dilution (= 3). Fourthly, information on the number of alcoholic of pooled DNA. Efficiencies ranged from 90 to 100% for drinks per week was used to categorise alcohol consump- single-copy gene runs, 100 to 110% for telomere runs tion. The different categories were set from high to low and 95 to 105% for mitochondrial DNA runs. Six inter- alcohol consumption: 22 drinks or more/week (= 0), run calibrators (IRCs) were used to account for inter-run 15–21 drinks/week (= 1), 8–14 drinks/week (= 2), 1–7 variability. Also, non-template controls were used in each drinks/week (= 3)and less than 1 drink/week (= 4). run. Raw data were processed and normalised to the ref- Finally, in line with the research by Benetou et  al., a erence gene using the qBase plus software (Biogazelle, diet score was calculated using the frequency of consum- Zwijnaarde, Belgium), taking into account the run-to-run ing fruit, vegetables, snacks and sodas [43]. For fruit as differences. well as vegetable consumption, the frequency was scored Leucocyte telomere length was expressed as the ratio as follows: never (= 0), < 1/week (= 1), 1–3/week (= 2), of telomere copy number to single-copy gene num- 4–6/week (= 3) and ≥ 1/day (= 4). The frequency of con - ber (T/S) relative to the mean T/S ratio of the entire suming snacks and sodas was scored as follows: never study population. Leucocyte mtDNAc was expressed as (= 4), < 1/week (= 3), 1–3/week (= 2), 4–6/week (= 1) the ratio of mtDNA copy number to single-copy gene and ≥ 1/day (= 0). The diet score was then divided into number (M/S) relative to the mean M/S ratio of the tertiles, in line with the research by Benetou et al. [43]. A entire study population. The reliability of our assay was diet score of 0–9 points was classified as the least healthy assessed by calculating the interclass correlation coef- behaviour (= 0). A diet score ranging from 10 to 12 made ficient (ICC) of the triplicate measures (T/S and M/S up the middle category (= 1), and a score from 13 to 16 ratios and T, M and S separately) as proposed by the Tel- was classified as the healthiest behaviour (= 2). omere Research Network, using RStudio version 1.1.463 All five previously described health behaviours were (RStudio PBC, Boston, MA, USA). The intra-plate ICCs combined into one healthy lifestyle score (Table  1). The of T/S ratios, TL runs, M/S ratios, mtDNAc runs and sum of the scores obtained for each health behaviour single-copy runs were respectively 0.804 (p < 0.0001), indicated the absolute lifestyle score. To calculate the 0.907 (p < 0.0001), 0.815 (p < 0.0001), 0.916 (p < 0.0001) relative lifestyle score, each absolute scored health behav- and 0.781 (p < 0.0001). Based on the IRCs, the inter-plate iour was given equal weight by recalculating its maxi- ICC was 0.714 (p < 0.0001) for TL and 0.762 (p < 0.0001) mum absolute score to a relative score of 1. The relative for mtDNAc. lifestyle scores were then summed up to achieve a final continuous lifestyle score, ranging from 0 to 5, with a Statistical analysis higher score representing a healthier lifestyle. Statistical analyses were performed using the SAS soft- ware (version 9.4; SAS Institute Inc., Cary, NC, USA). Telomere length and mitochondrial DNA content assay We performed a log(10) transformation of the TL and Blood samples were collected during the BELHES and mtDNAc data to reduce skewness and to better approxi- centrifuged for 15  min at 3000  rpm before storage mate a normal distribution. Three analyses were done: at − 80 °C. After extracting the buffy coat from the blood (1) In the BHIS subset (n = 6054), we evaluated the asso- sample, DNA was isolated using the QIAgen Mini Kit ciation between the lifestyle score and the mental health Haut ekiet et al. BMC Medicine (2022) 20:328 Page 5 of 13 and well-being outcomes (separately). These results are Results presented as the odds ratio (95% CI) of having a mental Population characteristics health condition or disorder for a one-point increment The characteristics of the BHIS and BELHES subset are in the lifestyle score. (2) In the BELHES subset (n = 739), presented in Table  2. In the BHIS subset, 48.8% of the we evaluated the association between the lifestyle score participants were men. The average age (SD) was 49.9 and both TL and mtDNAc (separately). These results are (17.5) years, and most participants were born in Belgium presented as the percentage difference in TL or mtD - (79.5%). The highest educational level in the household NAc (95% CI) for a one-point increment in the lifestyle was most often college or university degree (53.3%), and score. (3) In the BELHES subset (n = 739), we evaluated the most common household composition was couple the association between the mental health and well-being with child(ren) (37.7%). The proportion of participants in outcomes and both TL and mtDNAc (separately). These different regions of Belgium, i.e. Flanders, Brussels Capi - results are presented as the percentage difference in TL tal Region and Wallonia, was respectively 41.1%, 23.3% or mtDNAc (95% CI) when having a mental health condi- and 35.6%. For the BELHES subset, we found similar tion or disorder compared with the healthy group. results except for region and education. We noticed more For all three analyses, we performed multivariable lin- participants from Flanders and more participants with a ear mixed models (GLIMMIX; unstructured covariance high educational level in the household. The mean (SD) matrix) taking into account a priori selected covariates relative TL and mtDNAc were respectively 1.04 (0.23) including age (continuous), sex (male, female), region and 1.03 (0.24). TL and mtDNAc were positively corre- (Flanders, Brussels Capital Region, Wallonia), highest lated (Spearman’s correlation = 0.21, p < 0.0001). educational level of the household (up to lower second- We compared (1) the characteristics of the 6054 ary, higher secondary, college or university), country of eligible BHIS participants that were included in the birth (Belgium, EU, non-EU) and household type (single, BHIS subset with the 2539 eligible participants that one parent with child, couple without child, couple with were excluded from the BHIS subset (Additional file  1: child, others). To capture the non-linear effect of age, we Table  S2) and (2) the 739 participants from the BHIS included a quadratic term when the result of the analy- subset that were included in the BELHES subset with sis showed that both the linear and quadratic terms had the 5315 participants that were excluded from the BEL- a p-value < 0.1. For the two analyses on TL and mtDNAc, HES subset (Additional file  1: Table  S3). Except for sex we additionally adjusted for the date of participation in and nationality in the latter, all other covariates showed the BELHES. As multiple members of one household differences between the included and excluded groups. participated, we added household numbers in the ran- On the other hand, population data from 2018 indicates dom statement. that the average age (SD) of the adult Belgian population Bivariate analyses evaluating the associations between was 49.5 (18.9) with a distribution over Flanders, Brus- the characteristics and TL, mtDNAc, the lifestyle score sels Capital Region and Wallonia of respectively 58.2%, or psychological distress as a parameter of mental health 10.2% and 31.6% and that 48.7% were men. The distribu - and well-being are evaluated based on the same model. tion of our sample according to age and sex thus largely The chi-squared tests (categorical data) and t-tests (con - corresponds to the age and sex distribution of the adult tinuous data) were used to evaluate the characteristics of Belgian population figures. The large difference in the included and excluded participants. The lifestyle score regional distribution is due to the oversampling of the was validated by creating a ROC curve and calculating Brussels Capital Region in the BHIS. the area under the curve (AUC) of the adjusted associa- Bivariate associations evaluating the characteris- tion between the lifestyle score and self-perceived health. tics with TL, mtDNAc, the lifestyle score or psycho- Adjustments were made for age, sex, region, highest logical distress as a parameter of mental health are educational level of the household, country of birth and presented in Additional file  1: Table S4. Briefly, men had household type. a − 6.41% (95% CI: − 9.10 to − 3.65%, p < 0.0001) shorter In a sensitivity analysis, to evaluate the robustness of TL, a − 8.03% (95% CI: − 11.00 to − 4.96%, p < 0.0001) our findings, we additionally adjusted our main models lower mtDNAc, lower odds of psychological distress separately for perceived quality of social support (poor, (OR = 0.59, 95% CI: 0.53 to 0.66, p < 0.0001) and a life- moderate, strong) and chronic disease (suffering from style score of − 0.28 (95% CI: − 0.32 to − 0.24, p < 0.0001) any chronic disease or condition: yes, no). The third points less compared with women. Furthermore, a 1-year model, evaluating the biomarkers with the mental health increment in age was associated with a − 0.64% (− 0.73 outcomes, was also additionally adjusted for the lifestyle to − 0.55%, p < 0.0001) shorter TL and a − 0.19% (95% score. CI: − 0.31 to − 0.08%, p = 0.00074) lower mtDNAc. Hautekiet et al. BMC Medicine (2022) 20:328 Page 6 of 13 Table 2 Characteristics of the study population for the BHIS (n = 6054) and the BELHES subset (n = 739) Characteristics BHIS subset, n (%) or mean (SD) BELHES subset, n (%) or mean (SD) Male 2955 (48.8%) 369 (49.9%) Age, years 49.9 (17.5) 48.3 (15.5) Region Flanders 2488 (41.1%) 356 (48.2%) Brussels Capital Region 1410 (23.3%) 158 (21.4%) Wallonia 2156 (35.6%) 225 (30.5%) Highest educational level in the household Up to lower secondary school 1010 (16.7%) 92 (12.5%) Higher secondary school 1819 (30.1%) 196 (26.5%) College or university 3225 (53.3%) 451 (61.0%) Household composition Single 1339 (22.1%) 130 (17.6%) One parent with a child 514 (8.5%) 53 (7.2%) Couple without child 1674 (27.7%) 196 (26.5%) Couple with child(ren) 2283 (37.7%) 326 (44.1%) Others 244 (4.0%) 34 (4.6%) Country of birth Belgium 4812 (79.5%) 596 (80.7%) EU 619 (10.2%) 77 (10.4%) Non-EU 623 (10.3%) 66 (8.9%) Perceived quality of social support Poor 946 (15.7%) 116 (15.9%) Moderate 2978 (49.5%) 379 (51.9%) Strong 2093 (34.8%) 236 (32.3%) Chronic disease or condition 1776 (29.5%) 206 (28.1%) n = 6017 and 731 for the BHIS and BELHES subset, respectively n = 6017 and 733 for the BHIS and BELHES subset, respectively Mental health prevalence and lifestyle characteristics to have had suicidal thoughts in the past 12 months. Sim- Within the BHIS subset, 32.3% and 18.0% of the par- ilar results were found for the BELHES subset (Table 3). ticipants had respectively psychological and severe psy- Within the BHIS subset, the average lifestyle score chological distress. 86.7% had suboptimal vitality, 12.0% (SD) was 3.1 (0.9) (Table 4). A histogram of the lifestyle indicated low life satisfaction and 22.0% had very bad to score is shown in Additional file  1: Fig. S2. 16.6% were fair self-perceived health. The prevalence of depressive regular smokers, and 4.9% reported 22 alcoholic drinks and generalised anxiety disorders was respectively 9.0% per week or more. 29.7% reported that their main lei- and 10.8%, respectively. 4.4% of the participants indicated sure time included mainly sedentary activities, and Table 3 Prevalence of the mental health and well-being outcomes for the BHIS (n = 6054) and the BELHES subset (n = 739) Mental health and well-being outcomes BHIS subset, n (%) BELHES subset, n (%) Psychological distress 1954 (32.3%) 252 (34.1%) Severe psychological distress 1087 (18.0%) 131 (17.7%) Suboptimal vitality 5247 (86.7%) 661 (89.5%) Low life satisfaction 726 (12.0%) 82 (11.1%) Very bad to fair self-perceived health 1333 (22.0%) 135 (18.3%) Depressive disorder 544 (9.0%) 63 (8.5%) Generalised anxiety disorder 655 (10.8%) 76 (10.3%) Suicidal ideation 269 (4.4%) 38 (5.1%) Haut ekiet et al. BMC Medicine (2022) 20:328 Page 7 of 13 Table 4 Characteristics of the healthy lifestyle score for the BHIS subset (n = 6054) and the BELHES subset (n = 739) BHIS subset, n (%) or mean (SD) BELHES subset, n (%) or mean (SD) Lifestyle score 3.1 (0.9) 3.1 (0.9) Smoking status Regular smoker (0) 1002 (16.6%) 129 (17.5%) Occasional smoker (1) 159 (2.6%) 25 (3.4%) Former smoker (2) 1472 (24.3%) 180 (24.4%) Nonsmoker (3) 3421 (56.5%) 405 (54.8%) BMI Underweight/obese (0) 1123 (18.6%) 119 (16.1%) Overweight (1) 2073 (34.2%) 248 (33.6%) Normal weight (2) 2858 (47.2%) 372 (50.3%) Physical activity Sedentary activities (0) 1795 (29.7%) 185 (25.0%) Light activities < 4 h/week (1) 1391 (23.0%) 167 (22.6%) Light activities > 4 h/week or sport < 4 h/week (2) 1820 (30.1%) 227 (30.7%) Sport > 4 h/week or intense training (3) 1048 (17.3%) 160 (21.7%) Alcohol consumption ≥ 22 drinks/week (0) 294 (4.9%) 31 (4.2%) 15–21 drinks/week (1) 323 (5.3%) 52 (7.0%) 8–14 drinks/week (2) 738 (12.2%) 101 (13.7%) 1–7 drinks/week (3) 1777 (29.4%) 235 (31.8%) < 1 drink/week or abstainers (4) 2922 (48.3%) 320 (43.3%) Diet Diet score 0–9 (0) 1769 (29.2%) 210 (28.4%) Diet score 10–12 (1) 2621 (43.3%) 325 (44.0%) Diet score 13–16 (2) 1664 (27.5%) 204 (27.6%) 18.6% were underweight or obese. 29.2% were classified Table 5 Associations between the lifestyle score and the mental as having an unhealthy diet score. The participants of health and well-being outcomes the BELHES subset were slightly more active, but no Lifestyle score OR 95% CI p-value other dissimilarities were found (Table  4). The ROC curve shows an area under the curve (AUC) of 0.74, Psychological distress 0.74 0.69, 0.79 < 0.0001 indicating a 74% predictive accuracy for the lifestyle Severe psychological distress 0.69 0.64, 0.75 < 0.0001 score as a self-perceived health predictor (Additional Suboptimal vitality 0.62 0.56, 0.68 < 0.0001 file 1 : Fig. S3). Low life satisfaction 0.62 0.56, 0.68 < 0.0001 Very bad to fair self-perceived health 0.56 0.52, 0.61 < 0.0001 Depressive disorder 0.57 0.51, 0.63 < 0.0001 Healthy lifestyle and mental health and well-being Generalised anxiety disorder 0.63 0.57, 0.69 < 0.0001 Living a healthier lifestyle, indicated by having a higher Suicidal ideation 0.63 0.55, 0.72 < 0.0001 lifestyle score, was associated with lower odds of all Odds ratios (OR) and 95% confidence intervals (CI) for the different mental mental health and well-being outcomes (Table 5). After health and well-being outcomes for a one-point increment in the lifestyle score. Analyses were adjusted for age, sex, region, highest educational level in the adjustment, a one-point increment in the lifestyle household, household composition and country of birth score was associated with lower odds of psychological (OR = 0.74, 95% CI: 0.69, 0.79) and severe psychologi- cal distress (OR = 0.69, 95% CI: 0.64, 0.75). Similarly, 0.62 (95% CI: 0.56, 0.68) and 0.56 (95% CI: 0.52, 0.61). for the same increment, the odds of suboptimal vitality, Finally, the odds of having depressive disorder, general- low life satisfaction and very bad to fair self-perceived ised anxiety disorder or suicidal ideation were respec- health were respectively 0.62 (95% CI: 0.56, 0.68), tively 0.57 (95% CI: 0.51, 0.63), 0.63 (95% CI: 0.57, 0.69) Hautekiet et al. BMC Medicine (2022) 20:328 Page 8 of 13 Table 6 Associations between the biomarkers and both the lifestyle score and the mental health and well-being outcomes Lifestyle score/mental health and TL mtDNAc well-being outcome % difference 95% CI p-value % difference 95% CI p-value Lifestyle score 1.74 0.11, 3.40 0.037 4.07 2.01, 6.17 0.00012 Psychological distress − 0.12 − 3.04, 2.88 0.94 − 2.29 − 5.82, 1.36 0.21 Severe psychological distress 0.21 − 3.39, 3.94 0.91 − 4.62 − 8.85, − 0.20 0.041 Suboptimal vitality − 3.26 − 7.46, 1.12 0.14 − 2.37 − 7.62, 3.17 0.39 Low life satisfaction 0.20 − 4.18, 4.79 0.93 − 2.00 − 7.30, 3.59 0.47 Very bad to fair self-perceived health − 0.79 − 4.36, 2.90 0.67 − 2.53 − 6.87, 2.00 0.27 Depressive disorder 2.54 − 2.41, 7.73 0.32 2.84 − 3.28, 9.36 0.37 Generalised anxiety disorder 0.56 − 3.88, 5.20 0.81 2.05 − 3.52, 7.94 0.48 Suicidal ideation 0.73 − 5.44, 7.31 0.82 − 7.83 − 14.77, − 0.34 0.041 Difference (%) in relative telomere length (TL) and mitochondrial DNA content (mtDNAc) (with 95% CI) (1) for a one-point increment in the lifestyle score or (2) when having a mental health disorder or condition compared with the healthy group. Analyses were adjusted for age, sex, region, highest educational level in the household, household composition, country of birth and day of sample collection and 0.63 (95% CI: 0.55, 0.72) for a one-point increment telomere length and mitochondrial DNA content. Having in the lifestyle score. a healthy lifestyle was positively associated with all men- tal health and well-being indicators and the markers of The biomarkers of ageing biological ageing. Furthermore, having had suicidal idea- After adjustment, living a healthy lifestyle was positively tion or suffering from severe psychological distress was associated with both TL and mtDNAc (Table  6). A one- associated with a lower mtDNAc. However, no associa- point increment in the lifestyle score was associated with tion was found between mental health and TL. a 1.74 (95% CI: 0.11, 3.40%, p = 0.037) higher TL and a 4.07 (95% CI: 2.01, 6.17%, p = 0.00012) higher mtDNAc. Healthy lifestyle and mental health and well-being People suffering from severe psychological distress had In the first part of this research, we evaluated the associa - a − 4.62% (95% CI: − 8.85, − 0.20%, p = 0.041) lower mtD- tion between lifestyle and mental health and well-being and NAc compared with those who did not suffer from severe showed that living a healthy lifestyle was positively associ- psychological distress. Similarly, people with suicidal ide- ated with better mental health and well-being outcomes. ation had a − 7.83% (95% CI: − 14.77, − 0.34%, p = 0.041) Similar trends were found in previous studies for each of lower mtDNAc compared with those without suicidal the health behaviours separately [11, 12, 46–48]. Although ideation. No associations were found for the other men- evaluating these health behaviours separately provides tal health and well-being outcomes, and no associations valuable information, assessing them in combination with were found between mental health and TL (Table 6). each other rather than independently might better reflect the real-life situation as they often co-occur and may exert a synergistic effect on each other [24, 25, 49]. For exam- ple, 68% of the adults in England engaged in two or more Sensitivity analysis unhealthy behaviours [25]. Especially, smoking status and Additional adjustment of the main analyses for perceived alcohol consumption co-occurred, but half of the stud- quality of social support, chronic disease or lifestyle score ies in the review by Noble et  al. indicated clustering of all (in the association between the mental health outcomes included health behaviours [24]. and the biomarkers of ageing) did not strongly change To date, the number of studies evaluating the combina- the effect of our observations (Additional file  1: Tables tion of multiple health behaviours and mental health and S5-S7). However, we noticed that most of the associations well-being in adults is limited, and most of them use a between severe psychological distress or suicidal ideation different methodology to assess this association [50–56]. and mtDNAc showed marginally significant results. Firstly, differences are found between the included health behaviours. Most studies included the four “SNAP” risk factors, i.e. smoking, poor nutrition, excess alcohol con- Discussion sumption and physical inactivity. Other health behav- In this study, we evaluated the associations between iours that were sometimes included were BMI/obesity, eight mental health and well-being outcomes, a healthy sleep duration/quality and psychological distress [50, 53, lifestyle score and 2 biomarkers of biological ageing: Haut ekiet et al. BMC Medicine (2022) 20:328 Page 9 of 13 54, 56]. Secondly, differences are found in the scoring of our results, in a study population of 1661 men, the sum the health behaviours and the use of the lifestyle score. score of a healthier lifestyle was correlated with a longer Whereas in this study the health behaviours were scored TL [66]. Similar results were found by Sun et al. where a categorically, studies often dichotomised the health combination of healthy lifestyles in a female study pop- behaviours and/or the final lifestyle score [50, 52, 53, 56]. ulation was associated with a longer TL compared with Also, two studies performed clustering [54, 55]. Health the least healthy group [67]. Also, improvement in life- behaviours can cluster together at both ends of the risk style has been associated with TL maintenance in the spectrum, but less is known about the middle categories. elderly at risk for dementia [68], and a lifestyle interven- This is avoided by using the cluster method where partic - tion programme was positively associated with leucocyte ipants are clustered based on similar behaviours. On the telomere length in children and adolescents [69]. These other hand, a lifestyle score can be of better use and more results suggest that on a biological level, a healthy life- easily be interpreted when aiming to compare healthy style is associated with healthy ageing. Within this con- versus unhealthy lifestyles, as was the case for this study. text, a study on adults aged 60 and older showed that Despite these different methods, all previously men - maintaining a normal weight, not smoking and perform- tioned studies show similar results. Together with our ing regular physical activity were associated with slower findings, which also support these results, this pro - development of disability and a reduction in mortality vides clear evidence that an unhealthy lifestyle is associ- [70]. Similarly, midlife lifestyle factors like non-smoking, ated with poor mental health and well-being outcomes. higher levels of physical activity, non-obesity and good Important to notice is that, like our research, most stud- social support have been associated with successful age- ies in this field have a cross-sectional design and are ing, 22 years later [71]. therefore not able to assume causality. Therefore, men - tal health might be the cause or the consequence of an unhealthy lifestyle. Further prospective and longitudinal Mental health and well-being and biomarkers of ageing studies are warranted to confirm the direction of the Finally, we evaluated the association between the bio- association. markers  of ageing and the mental health and well-being outcomes. The hypothesis that biological ageing is associ - ated with mental health has been supported by observa- Healthy lifestyle and biomarkers of ageing tions showing that chronically stressed or psychiatrically How lifestyle affects our health is not yet fully under - ill persons have a higher risk for age-related diseases like stood. One possible pathway is through oxidative stress dementia, diabetes and hypertension [23, 72, 73]. Impor- and biological ageing. An unhealthy lifestyle has been tant to notice is that, like our research, the majority of associated with an increase in oxidative stress [57–59], studies on this topic have a cross-sectional design and and in turn, higher concentrations of oxidative stress are therefore are unable to identify causality. Therefore, it is known to negatively affect TL and mtDNAc [60]. In this currently unknown whether psychological diseases accel- study, we showed that living a healthy lifestyle was associ- erate biological ageing or whether biological ageing pre- ated with a longer TL and a higher mtDNAc. Our results cedes the onset of these diseases [74]. showed a stronger association of lifestyle with mtDNAc Our results showed a lower mtDNAc for individuals compared with TL. TL is strongly determined by TL at with suicidal ideation or severe psychological distress birth [61]. On the other hand, mtDNAc might be more but not for any of the other mental health outcomes. Evi- variable in shorter time periods. Although mtDNAc and dence on the association between mtDNAc and mental TL were strongly correlated, this could explain why life- health is inconsistent. Women above 60  years old with style is more strongly associated with mtDNAc. However, depression had a significantly lower mtDNAc compared we can only speculate about this, and further research is with the control group [75]. Furthermore, individuals necessary to confirm our results. with a low mtDNAc had poorer outcomes in terms of Similar as for the association with mental health, in self-rated health [76]. In contrast, Otsuka et  al. showed previous studies, the biomarkers have been associated a higher peripheral blood mtDNAc in suicide completers with health behaviours separately rather than combined [77], and studies on major depressive syndrome [78] and [62–65]. To our knowledge, we are the first to evaluate self-rated health [79] showed the same trend. Finally, the associations between a healthy lifestyle score and Vyas et  al. showed no significant association between mtDNAc. Our results are in line with our expectations. mtDNAc and depression status in mid-life and older As TL and mtDNAc are known to be correlated [60], adults [80]. These differences might be due to the dif - we would expect similar trends for both biomarkers. In ferences in study population and methods. For example, the case of TL, few studies included a combined lifestyle the two studies indicating lower mtDNAc in association score in association with this biomarker. Consistent with Hautekiet et al. BMC Medicine (2022) 20:328 Page 10 of 13 with poor mental health both had an elderly study popu- (SD) number of days is 52 (35). This is less than the lation, and one study population consisted of psychiatri- period for suicidal ideation, assessed over the 12 previ- cally ill patients. Next to that, differences were found in ous months, but there might be a more limited over- the type of samples, mtDNAc assays and questionnaires lap with the period for assessment of the other mental or diagnostics. The inconsistency of these studies and our health variables, such as vitality and psychological results calls for further research on this association and distress, assessed over the last few weeks, and depres- for standardisation of methods within studies to enable sive and generalised anxiety disorders, assessed over clear comparisons. the last 2  weeks. Fourthly, due to a non-response bias, As for TL, we did not find an association with any of the the lowest socio-economic classes are less represented mental health and well-being outcomes. Previous studies in our study population. This will not affect our dose– in adults showed a lower TL in association with current response associations but might affect the generalis - but not remitted anxiety disorder [81], depressive [82] ability of our findings to the overall population. Finally, and major depressive disorder [73, 83], childhood trauma we do not have data on blood cell counts, which has [84] suicide [77, 85], depressive symptoms in younger been associated with mtDNAc [95]. adults [86] and acculturative stress and postpartum depression in Latinx women [87]. Also, in a meta-anal- ysis, psychiatric disorders overall were associated with a Conclusions shorter leucocyte TL [88]. However, other studies failed In this large-scale study, we showed that living a healthy to demonstrate an association between TL and mental lifestyle was positively associated with mental health and health outcomes like major depressive disorder [89], late- well-being and, on a biological level, with a higher TL life depression [90] and anxiety disorder [91]. Again, this and mtDNAc, indicating healthy ageing. Furthermore, could be due to a different method to assess the mental individuals with suicidal ideation or suffering from health outcomes, a different study design, uncontrolled severe psychological distress had a lower mtDNAc. Our confounding factors and the type of telomere assay. For findings suggest that implementing strategies to incor - example, a meta-analysis showed stronger associations porate healthy lifestyle changes in the public’s daily life with depression when using southern blot or FISH assay could be beneficial for public health, and might offset compared with qPCR to measure telomere length [92]. the negative impact of environmental stressors. How- ever, further studies are necessary to confirm our results and especially prospective and longitudinal studies are Strengths and limitations essential to determine causality of the associations. An important strength of this study is the use of a vali- dated lifestyle score that can easily be reproduced and used for other research on lifestyle. Secondly, we were Abbreviations AUC: Ar ea under the curve; BMI: Body mass index; CI: Confidence intervals; able to use a large sample size for our analyses in the GAD-7: Generalised Anxiety Disorder Questionnaire; GHQ-12: General Health BHIS subset. Thirdly, by assessing multiple dimensions Questionnaire; IRC: Inter-run calibrator; mtDNAc: Mitochondrial DNA content; of mental health and well-being, we were able to give a OR: Odds ratio; PHQ-9: Patient Health Questionnaire; ROC curve: Relative operating characteristic curve; SF-36: Short Form Health Survey; TL: Telomere comprehensive overview of the mental health status. To length. our knowledge, we are the first to evaluate the associa - tions between a healthy lifestyle score and mtDNAc. Supplementary Information Our results should however be interpreted with con- The online version contains supplementary material available at https:// doi. sideration for some limitations. As mentioned before, org/ 10. 1186/ s12916- 022- 02524-9. the study has a cross-sectional design, and therefore, we cannot assume causality. Secondly, for the life- Additional file 1: Text S1. TL, mtDNAc and single copy-gene reaction mixture and PCR cycling conditions. Table S1. The mental health indica- style score, we used self-reported data, which might tors with their scores and uses. Table S2. Comparison of the characteris- not always represent the actual situation. For exam- tics of the 6,054 eligible BHIS participants that were included in the BHIS ple, BMI values tend to be underestimated due to the subset compared to the 1,838 eligible participants that were excluded from the BHIS subset. Table S3. Comparison of the characteristics of the overestimation of height and underestimation of weight 739 participants from the BHIS subset that were included in the BELHES [93], and also, smoking behaviour is often underesti- subset compared to the 5,315 participants that were excluded from the mated [94]. Also, equal weights were used for each of BELHES subset. Table S4. Bivariate associations between the characteris- tics and telomere length ( TL), mitochondrial DNA content (mtDNAc), the the health behaviours as no objective information was lifestyle score or psychological distress. Table S5. Results of the sensitivity available on which weight should be given to a spe- analysis of the association between lifestyle and mental health. Table S6. cific health behaviour. Thirdly, there is a distinct time Results of the sensitivity analysis of the association between lifestyle and the biomarkers of ageing. Table S7. Results of the sensitivity analysis of lag between the completion of the BHIS questionnaire the association between mental health and the biomarkers of ageing. Fig. and the collection of the BELHES samples. The mean Haut ekiet et al. BMC Medicine (2022) 20:328 Page 11 of 13 6. Whitman IR, Agarwal V, Nah G, Dukes JW, Vittinghoff E, Dewland TA, et al. S1. Exclusion criteria. The BHIS subset consisted of 6,055 BHIS participants Alcohol abuse and cardiac disease. J Am Coll Cardiol. 2017;69(1):13–24. and the BELHES subset consisted of 739 BELHES participants. Fig. S2. https:// doi. org/ 10. 1016/j. jacc. 2016. 10. 048. Histogram of the lifestyle score. Fig. S3. Validation of the lifestyle score. 7. Koliaki C, Liatis S, Kokkinos A. Obesity and cardiovascular disease: revisit- ROC curve for the lifestyle score as a predictor for good to very good ing an old relationship. Metabolism. 2019;92:98–107. https:// doi. org/ 10. self-perceived health. The model was adjusted for age, sex, region, highest 1016/j. metab ol. 2018. 10. 011. educational level in the household, household composition and country 8. Freisling H, Viallon V, Lennon H, Bagnardi V, Ricci C, Butterworth AS, et al. of birth. Lifestyle factors and risk of multimorbidity of cancer and cardiometa- bolic diseases: a multinational cohort study. BMC Med. 2020;18(1):5. https:// doi. org/ 10. 1186/ s12916- 019- 1474-7. Acknowledgements 9. Liu Y, Pleasants RA, Croft JB, Wheaton AG, Heidari K, Malarcher AM, et al. We are grateful to all BHIS and BELHES participants for contributing to this study. Smoking duration, respiratory symptoms, and COPD in adults aged ≥ 45 years with a smoking history. Int J Chron Obstruct Pulmon Dis. Authors’ contributions 2015;10:1409. https:// doi. org/ 10. 2147/ COPD. S82259. PH drafted the paper. PH, NS, MD and ED set up the design of the study. 10. Kirsch Micheletti J, Bláfoss R, Sundstrup E, Bay H, Pastre CM, Andersen LL. NS, DM, JvdH, TN and ED reviewed and commented on the manuscript. All Association between lifestyle and musculoskeletal pain: cross-sectional authors read and approved the final manuscript. study among 10,000 adults from the general working population. BMC Musculoskelet Disord. 2019;20(1):609. https:// doi. org/ 10. 1186/ Funding s12891- 019- 3002-5. The HuBiHIS project is financed by Sciensano (PJ ) N°: 1179–101. Dries Mar - 11. Bowe AK, Owens M, Codd MB, Lawlor BA, Glynn RW. Physical activity and tens is a postdoctoral fellow of the Research Foundation—Flanders (FWO mental health in an Irish population. Ir J Med Sci. 2019;188(2):625–31. 12X9620N). https:// doi. org/ 10. 1007/ s11845- 018- 1863-5. 12. Richardson S, McNeill A, Brose LS. Smoking and quitting behaviours by Availability of data and materials mental health conditions in Great Britain (1993–2014). Addict Behav. The dataset used for this study is available through a request to the Health 2019;90:14–9. https:// doi. org/ 10. 1016/j. addbeh. 2018. 10. 011. Committee of the Data Protection Authority. 13. Levy MZ, Allsopp RC, Futcher AB, Greider CW, Harley CB. Telomere end-replication problem and cell aging. J Mol Biol. 1992;225(4):951–60. https:// doi. org/ 10. 1016/ 0022- 2836(92) 90096-3. Declarations 14. Shaughnessy DT, McAllister K, Worth L, Haugen AC, Meyer JN, Domann FE, et al. Mitochondria, energetics, epigenetics, and cellular responses to Ethics approval and consent to participate stress. Environ Health Perspect. 2014;122(12):1271–8. https:// doi. org/ 10. As part of the BELHES, this project was approved by the Medical Ethics Com- 1289/ ehp. 14084 18. mittee of the University Hospital Ghent (registration number B670201834895). 15. Cui Y, Gao Y T, Cai Q, Qu S, Cai H, Li HL, et al. Associations of leukocyte The project was carried out in line with the recommendations of the Belgian telomere length with body anthropometric indices and weight change Privacy Commission. All participants have signed a consent form that was in Chinese women. Obesity. 2013;21(12):2582–8. https:// doi. org/ 10. 1002/ approved by the Medical Ethics Committee. oby. 20321. 16. Crous-Bou M, Fung T T, Prescott J, Julin B, Du M, Sun Q, et al. Mediter- Consent for publication ranean diet and telomere length in Nurses’ Health Study: population Not applicable. based cohort study. BMJ. 2014;349:g6674. https:// doi. org/ 10. 1136/ bmj. g6674. Competing interests 17. Janssen BG, Gyselaers W, Byun H-M, Roels HA, Cuypers A, Baccarelli AA, The authors declare that they have no competing interests. et al. Placental mitochondrial DNA and CYP1A1 gene methylation as molecular signatures for tobacco smoke exposure in pregnant women Author details and the relevance for birth weight. J Transl Med. 2017;15(1):5. https:// doi. Sciensano, Risk and Health Impact Assessment, Juliette Wytsmanstraat org/ 10. 1186/ s12967- 016- 1113-4. 14, 1050 Brussels, Belgium. Centre for Environmental Sciences, Hasselt 18. Navarro-Mateu F, Husky M, Cayuela-Fuentes P, Álvarez FJ, Roca-Vega A, University, 3500 Hasselt, Belgium. Sciensano, Epidemiology and Public Health, Rubio-Aparicio M, et al. The association of telomere length with substance Juliette Wytsmanstraat 14, 1050 Brussels, Belgium. Centre for Environment use disorders: a systematic review and meta-analysis of observational stud- and Health, Leuven University, 3000 Leuven, Belgium. ies. Addiction. 2020;116(8):1954–72. https:// doi. org/ 10. 1111/ add. 15312. 19. Pyle A, Anugrha H, Kurzawa-Akanbi M, Yarnall A, Burn D, Hudson G. Received: 4 February 2022 Accepted: 10 August 2022 Reduced mitochondrial DNA copy number is a biomarker of Parkinson’s disease. Neurobiol Aging. 2016;38:216.e7-.e10. https:// doi. org/ 10. 1016/j. neuro biola ging. 2015. 10. 033. 20. Ashar FN, Zhang Y, Longchamps RJ, Lane J, Moes A, Grove ML, et al. Asso- ciation of mitochondrial DNA copy number with cardiovascular disease. References JAMA Cardiol. 2017;2(11):1247–55. https:// doi. org/ 10. 1001/ jamac ardio. 1. World Health Organization. Healthy living: what is a healthy lifestyle? 2017. 3683. Copenhagen: WHO Regional Office for Europe; 1999. 21. Chen S, Lin J, Matsuguchi T, Blackburn E, Yeh F, Best LG, et al. Short 2. World Health Organization. Tackling NCDs: ‘best buys’ and other recom- leukocyte telomere length predicts incidence and progression of carotid mended interventions for the prevention and control of noncommunica- atherosclerosis in American Indians: the Strong Heart Family Study. ble diseases. Geneva: World Health Organization; 2017. Aging. 2014;6(5):414–27. https:// doi. org/ 10. 18632/ aging. 100671. 3. World Health Organization. WHO global report on trends in prevalence of 22. Mons U, Müezzinler A, Schöttker B, Dieffenbach AK, Butterbach K, tobacco smoking 2000–2025. 3rd ed. Geneva: World Health Organisation; Schick M, et al. Leukocyte telomere length and all-cause, cardiovascular disease, and cancer mortality: results from individual-participant-data 4. World Health Organization. 2021 Physical Activity Factsheets for the Euro- meta-analysis of 2 large prospective cohort studies. Am J Epidemiol. pean Union Member States in the WHO European Region. Copenhagen: 2017;185(12):1317–26. https:// doi. org/ 10. 1093/ aje/ kww210. WHO Regional Office for Europe; 2021. 23. Lindqvist D, Epel ES, Mellon SH, Penninx BW, Révész D, Verhoeven JE, 5. Eurostat. Database on health determinants 2021 [Available from: et al. Psychiatric disorders and leukocyte telomere length: underlying https:// ec. europa. eu/ euros tat/ web/ health/ data/ datab ase? node_ mechanisms linking mental illness with cellular aging. Neurosci Biobehav code= hlth_ det. Rev. 2015;55:333–64. https:// doi. org/ 10. 1016/j. neubi orev. 2015. 05. 007. Hautekiet et al. BMC Medicine (2022) 20:328 Page 12 of 13 24. Noble N, Paul C, Turon H, Oldmeadow C. Which modifiable health risk 46. Jacka FN, O’Neil A, Opie R, Itsiopoulos C, Cotton S, Mohebbi M, et al. A behaviours are related? A systematic review of the clustering of Smoking, randomised controlled trial of dietary improvement for adults with major Nutrition, Alcohol and Physical activity (‘SNAP’) health risk factors. Prev depression (the ‘SMILES’ trial). BMC Med. 2017;15(1):23. https:// doi. org/ 10. Med. 2015;81:16–41. https:// doi. org/ 10. 1016/j. ypmed. 2015. 07. 003.1186/ s12916- 017- 0791-y. 25. Poortinga W. The prevalence and clustering of four major lifestyle risk 47. Pavkovic B, Zaric M, Markovic M, Klacar M, Huljic A, Caricic A. Double factors in an English adult population. Prev Med. 2007;44(2):124–8. screening for dual disorder, alcoholism and depression. Psychiatry Res. https:// doi. org/ 10. 1016/j. ypmed. 2006. 10. 006. 2018;270:483–9. https:// doi. org/ 10. 1016/j. psych res. 2018. 10. 013. 26. Demarest S, Van der Heyden J, Charafeddine R, Drieskens S, Gisle 48. De Wit LM, Van Straten A, Van Herten M, Penninx BW, Cuijpers P. Depres- L, Tafforeau J. Methodological basics and evolution of the Belgian sion and body mass index, a u-shaped association. BMC Public Health. Health Interview Survey 1997–2008. Arch Public Health. 2013;71(1):24. 2009;9:14. https:// doi. org/ 10. 1186/ 1471- 2458-9- 14. https:// doi. org/ 10. 1186/ 0778- 7367- 71- 24. 49. Meader N, King K, Moe-Byrne T, Wright K, Graham H, Petticrew M, et al. A 27 Goldberg DP. User’s guide to the General Health Questionnaire. Windsor: systematic review on the clustering and co-occurrence of multiple risk NFER-Nelson; 1988. behaviours. BMC Public Health. 2016;16:657. https:// doi. org/ 10. 1186/ 28. Ware JE Jr, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: s12889- 016- 3373-6. construction of scales and preliminary tests of reliability and validity. 50. Saneei P, Esmaillzadeh A, Hassanzadeh Keshteli A, Reza Roohafza H, Afshar Med Care. 1996;34(3):220–33. https:// doi. org/ 10. 1097/ 00005 650- 19960 H, Feizi A, et al. Combined healthy lifestyle is inversely associated with 3000- 00003. psychological disorders among adults. PLoS ONE. 2016;11(1):e0146888. 29. Cantril H. Pattern of human concerns. New Brunswick: Rutgers University https:// doi. org/ 10. 1371/ journ al. pone. 01468 88. Press; 1965. 51. Bonnet F, Irving K, Terra J-L, Nony P, Berthezène F, Moulin P. Anxiety and 30. Kroenke K, Spitzer RL, Williams JB, Löwe B. The patient health question- depression are associated with unhealthy lifestyle in patients at risk of naire somatic, anxiety, and depressive symptom scales: a systematic cardiovascular disease. Atherosclerosis. 2005;178(2):339–44. https:// doi. review. Gen Hosp Psychiatry. 2010;32(4):345–59. https:// doi. org/ 10. 1016/j. org/ 10. 1016/j. ather oscle rosis. 2004. 08. 035. genho sppsy ch. 2010. 03. 006. 52. Loprinzi PD, Mahoney S. Concurrent occurrence of multiple positive 31. Löwe B, Decker O, Müller S, Brähler E, Schellberg D, Herzog W, et al. lifestyle behaviors and depression among adults in the United States. J Validation and standardization of the Generalized Anxiety Disorder Aec ff t Disord. 2014;165:126–30. https:// doi. org/ 10. 1016/j. jad. 2014. 04. 073. Screener (GAD-7) in the general population. Med Care. 2008;46(3):266– 53. Yang H, Gao J, Wang T, Yang L, Liu Y, Shen Y, et al. Association between 74. https:// doi. org/ 10. 1097/ MLR. 0b013 e3181 60d093. adverse mental health and an unhealthy lifestyle in rural-to-urban 32. Gisle L, Drieskens S, Demarest S, Van der Heyden J. Enquête de santé migrant workers in Shanghai. J Formos Med Assoc. 2017;116(2):90–8. 2018: Santé mentale. Bruxelles: Sciensano; 2018. (https:// www. enque https:// doi. org/ 10. 1016/j. jfma. 2016. 03. 004. tesan te. be. Available from. Numéro de rapport: D/2020/14.440/3). 54. Oftedal S, Kolt GS, Holliday EG, Stamatakis E, Vandelanotte C, Brown 33 Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey WJ, et al. Associations of health-behavior patterns, mental health and (SF-36). I. Conceptual framework and item selection. Medical Care. self-rated health. Prev Med. 2019;118:295–303. https:// doi. org/ 10. 1016/j. 1992;30(6):473–83.ypmed. 2018. 11. 017. 34 Braunholtz S, Davidson S, Myant K, O’Connor R. Well? What do you think?: 55. Conry MC, Morgan K, Curry P, McGee H, Harrington J, Ward M, et al. The the Third National Scottish Survey of Public Attitudes to Mental Health, clustering of health behaviours in Ireland and their relationship with Mental Wellbeing and Mental Health Problems. Scotland: Scottish Gov- mental health, self-rated health and quality of life. BMC Public Health. ernment Edinburgh; 2007. 2011;11(1):692. https:// doi. org/ 10. 1186/ 1471- 2458- 11- 692. 35. Van Lente E, Barry MM, Molcho M, Morgan K, Watson D, Harrington J, et al. 56. Buttery AK, Mensink GB, Busch MA. Healthy behaviours and mental Measuring population mental health and social well-being. Int J Public health: findings from the German Health Update (GEDA). The European Health. 2012;57(2):421–30. https:// doi. org/ 10. 1007/ s00038- 011- 0317-x. Journal of Public Health. 2015;25(2):219–25. https:// doi. org/ 10. 1093/ 36 de Bruin A, Picavet HSJ, Nossikov A. Health interview surveys: towards eurpub/ cku094. international harmonization of methods and instruments. Copenhagen: 57 Ahmed NJ, Husen AZ, Khoshnaw N, Getta HA, Hussein ZS, Yassin AK, et al. World Health Organization. Regional Office for Europe; 1996. The effects of smoking on IgE, oxidative stress and haemoglobin concen- 37. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression tration. Asian Pac J Cancer Prev. 2020;21(4):1069–72. https:// doi. org/ 10. severity measure. J Gen Intern Med. 2001;16(9):606–13.31557/ APJCP. 2020. 21.4. 1069. 38. Superior Health Council. Dietary guidelines for the Belgian adult popula- 58. Langley MR, Yoon H, Kim HN, Choi C-I, Simon W, Kleppe L, et al. High fat tion. Report 9284. Brussels: Superior Health Council; 2019. diet consumption results in mitochondrial dysfunction, oxidative stress, 39. World Health Organization. WHO guidelines on physical activity and and oligodendrocyte loss in the central nervous system. Biochim Biophys sedentary behaviour. Geneva: World Health Organization; 2020. Acta Mol Basis Dis. 2020;1866(3):165630. https:// doi. org/ 10. 1016/j. bbadis. 40. World Health Organization. HEARTS Technical package for cardiovascular 2019. 165630. disease management in primary health care: healthy-lifestyle counselling. 59. Tan HK, Yates E, Lilly K, Dhanda AD. Oxidative stress in alcohol-related liver Geneva: World Health Organization; 2018. disease. World J Hepatol. 2020;12(7):332–49. https:// doi. org/ 10. 4254/ wjh. 41 World Health Organization. Country profiles on nutrition, physical activity v12. i7. 332. and obesity in the 28 European Union Member States of the WHO Euro- 60. Martens DS, Nawrot TS. Air pollution stress and the aging phenotype: the pean Region. Copenhagen: WHO Regional Office for Europe; 2013. telomere connection. Curr Environ Health Rep. 2016;3(3):258–69. https:// 42 Bhaskaran K, dos-Santos-Silva I, Leon DA, Douglas IJ, Smeeth L. Associa-doi. org/ 10. 1007/ s40572- 016- 0098-8. tion of BMI with overall and cause-specific mortality: a population- 61. Martens DS, Van Der Stukken C, Derom C, Thiery E, Bijnens EM, Nawrot based cohort study of 3.6 million adults in the UK. Lancet Diabetes TS. Newborn telomere length predicts later life telomere length: track- Endocrinol. 2018;6(12):944–53. https:// doi. org/ 10. 1016/ S2213- 8587(18) ing telomere length from birth to child- and adulthood. EBioMedicine. 30288-2. 2021;63:103164. https:// doi. org/ 10. 1016/j. ebiom. 2020. 103164. 43. Benetou V, Kanellopoulou A, Kanavou E, Fotiou A, Stavrou M, Richardson 62. Hang D, Nan H, Kværner AS, De Vivo I, Chan AT, Hu Z, et al. Longitudinal C, et al. Diet-related behaviors and diet quality among school-aged associations of lifetime adiposity with leukocyte telomere length and adolescents living in Greece. Nutrients. 2020;12(12):3804. https:// doi. org/ mitochondrial DNA copy number. Eur J Epidemiol. 2018;33(5):485–95. 10. 3390/ nu121 23804.https:// doi. org/ 10. 1007/ s10654- 018- 0382-z. 44. Martens DS, Plusquin M, Gyselaers W, De Vivo I, Nawrot TS. Maternal pre- 63. Gu Y, Honig LS, Schupf N, Lee JH, Luchsinger JA, Stern Y, et al. Mediterra- pregnancy body mass index and newborn telomere length. BMC Med. nean diet and leukocyte telomere length in a multi-ethnic elderly popu- 2016;14(1):148. https:// doi. org/ 10. 1186/ s12916- 016- 0689-0. lation. Age. 2015;37(2):24. https:// doi. org/ 10. 1007/ s11357- 015- 9758-0. 45. Janssen BG, Munters E, Pieters N, Smeets K, Cox B, Cuypers A, et al. Pla- 64. Sellami M, Al-Muraikhy S, Al-Jaber H, Al-Amri H, Al-Mansoori L, Mazloum cental mitochondrial DNA content and particulate air pollution during in NA, et al. Age and sport intensity-dependent changes in cytokines and utero life. Environ Health Perspect. 2012;120(9):1346–52. https:// doi. org/ telomere length in elite athletes. Antioxidants. 2021;10(7):1035. https:// 10. 1289/ ehp. 11044 58.doi. org/ 10. 3390/ antio x1007 1035. Haut ekiet et al. BMC Medicine (2022) 20:328 Page 13 of 13 65. Savela S, Saijonmaa O, Strandberg TE, Koistinen P, Strandberg AY, Tilvis RS, patients with major depressive disorder. Psychiatr Genet. 2022;32(1):34–6. et al. Physical activity in midlife and telomere length measured in old age. https:// doi. org/ 10. 1097/ ypg. 00000 00000 000305. Exp Gerontol. 2013;48(1):81–4. https:// doi. org/ 10. 1016/j. exger. 2012. 02. 84. Aas M, Elvsåshagen T, Westlye LT, Kaufmann T, Athanasiu L, Djurovic S, 003. et al. Telomere length is associated with childhood trauma in patients 66. Mirabello L, Huang WY, Wong JY, Chatterjee N, Reding D, David Crawford with severe mental disorders. Transl Psychiatry. 2019;9(1):97. https:// doi. E, et al. The association between leukocyte telomere length and cigarette org/ 10. 1038/ s41398- 019- 0432-7. smoking, dietary and physical variables, and risk of prostate cancer. Aging 85. Birkenæs V, Elvsåshagen T, Westlye LT, Høegh MC, Haram M, Werner Cell. 2009;8(4):405–13. https:// doi. org/ 10. 1111/j. 1474- 9726. 2009. 00485.x. MCF, et al. Telomeres are shorter and associated with number of suicide 67. Sun Q, Shi L, Prescott J, Chiuve SE, Hu FB, De Vivo I, et al. Healthy attempts in affective disorders. J Aec ff t Disord. 2021;295:1032–9. https:// lifestyle and leukocyte telomere length in US women. PLoS ONE. doi. org/ 10. 1016/j. jad. 2021. 08. 135. 2012;7(5):e38374. https:// doi. org/ 10. 1371/ journ al. pone. 00383 74. 86. Phillips AC, Robertson T, Carroll D, Der G, Shiels PG, McGlynn L, et al. Do 68. Sindi S, Solomon A, Kåreholt I, Hovatta I, Antikainen R, Hänninen T, symptoms of depression predict telomere length? Evidence from the et al. Telomere length change in a multidomain lifestyle intervention West of Scotland Twenty-07 Study. Psychosom Med. 2013;75(3):288–96. to prevent cognitive decline: a randomized clinical trial. J Gerontol A. https:// doi. org/ 10. 1097/ PSY. 0b013 e3182 89e6b5. 2021;76(3):491–8. https:// doi. org/ 10. 1093/ gerona/ glaa2 79. 87. Incollingo Rodriguez AC, Polcari JJ, Nephew BC, Harris R, Zhang C, Mur- 69. Paltoglou G, Raftopoulou C, Nicolaides NC, Genitsaridi SM, Karampatsou gatroyd C, et al. Acculturative stress, telomere length, and postpartum SI, Papadopoulou M, et al. A comprehensive, multidisciplinary, personal- depression in Latinx mothers. J Psychiatr Res. 2022;147:301–6. https:// doi. ized, lifestyle intervention program is associated with increased leukocyte org/ 10. 1016/j. jpsyc hires. 2022. 01. 063. telomere length in children and adolescents with overweight and 88. Darrow SM, Verhoeven JE, Révész D, Lindqvist D, Penninx BW, Delucchi obesity. Nutrients. 2021;13(8):2682. https:// doi. org/ 10. 3390/ nu130 82682. KL, et al. The association between psychiatric disorders and telomere 70. Chakravarty EF, Hubert HB, Krishnan E, Bruce BB, Lingala VB, Fries JF. Life- length: a meta-analysis involving 14,827 persons. Psychosom Med. style risk factors predict disability and death in healthy aging adults. Am J 2016;78(7):776–87. https:// doi. org/ 10. 1097/ PSY. 00000 00000 000356. Med. 2012;125(2):190–7. https:// doi. org/ 10. 1016/j. amjmed. 2011. 08. 006. 89. Simon NM, Walton ZE, Bui E, Prescott J, Hoge E, Keshaviah A, et al. 71. Bosnes I, Nordahl HM, Stordal E, Bosnes O, Myklebust TÅ, Almkvist O. Life- Telomere length and telomerase in a well-characterized sample of indi- style predictors of successful aging: a 20-year prospective HUNT study. viduals with major depressive disorder compared to controls. Psychoneu- PLoS ONE. 2019;14(7):e0219200. https:// doi. org/ 10. 1371/ journ al. pone. roendocrinology. 2015;58:9–22. https:// doi. org/ 10. 1016/j. psyne uen. 2015. 02192 00.04. 004. 72. Epel ES, Prather AA. Stress, telomeres, and psychopathology: toward a 90. Schaakxs R, Verhoeven JE, Voshaar RCO, Comijs HC, Penninx BW. Leuko- deeper understanding of a triad of early aging. Annu Rev Clin Psychol. cyte telomere length and late-life depression. Am J Geriatr Psychiatry. 2018;14:371–97. https:// doi. org/ 10. 1146/ annur ev- clinp sy- 032816- 045054. 2015;23(4):423–32. https:// doi. org/ 10. 1016/j. jagp. 2014. 06. 003. 73. Verhoeven JE, Révész D, Epel ES, Lin J, Wolkowitz OM, Penninx BW. Major 91. de Baumont AC, Hoffmann MS, Bortoluzzi A, Fries GR, Lavandoski P, Grun depressive disorder and accelerated cellular aging: results from a large LK, et al. Telomere length and epigenetic age acceleration in adolescents psychiatric cohort study. Mol Psychiatry. 2014;19(8):895–901. https:// doi. with anxiety disorders. Sci Rep. 2021;11(1):7716. https:// doi. org/ 10. 1038/ org/ 10. 1038/ mp. 2013. 151.s41598- 021- 87045-w. 74. Han LK, Verhoeven JE, Tyrka AR, Penninx BW, Wolkowitz OM, Månsson 92. Schutte NS, Malouff JM. The association between depression and leuko - KN, et al. Accelerating research on biological aging and mental health: cyte telomere length: a meta-analysis. Depress Anxiety. 2015;32(4):229– current challenges and future directions. Psychoneuroendocrinology. 38. https:// doi. org/ 10. 1002/ da. 22351. 2019;106:293–311. https:// doi. org/ 10. 1016/j. psyne uen. 2019. 04. 004. 93. Drieskens S, Demarest S, Bel S, De Ridder K, Tafforeau J. Correction of 75. Kim MY, Lee JW, Kang HC, Kim E, Lee DC. Leukocyte mitochondrial DNA self-reported BMI based on objective measurements: a Belgian experi- (mtDNA) content is associated with depression in old women. Arch ence. Archives of Public Health. 2018;76:10. https:// doi. org/ 10. 1186/ Gerontol Geriatr. 2011;53(2):e218–21. https:// doi. org/ 10. 1016/j. archg er. s13690- 018- 0255-7. 2010. 11. 019. 94. Gorber SC, Schofield-Hurwitz S, Hardt J, Levasseur G, Tremblay M. The 76. Mengel-From J, Thinggaard M, Dalgard C, Kyvik KO, Christensen K, accuracy of self-reported smoking: a systematic review of the relationship Christiansen L. Mitochondrial DNA copy number in peripheral blood between self-reported and cotinine-assessed smoking status. Nicotine cells declines with age and is associated with general health among Tob Res. 2009;11(1):12–24. https:// doi. org/ 10. 1093/ ntr/ ntn010. elderly. Hum Genet. 2014;133(9):1149–59. https:// doi. org/ 10. 1007/ 95. Knez J, Winckelmans E, Plusquin M, Thijs L, Cauwenberghs N, Gu Y, et al. s00439- 014- 1458-9. Correlates of peripheral blood mitochondrial DNA content in a general 77. Otsuka I, Izumi T, Boku S, Kimura A, Zhang Y, Mouri K, et al. Aberrant population. Am J Epidemiol. 2016;183(2):138–46. https:// doi. org/ 10. 1093/ telomere length and mitochondrial DNA copy number in suicide com-aje/ kwv175. pleters. Sci Rep. 2017;7(1):3176. 78. Chung JK, Lee SY, Park M, Joo E-J, Kim SA. Investigation of mitochondrial Publisher’s Note DNA copy number in patients with major depressive disorder. Psychiatry Springer Nature remains neutral with regard to jurisdictional claims in pub- Res. 2019;282:112616. https:// doi. org/ 10. 1016/j. psych res. 2019. 112616. lished maps and institutional affiliations. 79. Takahashi PY, Jenkins GD, Welkie BP, McDonnell SK, Evans JM, Cerhan JR, et al. Association of mitochondrial DNA copy number with self-rated health status. Appl Clin Genet. 2018;11:121–7. https:// doi. org/ 10. 2147/ TACG. S1676 40. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : 80. Vyas CM, Ogata S, Reynolds CF 3rd, Mischoulon D, Chang G, Cook NR, et al. Lifestyle and behavioral factors and mitochondrial DNA copy fast, convenient online submission number in a diverse cohort of mid-life and older adults. PLoS ONE. thorough peer review by experienced researchers in your field 2020;15(8):e0237235. https:// doi. org/ 10. 1371/ journ al. pone. 02372 35. 81. Verhoeven JE, Révész D, van Oppen P, Epel ES, Wolkowitz OM, Penninx rapid publication on acceptance BW. Anxiety disorders and accelerated cellular ageing. Br J Psychiatry. support for research data, including large and complex data types 2015;206(5):371–8. https:// doi. org/ 10. 1192/ bjp. bp. 114. 151027. • gold Open Access which fosters wider collaboration and increased citations 82. Pisanu C, Vitali E, Meloni A, Congiu D, Severino G, Ardau R, et al. Inves- tigating the role of leukocyte telomere length in treatment-resistant maximum visibility for your research: over 100M website views per year depression and in response to electroconvulsive therapy. J Pers Med. 2021;11(11):1100. https:// doi. org/ 10. 3390/ jpm11 111100. At BMC, research is always in progress. 83. da Silva RS, de Moraes LS, da Rocha CAM, Ferreira-Fernandes H, Yoshioka Learn more biomedcentral.com/submissions FKN, Rey JA, et al. Telomere length and telomerase activity of leukocytes as biomarkers of selective serotonin reuptake inhibitor responses in http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Medicine Springer Journals

A healthy lifestyle is positively associated with mental health and well-being and core markers in ageing

Download PDF
13 pages

Loading next page...
 
/lp/springer-journals/a-healthy-lifestyle-is-positively-associated-with-mental-health-and-XTzz7faEbS

References (107)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2022
eISSN
1741-7015
DOI
10.1186/s12916-022-02524-9
Publisher site
See Article on Publisher Site

Abstract

Background: Studies often evaluate mental health and well-being in association with individual health behaviours although evaluating multiple health behaviours that co-occur in real life may reveal important insights into the overall association. Also, the underlying pathways of how lifestyle might affect our health are still under debate. Here, we studied the mediation of different health behaviours or lifestyle factors on mental health and its effect on core mark - ers of ageing: telomere length ( TL) and mitochondrial DNA content (mtDNAc). Methods: In this study, 6054 adults from the 2018 Belgian Health Interview Survey (BHIS) were included. Mental health and well-being outcomes included psychological and severe psychological distress, vitality, life satisfaction, self-perceived health, depressive and generalised anxiety disorder and suicidal ideation. A lifestyle score integrating diet, physical activity, smoking status, alcohol consumption and BMI was created and validated. On a subset of 739 participants, leucocyte TL and mtDNAc were assessed using qPCR. Generalised linear mixed models were used while adjusting for a priori chosen covariates. Results: The average age (SD) of the study population was 49.9 (17.5) years, and 48.8% were men. A one-point increment in the lifestyle score was associated with lower odds (ranging from 0.56 to 0.74) for all studied mental health outcomes and with a 1.74% (95% CI: 0.11, 3.40%) longer TL and 4.07% (95% CI: 2.01, 6.17%) higher mtDNAc. Psychological distress and suicidal ideation were associated with a lower mtDNAc of − 4.62% (95% CI: − 8.85, − 0.20%) and − 7.83% (95% CI: − 14.77, − 0.34%), respectively. No associations were found between mental health and TL. Conclusions: In this large-scale study, we showed the positive association between a healthy lifestyle and both biological ageing and different dimensions of mental health and well-being. We also indicated that living a healthy lifestyle contributes to more favourable biological ageing. Keywords: Mental health, Lifestyle, Biological ageing, Mitochondrial DNA content, Telomere length Background According to the World Health Organization (WHO), a healthy lifestyle is defined as “a way of living that low - ers the risk of being seriously ill or dying early” [1]. Public health authorities emphasise the importance of *Correspondence: [email protected] a healthy lifestyle, but despite this, many individuals Sciensano, Risk and Health Impact Assessment, Juliette Wytsmanstraat 14, worldwide still live an unhealthy lifestyle [2]. In Europe, 1050 Brussels, Belgium 26% of adults smoke [3], nearly half (46%) never exercise Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Hautekiet et al. BMC Medicine (2022) 20:328 Page 2 of 13 [4], 8.4% drink alcohol on a daily basis [5] and over half frame of the BHIS was the Belgian National Register, and (51%) are overweight [5]. These unhealthy behaviours participants were selected based on a multistage strati- have been associated with adverse health outcomes like fied sampling design including a geographical stratifica - cardiovascular diseases [6–8], respiratory diseases [9], tion and a selection of municipalities within provinces, musculoskeletal diseases [10] and, to a lesser extent, of households within municipalities and of respondents mental disorders [11, 12]. within households [26]. The study population for this Even though the association between lifestyle and cross-sectional study included 6054 BHIS participants health outcomes has been extensively investigated, bio- (see flowchart in Additional file  1: Fig. S1) [27–31]. logical mechanisms explaining these observed asso- Minors (< 18 years) and participants not eligible to com- ciations are not yet fully understood. One potential plete the mental health modules (participants who par- mechanism that can be suggested is biological ageing. ticipated through a proxy respondent, i.e. a person of Both telomere length (TL) and mitochondrial DNA con- confidence filled out the survey) were excluded (n = 2172 tent (mtDNAc) are known biomarkers of ageing. Tel- and n = 846, respectively). Furthermore, of the 8593 eligi- omeres are the end caps of chromosomes and consist of ble participants, those with missing information to cre- multiple TTA GGG s equence repeats. They protect chro - ate the mental health indicators, the lifestyle score or the mosomes from degradation and shorten with every cell covariates used in this study were excluded (n = 1642, division because of the “end-replication problem” [13]. 788 and 109, respectively). Mitochondria are crucial to the cell as they are respon- For the first time in 2018, a subset of 1184 BHIS par - sible for apoptosis, the control of cytosolic calcium lev- ticipants contributed to the 2018 Belgian Health Exami- els and cell signalling [14]. Living a healthy lifestyle can nation Survey (BELHES). All BHIS participants were be linked with healthy ageing as both TL and mtDNAc invited to participate except for minors (< 18  years), have been associated with health behaviours like obesity BHIS participants who participated through a proxy [15], diet [16], smoking [17] and alcohol abuse [18]. Fur- respondent and residents of the German Community of thermore, as biomarkers of ageing, both TL and mtD- Belgium, the latter representing 1% of the Belgian popu- NAc have been associated with age-related diseases like lation. Participants were recruited on a voluntary basis Parkinson’s disease [19], coronary heart disease [20], until the regional quotas were reached (450, 300 and atherosclerosis [21] and early mortality [22]. Also, early 350 in respectively Flanders, Brussels Capital Region mortality and higher risks for the aforementioned age- and Wallonia). These participants underwent a health related diseases are observed in psychiatric illnesses, and examination, including anthropological measurements it is suggested that advanced biological ageing underlies and completed an additional questionnaire. Also, blood these observations [23]. and urine samples were collected. Of the 6054 included Multiple studies evaluated individual health behav- BHIS participants, 909 participated in the BELHES. Par- iours, but research on the combination of these health ticipants for whom we could not calculate both TL and behaviours is limited. As they often co-occur and may mtDNAc were excluded (n = 170). More specifically, cause synergistic effects, assessing them in combination participants were excluded because they did not provide with each other rather than independently might bet- a blood sample (n = 91) or because they did not provide ter reflect the real-life situation [24, 25]. Therefore, in a permission for DNA research (n = 32). Twenty samples general adult population, we combined five commonly were excluded from DNA extraction because either total studied health behaviours including diet, smoking sta- blood volume was too low (n = 7), samples were clothed tus, alcohol consumption, BMI and physical activity (n = 1) or tubes were broken due to freezing conditions into one healthy lifestyle score to evaluate its association (n = 12). Twenty-seven samples were excluded because with mental health and well-being and biological ageing. they did not meet the biomarker quality control criteria Furthermore, we evaluated the association between the (high technical variation in qPCR triplicates). This was markers of biological ageing and mental health and well- not met for 3 TL samples, 20 mtDNAc samples and 4 being. We hypothesise that individuals living a healthy samples for both biomarkers. For this subset, we ended lifestyle have a better mental health status, a longer TL up with a final number of 739 participants. Further in this and a higher mtDNAc and that these biomarkers are pos- paper, we refer to “the BHIS subset” for the BHIS partici- itively associated with mental health and well-being. pants (n = 6054) and the “BELHES subset” for the BEL- HES participants (n = 739). Methods As part of the BELHES, this project was approved by Study population the Medical Ethics Committee of the University Hospital In 2018, 11611 Belgian residents participated in the 2018 Ghent (registration number B670201834895). The pro - Belgian Health Interview Survey (BHIS). The sampling ject was carried out in line with the recommendations of Haut ekiet et al. BMC Medicine (2022) 20:328 Page 3 of 13 the Belgian Privacy Commission. All participants have Fourthly, depressive and generalised anxiety disorders signed a consent form that was approved by the Medical were defined using respectively the Patient Health Ques - Ethics Committee. tionnaire (PHQ-9) and the Generalised Anxiety Disorder Questionnaire (GAD-7). We identified individuals who Health interview survey suffer from major depressive syndrome or any other type The BHIS is a comprehensive survey which aims to gain of depressive syndrome according to the criteria of the insight into the health status of the Belgian popula- PHQ-9 [37]. A cut-off point of + 10 on the total sum of tion. The questions on the different dimensions of men - the GAD-7 score was used to indicate generalised anxi- tal health and well-being were based on international ety disorder [31]. Additionally, a dichotomous question standardised and validated questionnaires [32], and this on suicidal ideation was used: “Have you ever seriously resulted in eight mental health outcomes that were used thought of ending your life?”; “If yes, did you have such in this study. Detailed information on each indicator thoughts in the past 12  months?”. Finally, the BHIS also score and its use is addressed in Additional file  1: Table. includes personal, socio-economic and lifestyle informa- S1. Firstly, the General Health Questionnaire (GHQ-12) tion. The standardised Cronbach’s alpha coefficients for provides the prevalence of psychological and severe psy- the PHQ-9, GHQ-12, GAD-7 and questions on vitality of chological distress in the population [27]. On the total the SF-36 ranged between 0.80 and 0.90. GHQ score, cut-off points of + 2 and + 4 were used to identify respectively psychological and severe psycho- logical distress. Healthy lifestyle score Secondly, we used two indicators for the positive We developed a healthy lifestyle score based on five dif - dimensions of mental health: vitality and life satisfaction. ferent health behaviours: body mass index (BMI), smok- Four questions of the short form health survey (SF-36) ing status, physical activity, alcohol consumption and diet indicate the participant’s vital energy level [28, 33]. We (Table 1). These health behaviours were defined as much used a cut-off point to identify participants with an opti - as possible according to the existing guidelines for healthy mal vitality score, which is a score equal to or above one living issued by the Belgian Superior Health Council standard deviation above the mean, as used in previous [38] and the World Health Organisation [39–41]. Firstly, studies [34, 35]. Life satisfaction was measured by the BMI was calculated as a person’s self-reported weight in Cantril Scale, which ranges from 0 to 10 [29]. A cut-off kilogrammes divided by the square of the person’s self- point of + 6 was used to indicate participants with high reported height in metres (kg/m ). BMI was classified or medium life satisfaction versus low life satisfaction. into four categories: underweight (BMI < 18.5  kg/m ), Thirdly, the question “How is your health in general? normal weight (BMI 18.5–24.9 kg/m ), overweight (BMI 2 2 Is it very good, good, fair, bad or very bad?” was used 25.0–29.9  kg/m ) and obese (BMI ≥ 30.0  kg/m ). Due to to assess self-perceived health, also known as self-rated a J-shaped association of BMI with the overall mortality health. Based on WHO recommendations [36], the and multiple specific causes of death, obesity and under - answer categories were dichotomised into “good to very weight were both classified as least healthy [42]. BMI good self-perceived health” and “very bad to fair self-per- was scored as follows: obese and underweight = 0, over- ceived health”. weight = 1 and normal weight = 2. Table 1 Healthy lifestyle score, where each health behaviour is scored from the least healthy to the healthiest Hautekiet et al. BMC Medicine (2022) 20:328 Page 4 of 13 Secondly, smoking status was divided into four cat- (Qiagen, N.V.V Venlo, The Netherlands). The purity and egories. Participants were categorised as regular smok- quantity of the sample were measured with a NanoDrop ers if they smoked a minimum of 4  days per week or if spectrophotometer (ND-2000; Thermo Fisher Scientific, they quit smoking less than 1 month before participation Wilmington, DE, USA). DNA integrity was assessed by (= 0). Occasional smokers were defined as smoking more agarose gel electrophoresis. To ensure a uniform DNA than once per month up to 3  days per week (= 1). Par- input of 6  ng for each qPCR reaction, samples were ™ ® ticipants were classified as former smokers if they quit diluted and checked using the Quant-iT PicoGreen smoking at least 1  month before the questionnaire or if dsDNA Assay Kit (Life Technologies, Europe). they smoked less than once a month (= 2). The final cat - Relative TL and mtDNAc were measured in triplicate egory included people who never smoked (= 3). using a previously described quantitative real-time PCR Thirdly, physical activity was assessed by the question: (qPCR) assay with minor modifications [44, 45]. All reac- “What describes best your leisure time activities dur- tions were performed on a 7900HT Fast Real-Time PCR ing the last year?”. Four categories were established and System (Applied Biosystems, Foster City, CA, USA) in scored as follows: sedentary activities (= 0), light activi- a 384-well format. Used telomere, mtDNAc and single ties less than 4  h/week (= 1), light activities more than copy-gene reaction mixtures and PCR cycles are given 4  h/week or recreational sport less than 4  h/week (= 2) in Additional file  1: Text. S1. Reaction efficiency was and recreational sport more than 4 h or intense training assessed on each plate by using a 6-point serial dilution (= 3). Fourthly, information on the number of alcoholic of pooled DNA. Efficiencies ranged from 90 to 100% for drinks per week was used to categorise alcohol consump- single-copy gene runs, 100 to 110% for telomere runs tion. The different categories were set from high to low and 95 to 105% for mitochondrial DNA runs. Six inter- alcohol consumption: 22 drinks or more/week (= 0), run calibrators (IRCs) were used to account for inter-run 15–21 drinks/week (= 1), 8–14 drinks/week (= 2), 1–7 variability. Also, non-template controls were used in each drinks/week (= 3)and less than 1 drink/week (= 4). run. Raw data were processed and normalised to the ref- Finally, in line with the research by Benetou et  al., a erence gene using the qBase plus software (Biogazelle, diet score was calculated using the frequency of consum- Zwijnaarde, Belgium), taking into account the run-to-run ing fruit, vegetables, snacks and sodas [43]. For fruit as differences. well as vegetable consumption, the frequency was scored Leucocyte telomere length was expressed as the ratio as follows: never (= 0), < 1/week (= 1), 1–3/week (= 2), of telomere copy number to single-copy gene num- 4–6/week (= 3) and ≥ 1/day (= 4). The frequency of con - ber (T/S) relative to the mean T/S ratio of the entire suming snacks and sodas was scored as follows: never study population. Leucocyte mtDNAc was expressed as (= 4), < 1/week (= 3), 1–3/week (= 2), 4–6/week (= 1) the ratio of mtDNA copy number to single-copy gene and ≥ 1/day (= 0). The diet score was then divided into number (M/S) relative to the mean M/S ratio of the tertiles, in line with the research by Benetou et al. [43]. A entire study population. The reliability of our assay was diet score of 0–9 points was classified as the least healthy assessed by calculating the interclass correlation coef- behaviour (= 0). A diet score ranging from 10 to 12 made ficient (ICC) of the triplicate measures (T/S and M/S up the middle category (= 1), and a score from 13 to 16 ratios and T, M and S separately) as proposed by the Tel- was classified as the healthiest behaviour (= 2). omere Research Network, using RStudio version 1.1.463 All five previously described health behaviours were (RStudio PBC, Boston, MA, USA). The intra-plate ICCs combined into one healthy lifestyle score (Table  1). The of T/S ratios, TL runs, M/S ratios, mtDNAc runs and sum of the scores obtained for each health behaviour single-copy runs were respectively 0.804 (p < 0.0001), indicated the absolute lifestyle score. To calculate the 0.907 (p < 0.0001), 0.815 (p < 0.0001), 0.916 (p < 0.0001) relative lifestyle score, each absolute scored health behav- and 0.781 (p < 0.0001). Based on the IRCs, the inter-plate iour was given equal weight by recalculating its maxi- ICC was 0.714 (p < 0.0001) for TL and 0.762 (p < 0.0001) mum absolute score to a relative score of 1. The relative for mtDNAc. lifestyle scores were then summed up to achieve a final continuous lifestyle score, ranging from 0 to 5, with a Statistical analysis higher score representing a healthier lifestyle. Statistical analyses were performed using the SAS soft- ware (version 9.4; SAS Institute Inc., Cary, NC, USA). Telomere length and mitochondrial DNA content assay We performed a log(10) transformation of the TL and Blood samples were collected during the BELHES and mtDNAc data to reduce skewness and to better approxi- centrifuged for 15  min at 3000  rpm before storage mate a normal distribution. Three analyses were done: at − 80 °C. After extracting the buffy coat from the blood (1) In the BHIS subset (n = 6054), we evaluated the asso- sample, DNA was isolated using the QIAgen Mini Kit ciation between the lifestyle score and the mental health Haut ekiet et al. BMC Medicine (2022) 20:328 Page 5 of 13 and well-being outcomes (separately). These results are Results presented as the odds ratio (95% CI) of having a mental Population characteristics health condition or disorder for a one-point increment The characteristics of the BHIS and BELHES subset are in the lifestyle score. (2) In the BELHES subset (n = 739), presented in Table  2. In the BHIS subset, 48.8% of the we evaluated the association between the lifestyle score participants were men. The average age (SD) was 49.9 and both TL and mtDNAc (separately). These results are (17.5) years, and most participants were born in Belgium presented as the percentage difference in TL or mtD - (79.5%). The highest educational level in the household NAc (95% CI) for a one-point increment in the lifestyle was most often college or university degree (53.3%), and score. (3) In the BELHES subset (n = 739), we evaluated the most common household composition was couple the association between the mental health and well-being with child(ren) (37.7%). The proportion of participants in outcomes and both TL and mtDNAc (separately). These different regions of Belgium, i.e. Flanders, Brussels Capi - results are presented as the percentage difference in TL tal Region and Wallonia, was respectively 41.1%, 23.3% or mtDNAc (95% CI) when having a mental health condi- and 35.6%. For the BELHES subset, we found similar tion or disorder compared with the healthy group. results except for region and education. We noticed more For all three analyses, we performed multivariable lin- participants from Flanders and more participants with a ear mixed models (GLIMMIX; unstructured covariance high educational level in the household. The mean (SD) matrix) taking into account a priori selected covariates relative TL and mtDNAc were respectively 1.04 (0.23) including age (continuous), sex (male, female), region and 1.03 (0.24). TL and mtDNAc were positively corre- (Flanders, Brussels Capital Region, Wallonia), highest lated (Spearman’s correlation = 0.21, p < 0.0001). educational level of the household (up to lower second- We compared (1) the characteristics of the 6054 ary, higher secondary, college or university), country of eligible BHIS participants that were included in the birth (Belgium, EU, non-EU) and household type (single, BHIS subset with the 2539 eligible participants that one parent with child, couple without child, couple with were excluded from the BHIS subset (Additional file  1: child, others). To capture the non-linear effect of age, we Table  S2) and (2) the 739 participants from the BHIS included a quadratic term when the result of the analy- subset that were included in the BELHES subset with sis showed that both the linear and quadratic terms had the 5315 participants that were excluded from the BEL- a p-value < 0.1. For the two analyses on TL and mtDNAc, HES subset (Additional file  1: Table  S3). Except for sex we additionally adjusted for the date of participation in and nationality in the latter, all other covariates showed the BELHES. As multiple members of one household differences between the included and excluded groups. participated, we added household numbers in the ran- On the other hand, population data from 2018 indicates dom statement. that the average age (SD) of the adult Belgian population Bivariate analyses evaluating the associations between was 49.5 (18.9) with a distribution over Flanders, Brus- the characteristics and TL, mtDNAc, the lifestyle score sels Capital Region and Wallonia of respectively 58.2%, or psychological distress as a parameter of mental health 10.2% and 31.6% and that 48.7% were men. The distribu - and well-being are evaluated based on the same model. tion of our sample according to age and sex thus largely The chi-squared tests (categorical data) and t-tests (con - corresponds to the age and sex distribution of the adult tinuous data) were used to evaluate the characteristics of Belgian population figures. The large difference in the included and excluded participants. The lifestyle score regional distribution is due to the oversampling of the was validated by creating a ROC curve and calculating Brussels Capital Region in the BHIS. the area under the curve (AUC) of the adjusted associa- Bivariate associations evaluating the characteris- tion between the lifestyle score and self-perceived health. tics with TL, mtDNAc, the lifestyle score or psycho- Adjustments were made for age, sex, region, highest logical distress as a parameter of mental health are educational level of the household, country of birth and presented in Additional file  1: Table S4. Briefly, men had household type. a − 6.41% (95% CI: − 9.10 to − 3.65%, p < 0.0001) shorter In a sensitivity analysis, to evaluate the robustness of TL, a − 8.03% (95% CI: − 11.00 to − 4.96%, p < 0.0001) our findings, we additionally adjusted our main models lower mtDNAc, lower odds of psychological distress separately for perceived quality of social support (poor, (OR = 0.59, 95% CI: 0.53 to 0.66, p < 0.0001) and a life- moderate, strong) and chronic disease (suffering from style score of − 0.28 (95% CI: − 0.32 to − 0.24, p < 0.0001) any chronic disease or condition: yes, no). The third points less compared with women. Furthermore, a 1-year model, evaluating the biomarkers with the mental health increment in age was associated with a − 0.64% (− 0.73 outcomes, was also additionally adjusted for the lifestyle to − 0.55%, p < 0.0001) shorter TL and a − 0.19% (95% score. CI: − 0.31 to − 0.08%, p = 0.00074) lower mtDNAc. Hautekiet et al. BMC Medicine (2022) 20:328 Page 6 of 13 Table 2 Characteristics of the study population for the BHIS (n = 6054) and the BELHES subset (n = 739) Characteristics BHIS subset, n (%) or mean (SD) BELHES subset, n (%) or mean (SD) Male 2955 (48.8%) 369 (49.9%) Age, years 49.9 (17.5) 48.3 (15.5) Region Flanders 2488 (41.1%) 356 (48.2%) Brussels Capital Region 1410 (23.3%) 158 (21.4%) Wallonia 2156 (35.6%) 225 (30.5%) Highest educational level in the household Up to lower secondary school 1010 (16.7%) 92 (12.5%) Higher secondary school 1819 (30.1%) 196 (26.5%) College or university 3225 (53.3%) 451 (61.0%) Household composition Single 1339 (22.1%) 130 (17.6%) One parent with a child 514 (8.5%) 53 (7.2%) Couple without child 1674 (27.7%) 196 (26.5%) Couple with child(ren) 2283 (37.7%) 326 (44.1%) Others 244 (4.0%) 34 (4.6%) Country of birth Belgium 4812 (79.5%) 596 (80.7%) EU 619 (10.2%) 77 (10.4%) Non-EU 623 (10.3%) 66 (8.9%) Perceived quality of social support Poor 946 (15.7%) 116 (15.9%) Moderate 2978 (49.5%) 379 (51.9%) Strong 2093 (34.8%) 236 (32.3%) Chronic disease or condition 1776 (29.5%) 206 (28.1%) n = 6017 and 731 for the BHIS and BELHES subset, respectively n = 6017 and 733 for the BHIS and BELHES subset, respectively Mental health prevalence and lifestyle characteristics to have had suicidal thoughts in the past 12 months. Sim- Within the BHIS subset, 32.3% and 18.0% of the par- ilar results were found for the BELHES subset (Table 3). ticipants had respectively psychological and severe psy- Within the BHIS subset, the average lifestyle score chological distress. 86.7% had suboptimal vitality, 12.0% (SD) was 3.1 (0.9) (Table 4). A histogram of the lifestyle indicated low life satisfaction and 22.0% had very bad to score is shown in Additional file  1: Fig. S2. 16.6% were fair self-perceived health. The prevalence of depressive regular smokers, and 4.9% reported 22 alcoholic drinks and generalised anxiety disorders was respectively 9.0% per week or more. 29.7% reported that their main lei- and 10.8%, respectively. 4.4% of the participants indicated sure time included mainly sedentary activities, and Table 3 Prevalence of the mental health and well-being outcomes for the BHIS (n = 6054) and the BELHES subset (n = 739) Mental health and well-being outcomes BHIS subset, n (%) BELHES subset, n (%) Psychological distress 1954 (32.3%) 252 (34.1%) Severe psychological distress 1087 (18.0%) 131 (17.7%) Suboptimal vitality 5247 (86.7%) 661 (89.5%) Low life satisfaction 726 (12.0%) 82 (11.1%) Very bad to fair self-perceived health 1333 (22.0%) 135 (18.3%) Depressive disorder 544 (9.0%) 63 (8.5%) Generalised anxiety disorder 655 (10.8%) 76 (10.3%) Suicidal ideation 269 (4.4%) 38 (5.1%) Haut ekiet et al. BMC Medicine (2022) 20:328 Page 7 of 13 Table 4 Characteristics of the healthy lifestyle score for the BHIS subset (n = 6054) and the BELHES subset (n = 739) BHIS subset, n (%) or mean (SD) BELHES subset, n (%) or mean (SD) Lifestyle score 3.1 (0.9) 3.1 (0.9) Smoking status Regular smoker (0) 1002 (16.6%) 129 (17.5%) Occasional smoker (1) 159 (2.6%) 25 (3.4%) Former smoker (2) 1472 (24.3%) 180 (24.4%) Nonsmoker (3) 3421 (56.5%) 405 (54.8%) BMI Underweight/obese (0) 1123 (18.6%) 119 (16.1%) Overweight (1) 2073 (34.2%) 248 (33.6%) Normal weight (2) 2858 (47.2%) 372 (50.3%) Physical activity Sedentary activities (0) 1795 (29.7%) 185 (25.0%) Light activities < 4 h/week (1) 1391 (23.0%) 167 (22.6%) Light activities > 4 h/week or sport < 4 h/week (2) 1820 (30.1%) 227 (30.7%) Sport > 4 h/week or intense training (3) 1048 (17.3%) 160 (21.7%) Alcohol consumption ≥ 22 drinks/week (0) 294 (4.9%) 31 (4.2%) 15–21 drinks/week (1) 323 (5.3%) 52 (7.0%) 8–14 drinks/week (2) 738 (12.2%) 101 (13.7%) 1–7 drinks/week (3) 1777 (29.4%) 235 (31.8%) < 1 drink/week or abstainers (4) 2922 (48.3%) 320 (43.3%) Diet Diet score 0–9 (0) 1769 (29.2%) 210 (28.4%) Diet score 10–12 (1) 2621 (43.3%) 325 (44.0%) Diet score 13–16 (2) 1664 (27.5%) 204 (27.6%) 18.6% were underweight or obese. 29.2% were classified Table 5 Associations between the lifestyle score and the mental as having an unhealthy diet score. The participants of health and well-being outcomes the BELHES subset were slightly more active, but no Lifestyle score OR 95% CI p-value other dissimilarities were found (Table  4). The ROC curve shows an area under the curve (AUC) of 0.74, Psychological distress 0.74 0.69, 0.79 < 0.0001 indicating a 74% predictive accuracy for the lifestyle Severe psychological distress 0.69 0.64, 0.75 < 0.0001 score as a self-perceived health predictor (Additional Suboptimal vitality 0.62 0.56, 0.68 < 0.0001 file 1 : Fig. S3). Low life satisfaction 0.62 0.56, 0.68 < 0.0001 Very bad to fair self-perceived health 0.56 0.52, 0.61 < 0.0001 Depressive disorder 0.57 0.51, 0.63 < 0.0001 Healthy lifestyle and mental health and well-being Generalised anxiety disorder 0.63 0.57, 0.69 < 0.0001 Living a healthier lifestyle, indicated by having a higher Suicidal ideation 0.63 0.55, 0.72 < 0.0001 lifestyle score, was associated with lower odds of all Odds ratios (OR) and 95% confidence intervals (CI) for the different mental mental health and well-being outcomes (Table 5). After health and well-being outcomes for a one-point increment in the lifestyle score. Analyses were adjusted for age, sex, region, highest educational level in the adjustment, a one-point increment in the lifestyle household, household composition and country of birth score was associated with lower odds of psychological (OR = 0.74, 95% CI: 0.69, 0.79) and severe psychologi- cal distress (OR = 0.69, 95% CI: 0.64, 0.75). Similarly, 0.62 (95% CI: 0.56, 0.68) and 0.56 (95% CI: 0.52, 0.61). for the same increment, the odds of suboptimal vitality, Finally, the odds of having depressive disorder, general- low life satisfaction and very bad to fair self-perceived ised anxiety disorder or suicidal ideation were respec- health were respectively 0.62 (95% CI: 0.56, 0.68), tively 0.57 (95% CI: 0.51, 0.63), 0.63 (95% CI: 0.57, 0.69) Hautekiet et al. BMC Medicine (2022) 20:328 Page 8 of 13 Table 6 Associations between the biomarkers and both the lifestyle score and the mental health and well-being outcomes Lifestyle score/mental health and TL mtDNAc well-being outcome % difference 95% CI p-value % difference 95% CI p-value Lifestyle score 1.74 0.11, 3.40 0.037 4.07 2.01, 6.17 0.00012 Psychological distress − 0.12 − 3.04, 2.88 0.94 − 2.29 − 5.82, 1.36 0.21 Severe psychological distress 0.21 − 3.39, 3.94 0.91 − 4.62 − 8.85, − 0.20 0.041 Suboptimal vitality − 3.26 − 7.46, 1.12 0.14 − 2.37 − 7.62, 3.17 0.39 Low life satisfaction 0.20 − 4.18, 4.79 0.93 − 2.00 − 7.30, 3.59 0.47 Very bad to fair self-perceived health − 0.79 − 4.36, 2.90 0.67 − 2.53 − 6.87, 2.00 0.27 Depressive disorder 2.54 − 2.41, 7.73 0.32 2.84 − 3.28, 9.36 0.37 Generalised anxiety disorder 0.56 − 3.88, 5.20 0.81 2.05 − 3.52, 7.94 0.48 Suicidal ideation 0.73 − 5.44, 7.31 0.82 − 7.83 − 14.77, − 0.34 0.041 Difference (%) in relative telomere length (TL) and mitochondrial DNA content (mtDNAc) (with 95% CI) (1) for a one-point increment in the lifestyle score or (2) when having a mental health disorder or condition compared with the healthy group. Analyses were adjusted for age, sex, region, highest educational level in the household, household composition, country of birth and day of sample collection and 0.63 (95% CI: 0.55, 0.72) for a one-point increment telomere length and mitochondrial DNA content. Having in the lifestyle score. a healthy lifestyle was positively associated with all men- tal health and well-being indicators and the markers of The biomarkers of ageing biological ageing. Furthermore, having had suicidal idea- After adjustment, living a healthy lifestyle was positively tion or suffering from severe psychological distress was associated with both TL and mtDNAc (Table  6). A one- associated with a lower mtDNAc. However, no associa- point increment in the lifestyle score was associated with tion was found between mental health and TL. a 1.74 (95% CI: 0.11, 3.40%, p = 0.037) higher TL and a 4.07 (95% CI: 2.01, 6.17%, p = 0.00012) higher mtDNAc. Healthy lifestyle and mental health and well-being People suffering from severe psychological distress had In the first part of this research, we evaluated the associa - a − 4.62% (95% CI: − 8.85, − 0.20%, p = 0.041) lower mtD- tion between lifestyle and mental health and well-being and NAc compared with those who did not suffer from severe showed that living a healthy lifestyle was positively associ- psychological distress. Similarly, people with suicidal ide- ated with better mental health and well-being outcomes. ation had a − 7.83% (95% CI: − 14.77, − 0.34%, p = 0.041) Similar trends were found in previous studies for each of lower mtDNAc compared with those without suicidal the health behaviours separately [11, 12, 46–48]. Although ideation. No associations were found for the other men- evaluating these health behaviours separately provides tal health and well-being outcomes, and no associations valuable information, assessing them in combination with were found between mental health and TL (Table 6). each other rather than independently might better reflect the real-life situation as they often co-occur and may exert a synergistic effect on each other [24, 25, 49]. For exam- ple, 68% of the adults in England engaged in two or more Sensitivity analysis unhealthy behaviours [25]. Especially, smoking status and Additional adjustment of the main analyses for perceived alcohol consumption co-occurred, but half of the stud- quality of social support, chronic disease or lifestyle score ies in the review by Noble et  al. indicated clustering of all (in the association between the mental health outcomes included health behaviours [24]. and the biomarkers of ageing) did not strongly change To date, the number of studies evaluating the combina- the effect of our observations (Additional file  1: Tables tion of multiple health behaviours and mental health and S5-S7). However, we noticed that most of the associations well-being in adults is limited, and most of them use a between severe psychological distress or suicidal ideation different methodology to assess this association [50–56]. and mtDNAc showed marginally significant results. Firstly, differences are found between the included health behaviours. Most studies included the four “SNAP” risk factors, i.e. smoking, poor nutrition, excess alcohol con- Discussion sumption and physical inactivity. Other health behav- In this study, we evaluated the associations between iours that were sometimes included were BMI/obesity, eight mental health and well-being outcomes, a healthy sleep duration/quality and psychological distress [50, 53, lifestyle score and 2 biomarkers of biological ageing: Haut ekiet et al. BMC Medicine (2022) 20:328 Page 9 of 13 54, 56]. Secondly, differences are found in the scoring of our results, in a study population of 1661 men, the sum the health behaviours and the use of the lifestyle score. score of a healthier lifestyle was correlated with a longer Whereas in this study the health behaviours were scored TL [66]. Similar results were found by Sun et al. where a categorically, studies often dichotomised the health combination of healthy lifestyles in a female study pop- behaviours and/or the final lifestyle score [50, 52, 53, 56]. ulation was associated with a longer TL compared with Also, two studies performed clustering [54, 55]. Health the least healthy group [67]. Also, improvement in life- behaviours can cluster together at both ends of the risk style has been associated with TL maintenance in the spectrum, but less is known about the middle categories. elderly at risk for dementia [68], and a lifestyle interven- This is avoided by using the cluster method where partic - tion programme was positively associated with leucocyte ipants are clustered based on similar behaviours. On the telomere length in children and adolescents [69]. These other hand, a lifestyle score can be of better use and more results suggest that on a biological level, a healthy life- easily be interpreted when aiming to compare healthy style is associated with healthy ageing. Within this con- versus unhealthy lifestyles, as was the case for this study. text, a study on adults aged 60 and older showed that Despite these different methods, all previously men - maintaining a normal weight, not smoking and perform- tioned studies show similar results. Together with our ing regular physical activity were associated with slower findings, which also support these results, this pro - development of disability and a reduction in mortality vides clear evidence that an unhealthy lifestyle is associ- [70]. Similarly, midlife lifestyle factors like non-smoking, ated with poor mental health and well-being outcomes. higher levels of physical activity, non-obesity and good Important to notice is that, like our research, most stud- social support have been associated with successful age- ies in this field have a cross-sectional design and are ing, 22 years later [71]. therefore not able to assume causality. Therefore, men - tal health might be the cause or the consequence of an unhealthy lifestyle. Further prospective and longitudinal Mental health and well-being and biomarkers of ageing studies are warranted to confirm the direction of the Finally, we evaluated the association between the bio- association. markers  of ageing and the mental health and well-being outcomes. The hypothesis that biological ageing is associ - ated with mental health has been supported by observa- Healthy lifestyle and biomarkers of ageing tions showing that chronically stressed or psychiatrically How lifestyle affects our health is not yet fully under - ill persons have a higher risk for age-related diseases like stood. One possible pathway is through oxidative stress dementia, diabetes and hypertension [23, 72, 73]. Impor- and biological ageing. An unhealthy lifestyle has been tant to notice is that, like our research, the majority of associated with an increase in oxidative stress [57–59], studies on this topic have a cross-sectional design and and in turn, higher concentrations of oxidative stress are therefore are unable to identify causality. Therefore, it is known to negatively affect TL and mtDNAc [60]. In this currently unknown whether psychological diseases accel- study, we showed that living a healthy lifestyle was associ- erate biological ageing or whether biological ageing pre- ated with a longer TL and a higher mtDNAc. Our results cedes the onset of these diseases [74]. showed a stronger association of lifestyle with mtDNAc Our results showed a lower mtDNAc for individuals compared with TL. TL is strongly determined by TL at with suicidal ideation or severe psychological distress birth [61]. On the other hand, mtDNAc might be more but not for any of the other mental health outcomes. Evi- variable in shorter time periods. Although mtDNAc and dence on the association between mtDNAc and mental TL were strongly correlated, this could explain why life- health is inconsistent. Women above 60  years old with style is more strongly associated with mtDNAc. However, depression had a significantly lower mtDNAc compared we can only speculate about this, and further research is with the control group [75]. Furthermore, individuals necessary to confirm our results. with a low mtDNAc had poorer outcomes in terms of Similar as for the association with mental health, in self-rated health [76]. In contrast, Otsuka et  al. showed previous studies, the biomarkers have been associated a higher peripheral blood mtDNAc in suicide completers with health behaviours separately rather than combined [77], and studies on major depressive syndrome [78] and [62–65]. To our knowledge, we are the first to evaluate self-rated health [79] showed the same trend. Finally, the associations between a healthy lifestyle score and Vyas et  al. showed no significant association between mtDNAc. Our results are in line with our expectations. mtDNAc and depression status in mid-life and older As TL and mtDNAc are known to be correlated [60], adults [80]. These differences might be due to the dif - we would expect similar trends for both biomarkers. In ferences in study population and methods. For example, the case of TL, few studies included a combined lifestyle the two studies indicating lower mtDNAc in association score in association with this biomarker. Consistent with Hautekiet et al. BMC Medicine (2022) 20:328 Page 10 of 13 with poor mental health both had an elderly study popu- (SD) number of days is 52 (35). This is less than the lation, and one study population consisted of psychiatri- period for suicidal ideation, assessed over the 12 previ- cally ill patients. Next to that, differences were found in ous months, but there might be a more limited over- the type of samples, mtDNAc assays and questionnaires lap with the period for assessment of the other mental or diagnostics. The inconsistency of these studies and our health variables, such as vitality and psychological results calls for further research on this association and distress, assessed over the last few weeks, and depres- for standardisation of methods within studies to enable sive and generalised anxiety disorders, assessed over clear comparisons. the last 2  weeks. Fourthly, due to a non-response bias, As for TL, we did not find an association with any of the the lowest socio-economic classes are less represented mental health and well-being outcomes. Previous studies in our study population. This will not affect our dose– in adults showed a lower TL in association with current response associations but might affect the generalis - but not remitted anxiety disorder [81], depressive [82] ability of our findings to the overall population. Finally, and major depressive disorder [73, 83], childhood trauma we do not have data on blood cell counts, which has [84] suicide [77, 85], depressive symptoms in younger been associated with mtDNAc [95]. adults [86] and acculturative stress and postpartum depression in Latinx women [87]. Also, in a meta-anal- ysis, psychiatric disorders overall were associated with a Conclusions shorter leucocyte TL [88]. However, other studies failed In this large-scale study, we showed that living a healthy to demonstrate an association between TL and mental lifestyle was positively associated with mental health and health outcomes like major depressive disorder [89], late- well-being and, on a biological level, with a higher TL life depression [90] and anxiety disorder [91]. Again, this and mtDNAc, indicating healthy ageing. Furthermore, could be due to a different method to assess the mental individuals with suicidal ideation or suffering from health outcomes, a different study design, uncontrolled severe psychological distress had a lower mtDNAc. Our confounding factors and the type of telomere assay. For findings suggest that implementing strategies to incor - example, a meta-analysis showed stronger associations porate healthy lifestyle changes in the public’s daily life with depression when using southern blot or FISH assay could be beneficial for public health, and might offset compared with qPCR to measure telomere length [92]. the negative impact of environmental stressors. How- ever, further studies are necessary to confirm our results and especially prospective and longitudinal studies are Strengths and limitations essential to determine causality of the associations. An important strength of this study is the use of a vali- dated lifestyle score that can easily be reproduced and used for other research on lifestyle. Secondly, we were Abbreviations AUC: Ar ea under the curve; BMI: Body mass index; CI: Confidence intervals; able to use a large sample size for our analyses in the GAD-7: Generalised Anxiety Disorder Questionnaire; GHQ-12: General Health BHIS subset. Thirdly, by assessing multiple dimensions Questionnaire; IRC: Inter-run calibrator; mtDNAc: Mitochondrial DNA content; of mental health and well-being, we were able to give a OR: Odds ratio; PHQ-9: Patient Health Questionnaire; ROC curve: Relative operating characteristic curve; SF-36: Short Form Health Survey; TL: Telomere comprehensive overview of the mental health status. To length. our knowledge, we are the first to evaluate the associa - tions between a healthy lifestyle score and mtDNAc. Supplementary Information Our results should however be interpreted with con- The online version contains supplementary material available at https:// doi. sideration for some limitations. As mentioned before, org/ 10. 1186/ s12916- 022- 02524-9. the study has a cross-sectional design, and therefore, we cannot assume causality. Secondly, for the life- Additional file 1: Text S1. TL, mtDNAc and single copy-gene reaction mixture and PCR cycling conditions. Table S1. The mental health indica- style score, we used self-reported data, which might tors with their scores and uses. Table S2. Comparison of the characteris- not always represent the actual situation. For exam- tics of the 6,054 eligible BHIS participants that were included in the BHIS ple, BMI values tend to be underestimated due to the subset compared to the 1,838 eligible participants that were excluded from the BHIS subset. Table S3. Comparison of the characteristics of the overestimation of height and underestimation of weight 739 participants from the BHIS subset that were included in the BELHES [93], and also, smoking behaviour is often underesti- subset compared to the 5,315 participants that were excluded from the mated [94]. Also, equal weights were used for each of BELHES subset. Table S4. Bivariate associations between the characteris- tics and telomere length ( TL), mitochondrial DNA content (mtDNAc), the the health behaviours as no objective information was lifestyle score or psychological distress. Table S5. Results of the sensitivity available on which weight should be given to a spe- analysis of the association between lifestyle and mental health. Table S6. cific health behaviour. Thirdly, there is a distinct time Results of the sensitivity analysis of the association between lifestyle and the biomarkers of ageing. Table S7. Results of the sensitivity analysis of lag between the completion of the BHIS questionnaire the association between mental health and the biomarkers of ageing. Fig. and the collection of the BELHES samples. The mean Haut ekiet et al. BMC Medicine (2022) 20:328 Page 11 of 13 6. Whitman IR, Agarwal V, Nah G, Dukes JW, Vittinghoff E, Dewland TA, et al. S1. Exclusion criteria. The BHIS subset consisted of 6,055 BHIS participants Alcohol abuse and cardiac disease. J Am Coll Cardiol. 2017;69(1):13–24. and the BELHES subset consisted of 739 BELHES participants. Fig. S2. https:// doi. org/ 10. 1016/j. jacc. 2016. 10. 048. Histogram of the lifestyle score. Fig. S3. Validation of the lifestyle score. 7. Koliaki C, Liatis S, Kokkinos A. Obesity and cardiovascular disease: revisit- ROC curve for the lifestyle score as a predictor for good to very good ing an old relationship. Metabolism. 2019;92:98–107. https:// doi. org/ 10. self-perceived health. The model was adjusted for age, sex, region, highest 1016/j. metab ol. 2018. 10. 011. educational level in the household, household composition and country 8. Freisling H, Viallon V, Lennon H, Bagnardi V, Ricci C, Butterworth AS, et al. of birth. Lifestyle factors and risk of multimorbidity of cancer and cardiometa- bolic diseases: a multinational cohort study. BMC Med. 2020;18(1):5. https:// doi. org/ 10. 1186/ s12916- 019- 1474-7. Acknowledgements 9. Liu Y, Pleasants RA, Croft JB, Wheaton AG, Heidari K, Malarcher AM, et al. We are grateful to all BHIS and BELHES participants for contributing to this study. Smoking duration, respiratory symptoms, and COPD in adults aged ≥ 45 years with a smoking history. Int J Chron Obstruct Pulmon Dis. Authors’ contributions 2015;10:1409. https:// doi. org/ 10. 2147/ COPD. S82259. PH drafted the paper. PH, NS, MD and ED set up the design of the study. 10. Kirsch Micheletti J, Bláfoss R, Sundstrup E, Bay H, Pastre CM, Andersen LL. NS, DM, JvdH, TN and ED reviewed and commented on the manuscript. All Association between lifestyle and musculoskeletal pain: cross-sectional authors read and approved the final manuscript. study among 10,000 adults from the general working population. BMC Musculoskelet Disord. 2019;20(1):609. https:// doi. org/ 10. 1186/ Funding s12891- 019- 3002-5. The HuBiHIS project is financed by Sciensano (PJ ) N°: 1179–101. Dries Mar - 11. Bowe AK, Owens M, Codd MB, Lawlor BA, Glynn RW. Physical activity and tens is a postdoctoral fellow of the Research Foundation—Flanders (FWO mental health in an Irish population. Ir J Med Sci. 2019;188(2):625–31. 12X9620N). https:// doi. org/ 10. 1007/ s11845- 018- 1863-5. 12. Richardson S, McNeill A, Brose LS. Smoking and quitting behaviours by Availability of data and materials mental health conditions in Great Britain (1993–2014). Addict Behav. The dataset used for this study is available through a request to the Health 2019;90:14–9. https:// doi. org/ 10. 1016/j. addbeh. 2018. 10. 011. Committee of the Data Protection Authority. 13. Levy MZ, Allsopp RC, Futcher AB, Greider CW, Harley CB. Telomere end-replication problem and cell aging. J Mol Biol. 1992;225(4):951–60. https:// doi. org/ 10. 1016/ 0022- 2836(92) 90096-3. Declarations 14. Shaughnessy DT, McAllister K, Worth L, Haugen AC, Meyer JN, Domann FE, et al. Mitochondria, energetics, epigenetics, and cellular responses to Ethics approval and consent to participate stress. Environ Health Perspect. 2014;122(12):1271–8. https:// doi. org/ 10. As part of the BELHES, this project was approved by the Medical Ethics Com- 1289/ ehp. 14084 18. mittee of the University Hospital Ghent (registration number B670201834895). 15. Cui Y, Gao Y T, Cai Q, Qu S, Cai H, Li HL, et al. Associations of leukocyte The project was carried out in line with the recommendations of the Belgian telomere length with body anthropometric indices and weight change Privacy Commission. All participants have signed a consent form that was in Chinese women. Obesity. 2013;21(12):2582–8. https:// doi. org/ 10. 1002/ approved by the Medical Ethics Committee. oby. 20321. 16. Crous-Bou M, Fung T T, Prescott J, Julin B, Du M, Sun Q, et al. Mediter- Consent for publication ranean diet and telomere length in Nurses’ Health Study: population Not applicable. based cohort study. BMJ. 2014;349:g6674. https:// doi. org/ 10. 1136/ bmj. g6674. Competing interests 17. Janssen BG, Gyselaers W, Byun H-M, Roels HA, Cuypers A, Baccarelli AA, The authors declare that they have no competing interests. et al. Placental mitochondrial DNA and CYP1A1 gene methylation as molecular signatures for tobacco smoke exposure in pregnant women Author details and the relevance for birth weight. J Transl Med. 2017;15(1):5. https:// doi. Sciensano, Risk and Health Impact Assessment, Juliette Wytsmanstraat org/ 10. 1186/ s12967- 016- 1113-4. 14, 1050 Brussels, Belgium. Centre for Environmental Sciences, Hasselt 18. Navarro-Mateu F, Husky M, Cayuela-Fuentes P, Álvarez FJ, Roca-Vega A, University, 3500 Hasselt, Belgium. Sciensano, Epidemiology and Public Health, Rubio-Aparicio M, et al. The association of telomere length with substance Juliette Wytsmanstraat 14, 1050 Brussels, Belgium. Centre for Environment use disorders: a systematic review and meta-analysis of observational stud- and Health, Leuven University, 3000 Leuven, Belgium. ies. Addiction. 2020;116(8):1954–72. https:// doi. org/ 10. 1111/ add. 15312. 19. Pyle A, Anugrha H, Kurzawa-Akanbi M, Yarnall A, Burn D, Hudson G. Received: 4 February 2022 Accepted: 10 August 2022 Reduced mitochondrial DNA copy number is a biomarker of Parkinson’s disease. Neurobiol Aging. 2016;38:216.e7-.e10. https:// doi. org/ 10. 1016/j. neuro biola ging. 2015. 10. 033. 20. Ashar FN, Zhang Y, Longchamps RJ, Lane J, Moes A, Grove ML, et al. Asso- ciation of mitochondrial DNA copy number with cardiovascular disease. References JAMA Cardiol. 2017;2(11):1247–55. https:// doi. org/ 10. 1001/ jamac ardio. 1. World Health Organization. Healthy living: what is a healthy lifestyle? 2017. 3683. Copenhagen: WHO Regional Office for Europe; 1999. 21. Chen S, Lin J, Matsuguchi T, Blackburn E, Yeh F, Best LG, et al. Short 2. World Health Organization. Tackling NCDs: ‘best buys’ and other recom- leukocyte telomere length predicts incidence and progression of carotid mended interventions for the prevention and control of noncommunica- atherosclerosis in American Indians: the Strong Heart Family Study. ble diseases. Geneva: World Health Organization; 2017. Aging. 2014;6(5):414–27. https:// doi. org/ 10. 18632/ aging. 100671. 3. World Health Organization. WHO global report on trends in prevalence of 22. Mons U, Müezzinler A, Schöttker B, Dieffenbach AK, Butterbach K, tobacco smoking 2000–2025. 3rd ed. Geneva: World Health Organisation; Schick M, et al. Leukocyte telomere length and all-cause, cardiovascular disease, and cancer mortality: results from individual-participant-data 4. World Health Organization. 2021 Physical Activity Factsheets for the Euro- meta-analysis of 2 large prospective cohort studies. Am J Epidemiol. pean Union Member States in the WHO European Region. Copenhagen: 2017;185(12):1317–26. https:// doi. org/ 10. 1093/ aje/ kww210. WHO Regional Office for Europe; 2021. 23. Lindqvist D, Epel ES, Mellon SH, Penninx BW, Révész D, Verhoeven JE, 5. Eurostat. Database on health determinants 2021 [Available from: et al. Psychiatric disorders and leukocyte telomere length: underlying https:// ec. europa. eu/ euros tat/ web/ health/ data/ datab ase? node_ mechanisms linking mental illness with cellular aging. Neurosci Biobehav code= hlth_ det. Rev. 2015;55:333–64. https:// doi. org/ 10. 1016/j. neubi orev. 2015. 05. 007. Hautekiet et al. BMC Medicine (2022) 20:328 Page 12 of 13 24. Noble N, Paul C, Turon H, Oldmeadow C. Which modifiable health risk 46. Jacka FN, O’Neil A, Opie R, Itsiopoulos C, Cotton S, Mohebbi M, et al. A behaviours are related? A systematic review of the clustering of Smoking, randomised controlled trial of dietary improvement for adults with major Nutrition, Alcohol and Physical activity (‘SNAP’) health risk factors. Prev depression (the ‘SMILES’ trial). BMC Med. 2017;15(1):23. https:// doi. org/ 10. Med. 2015;81:16–41. https:// doi. org/ 10. 1016/j. ypmed. 2015. 07. 003.1186/ s12916- 017- 0791-y. 25. Poortinga W. The prevalence and clustering of four major lifestyle risk 47. Pavkovic B, Zaric M, Markovic M, Klacar M, Huljic A, Caricic A. Double factors in an English adult population. Prev Med. 2007;44(2):124–8. screening for dual disorder, alcoholism and depression. Psychiatry Res. https:// doi. org/ 10. 1016/j. ypmed. 2006. 10. 006. 2018;270:483–9. https:// doi. org/ 10. 1016/j. psych res. 2018. 10. 013. 26. Demarest S, Van der Heyden J, Charafeddine R, Drieskens S, Gisle 48. De Wit LM, Van Straten A, Van Herten M, Penninx BW, Cuijpers P. Depres- L, Tafforeau J. Methodological basics and evolution of the Belgian sion and body mass index, a u-shaped association. BMC Public Health. Health Interview Survey 1997–2008. Arch Public Health. 2013;71(1):24. 2009;9:14. https:// doi. org/ 10. 1186/ 1471- 2458-9- 14. https:// doi. org/ 10. 1186/ 0778- 7367- 71- 24. 49. Meader N, King K, Moe-Byrne T, Wright K, Graham H, Petticrew M, et al. A 27 Goldberg DP. User’s guide to the General Health Questionnaire. Windsor: systematic review on the clustering and co-occurrence of multiple risk NFER-Nelson; 1988. behaviours. BMC Public Health. 2016;16:657. https:// doi. org/ 10. 1186/ 28. Ware JE Jr, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: s12889- 016- 3373-6. construction of scales and preliminary tests of reliability and validity. 50. Saneei P, Esmaillzadeh A, Hassanzadeh Keshteli A, Reza Roohafza H, Afshar Med Care. 1996;34(3):220–33. https:// doi. org/ 10. 1097/ 00005 650- 19960 H, Feizi A, et al. Combined healthy lifestyle is inversely associated with 3000- 00003. psychological disorders among adults. PLoS ONE. 2016;11(1):e0146888. 29. Cantril H. Pattern of human concerns. New Brunswick: Rutgers University https:// doi. org/ 10. 1371/ journ al. pone. 01468 88. Press; 1965. 51. Bonnet F, Irving K, Terra J-L, Nony P, Berthezène F, Moulin P. Anxiety and 30. Kroenke K, Spitzer RL, Williams JB, Löwe B. The patient health question- depression are associated with unhealthy lifestyle in patients at risk of naire somatic, anxiety, and depressive symptom scales: a systematic cardiovascular disease. Atherosclerosis. 2005;178(2):339–44. https:// doi. review. Gen Hosp Psychiatry. 2010;32(4):345–59. https:// doi. org/ 10. 1016/j. org/ 10. 1016/j. ather oscle rosis. 2004. 08. 035. genho sppsy ch. 2010. 03. 006. 52. Loprinzi PD, Mahoney S. Concurrent occurrence of multiple positive 31. Löwe B, Decker O, Müller S, Brähler E, Schellberg D, Herzog W, et al. lifestyle behaviors and depression among adults in the United States. J Validation and standardization of the Generalized Anxiety Disorder Aec ff t Disord. 2014;165:126–30. https:// doi. org/ 10. 1016/j. jad. 2014. 04. 073. Screener (GAD-7) in the general population. Med Care. 2008;46(3):266– 53. Yang H, Gao J, Wang T, Yang L, Liu Y, Shen Y, et al. Association between 74. https:// doi. org/ 10. 1097/ MLR. 0b013 e3181 60d093. adverse mental health and an unhealthy lifestyle in rural-to-urban 32. Gisle L, Drieskens S, Demarest S, Van der Heyden J. Enquête de santé migrant workers in Shanghai. J Formos Med Assoc. 2017;116(2):90–8. 2018: Santé mentale. Bruxelles: Sciensano; 2018. (https:// www. enque https:// doi. org/ 10. 1016/j. jfma. 2016. 03. 004. tesan te. be. Available from. Numéro de rapport: D/2020/14.440/3). 54. Oftedal S, Kolt GS, Holliday EG, Stamatakis E, Vandelanotte C, Brown 33 Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey WJ, et al. Associations of health-behavior patterns, mental health and (SF-36). I. Conceptual framework and item selection. Medical Care. self-rated health. Prev Med. 2019;118:295–303. https:// doi. org/ 10. 1016/j. 1992;30(6):473–83.ypmed. 2018. 11. 017. 34 Braunholtz S, Davidson S, Myant K, O’Connor R. Well? What do you think?: 55. Conry MC, Morgan K, Curry P, McGee H, Harrington J, Ward M, et al. The the Third National Scottish Survey of Public Attitudes to Mental Health, clustering of health behaviours in Ireland and their relationship with Mental Wellbeing and Mental Health Problems. Scotland: Scottish Gov- mental health, self-rated health and quality of life. BMC Public Health. ernment Edinburgh; 2007. 2011;11(1):692. https:// doi. org/ 10. 1186/ 1471- 2458- 11- 692. 35. Van Lente E, Barry MM, Molcho M, Morgan K, Watson D, Harrington J, et al. 56. Buttery AK, Mensink GB, Busch MA. Healthy behaviours and mental Measuring population mental health and social well-being. Int J Public health: findings from the German Health Update (GEDA). The European Health. 2012;57(2):421–30. https:// doi. org/ 10. 1007/ s00038- 011- 0317-x. Journal of Public Health. 2015;25(2):219–25. https:// doi. org/ 10. 1093/ 36 de Bruin A, Picavet HSJ, Nossikov A. Health interview surveys: towards eurpub/ cku094. international harmonization of methods and instruments. Copenhagen: 57 Ahmed NJ, Husen AZ, Khoshnaw N, Getta HA, Hussein ZS, Yassin AK, et al. World Health Organization. Regional Office for Europe; 1996. The effects of smoking on IgE, oxidative stress and haemoglobin concen- 37. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression tration. Asian Pac J Cancer Prev. 2020;21(4):1069–72. https:// doi. org/ 10. severity measure. J Gen Intern Med. 2001;16(9):606–13.31557/ APJCP. 2020. 21.4. 1069. 38. Superior Health Council. Dietary guidelines for the Belgian adult popula- 58. Langley MR, Yoon H, Kim HN, Choi C-I, Simon W, Kleppe L, et al. High fat tion. Report 9284. Brussels: Superior Health Council; 2019. diet consumption results in mitochondrial dysfunction, oxidative stress, 39. World Health Organization. WHO guidelines on physical activity and and oligodendrocyte loss in the central nervous system. Biochim Biophys sedentary behaviour. Geneva: World Health Organization; 2020. Acta Mol Basis Dis. 2020;1866(3):165630. https:// doi. org/ 10. 1016/j. bbadis. 40. World Health Organization. HEARTS Technical package for cardiovascular 2019. 165630. disease management in primary health care: healthy-lifestyle counselling. 59. Tan HK, Yates E, Lilly K, Dhanda AD. Oxidative stress in alcohol-related liver Geneva: World Health Organization; 2018. disease. World J Hepatol. 2020;12(7):332–49. https:// doi. org/ 10. 4254/ wjh. 41 World Health Organization. Country profiles on nutrition, physical activity v12. i7. 332. and obesity in the 28 European Union Member States of the WHO Euro- 60. Martens DS, Nawrot TS. Air pollution stress and the aging phenotype: the pean Region. Copenhagen: WHO Regional Office for Europe; 2013. telomere connection. Curr Environ Health Rep. 2016;3(3):258–69. https:// 42 Bhaskaran K, dos-Santos-Silva I, Leon DA, Douglas IJ, Smeeth L. Associa-doi. org/ 10. 1007/ s40572- 016- 0098-8. tion of BMI with overall and cause-specific mortality: a population- 61. Martens DS, Van Der Stukken C, Derom C, Thiery E, Bijnens EM, Nawrot based cohort study of 3.6 million adults in the UK. Lancet Diabetes TS. Newborn telomere length predicts later life telomere length: track- Endocrinol. 2018;6(12):944–53. https:// doi. org/ 10. 1016/ S2213- 8587(18) ing telomere length from birth to child- and adulthood. EBioMedicine. 30288-2. 2021;63:103164. https:// doi. org/ 10. 1016/j. ebiom. 2020. 103164. 43. Benetou V, Kanellopoulou A, Kanavou E, Fotiou A, Stavrou M, Richardson 62. Hang D, Nan H, Kværner AS, De Vivo I, Chan AT, Hu Z, et al. Longitudinal C, et al. Diet-related behaviors and diet quality among school-aged associations of lifetime adiposity with leukocyte telomere length and adolescents living in Greece. Nutrients. 2020;12(12):3804. https:// doi. org/ mitochondrial DNA copy number. Eur J Epidemiol. 2018;33(5):485–95. 10. 3390/ nu121 23804.https:// doi. org/ 10. 1007/ s10654- 018- 0382-z. 44. Martens DS, Plusquin M, Gyselaers W, De Vivo I, Nawrot TS. Maternal pre- 63. Gu Y, Honig LS, Schupf N, Lee JH, Luchsinger JA, Stern Y, et al. Mediterra- pregnancy body mass index and newborn telomere length. BMC Med. nean diet and leukocyte telomere length in a multi-ethnic elderly popu- 2016;14(1):148. https:// doi. org/ 10. 1186/ s12916- 016- 0689-0. lation. Age. 2015;37(2):24. https:// doi. org/ 10. 1007/ s11357- 015- 9758-0. 45. Janssen BG, Munters E, Pieters N, Smeets K, Cox B, Cuypers A, et al. Pla- 64. Sellami M, Al-Muraikhy S, Al-Jaber H, Al-Amri H, Al-Mansoori L, Mazloum cental mitochondrial DNA content and particulate air pollution during in NA, et al. Age and sport intensity-dependent changes in cytokines and utero life. Environ Health Perspect. 2012;120(9):1346–52. https:// doi. org/ telomere length in elite athletes. Antioxidants. 2021;10(7):1035. https:// 10. 1289/ ehp. 11044 58.doi. org/ 10. 3390/ antio x1007 1035. Haut ekiet et al. BMC Medicine (2022) 20:328 Page 13 of 13 65. Savela S, Saijonmaa O, Strandberg TE, Koistinen P, Strandberg AY, Tilvis RS, patients with major depressive disorder. Psychiatr Genet. 2022;32(1):34–6. et al. Physical activity in midlife and telomere length measured in old age. https:// doi. org/ 10. 1097/ ypg. 00000 00000 000305. Exp Gerontol. 2013;48(1):81–4. https:// doi. org/ 10. 1016/j. exger. 2012. 02. 84. Aas M, Elvsåshagen T, Westlye LT, Kaufmann T, Athanasiu L, Djurovic S, 003. et al. Telomere length is associated with childhood trauma in patients 66. Mirabello L, Huang WY, Wong JY, Chatterjee N, Reding D, David Crawford with severe mental disorders. Transl Psychiatry. 2019;9(1):97. https:// doi. E, et al. The association between leukocyte telomere length and cigarette org/ 10. 1038/ s41398- 019- 0432-7. smoking, dietary and physical variables, and risk of prostate cancer. Aging 85. Birkenæs V, Elvsåshagen T, Westlye LT, Høegh MC, Haram M, Werner Cell. 2009;8(4):405–13. https:// doi. org/ 10. 1111/j. 1474- 9726. 2009. 00485.x. MCF, et al. Telomeres are shorter and associated with number of suicide 67. Sun Q, Shi L, Prescott J, Chiuve SE, Hu FB, De Vivo I, et al. Healthy attempts in affective disorders. J Aec ff t Disord. 2021;295:1032–9. https:// lifestyle and leukocyte telomere length in US women. PLoS ONE. doi. org/ 10. 1016/j. jad. 2021. 08. 135. 2012;7(5):e38374. https:// doi. org/ 10. 1371/ journ al. pone. 00383 74. 86. Phillips AC, Robertson T, Carroll D, Der G, Shiels PG, McGlynn L, et al. Do 68. Sindi S, Solomon A, Kåreholt I, Hovatta I, Antikainen R, Hänninen T, symptoms of depression predict telomere length? Evidence from the et al. Telomere length change in a multidomain lifestyle intervention West of Scotland Twenty-07 Study. Psychosom Med. 2013;75(3):288–96. to prevent cognitive decline: a randomized clinical trial. J Gerontol A. https:// doi. org/ 10. 1097/ PSY. 0b013 e3182 89e6b5. 2021;76(3):491–8. https:// doi. org/ 10. 1093/ gerona/ glaa2 79. 87. Incollingo Rodriguez AC, Polcari JJ, Nephew BC, Harris R, Zhang C, Mur- 69. Paltoglou G, Raftopoulou C, Nicolaides NC, Genitsaridi SM, Karampatsou gatroyd C, et al. Acculturative stress, telomere length, and postpartum SI, Papadopoulou M, et al. A comprehensive, multidisciplinary, personal- depression in Latinx mothers. J Psychiatr Res. 2022;147:301–6. https:// doi. ized, lifestyle intervention program is associated with increased leukocyte org/ 10. 1016/j. jpsyc hires. 2022. 01. 063. telomere length in children and adolescents with overweight and 88. Darrow SM, Verhoeven JE, Révész D, Lindqvist D, Penninx BW, Delucchi obesity. Nutrients. 2021;13(8):2682. https:// doi. org/ 10. 3390/ nu130 82682. KL, et al. The association between psychiatric disorders and telomere 70. Chakravarty EF, Hubert HB, Krishnan E, Bruce BB, Lingala VB, Fries JF. Life- length: a meta-analysis involving 14,827 persons. Psychosom Med. style risk factors predict disability and death in healthy aging adults. Am J 2016;78(7):776–87. https:// doi. org/ 10. 1097/ PSY. 00000 00000 000356. Med. 2012;125(2):190–7. https:// doi. org/ 10. 1016/j. amjmed. 2011. 08. 006. 89. Simon NM, Walton ZE, Bui E, Prescott J, Hoge E, Keshaviah A, et al. 71. Bosnes I, Nordahl HM, Stordal E, Bosnes O, Myklebust TÅ, Almkvist O. Life- Telomere length and telomerase in a well-characterized sample of indi- style predictors of successful aging: a 20-year prospective HUNT study. viduals with major depressive disorder compared to controls. Psychoneu- PLoS ONE. 2019;14(7):e0219200. https:// doi. org/ 10. 1371/ journ al. pone. roendocrinology. 2015;58:9–22. https:// doi. org/ 10. 1016/j. psyne uen. 2015. 02192 00.04. 004. 72. Epel ES, Prather AA. Stress, telomeres, and psychopathology: toward a 90. Schaakxs R, Verhoeven JE, Voshaar RCO, Comijs HC, Penninx BW. Leuko- deeper understanding of a triad of early aging. Annu Rev Clin Psychol. cyte telomere length and late-life depression. Am J Geriatr Psychiatry. 2018;14:371–97. https:// doi. org/ 10. 1146/ annur ev- clinp sy- 032816- 045054. 2015;23(4):423–32. https:// doi. org/ 10. 1016/j. jagp. 2014. 06. 003. 73. Verhoeven JE, Révész D, Epel ES, Lin J, Wolkowitz OM, Penninx BW. Major 91. de Baumont AC, Hoffmann MS, Bortoluzzi A, Fries GR, Lavandoski P, Grun depressive disorder and accelerated cellular aging: results from a large LK, et al. Telomere length and epigenetic age acceleration in adolescents psychiatric cohort study. Mol Psychiatry. 2014;19(8):895–901. https:// doi. with anxiety disorders. Sci Rep. 2021;11(1):7716. https:// doi. org/ 10. 1038/ org/ 10. 1038/ mp. 2013. 151.s41598- 021- 87045-w. 74. Han LK, Verhoeven JE, Tyrka AR, Penninx BW, Wolkowitz OM, Månsson 92. Schutte NS, Malouff JM. The association between depression and leuko - KN, et al. Accelerating research on biological aging and mental health: cyte telomere length: a meta-analysis. Depress Anxiety. 2015;32(4):229– current challenges and future directions. Psychoneuroendocrinology. 38. https:// doi. org/ 10. 1002/ da. 22351. 2019;106:293–311. https:// doi. org/ 10. 1016/j. psyne uen. 2019. 04. 004. 93. Drieskens S, Demarest S, Bel S, De Ridder K, Tafforeau J. Correction of 75. Kim MY, Lee JW, Kang HC, Kim E, Lee DC. Leukocyte mitochondrial DNA self-reported BMI based on objective measurements: a Belgian experi- (mtDNA) content is associated with depression in old women. Arch ence. Archives of Public Health. 2018;76:10. https:// doi. org/ 10. 1186/ Gerontol Geriatr. 2011;53(2):e218–21. https:// doi. org/ 10. 1016/j. archg er. s13690- 018- 0255-7. 2010. 11. 019. 94. Gorber SC, Schofield-Hurwitz S, Hardt J, Levasseur G, Tremblay M. The 76. Mengel-From J, Thinggaard M, Dalgard C, Kyvik KO, Christensen K, accuracy of self-reported smoking: a systematic review of the relationship Christiansen L. Mitochondrial DNA copy number in peripheral blood between self-reported and cotinine-assessed smoking status. Nicotine cells declines with age and is associated with general health among Tob Res. 2009;11(1):12–24. https:// doi. org/ 10. 1093/ ntr/ ntn010. elderly. Hum Genet. 2014;133(9):1149–59. https:// doi. org/ 10. 1007/ 95. Knez J, Winckelmans E, Plusquin M, Thijs L, Cauwenberghs N, Gu Y, et al. s00439- 014- 1458-9. Correlates of peripheral blood mitochondrial DNA content in a general 77. Otsuka I, Izumi T, Boku S, Kimura A, Zhang Y, Mouri K, et al. Aberrant population. Am J Epidemiol. 2016;183(2):138–46. https:// doi. org/ 10. 1093/ telomere length and mitochondrial DNA copy number in suicide com-aje/ kwv175. pleters. Sci Rep. 2017;7(1):3176. 78. Chung JK, Lee SY, Park M, Joo E-J, Kim SA. Investigation of mitochondrial Publisher’s Note DNA copy number in patients with major depressive disorder. Psychiatry Springer Nature remains neutral with regard to jurisdictional claims in pub- Res. 2019;282:112616. https:// doi. org/ 10. 1016/j. psych res. 2019. 112616. lished maps and institutional affiliations. 79. Takahashi PY, Jenkins GD, Welkie BP, McDonnell SK, Evans JM, Cerhan JR, et al. Association of mitochondrial DNA copy number with self-rated health status. Appl Clin Genet. 2018;11:121–7. https:// doi. org/ 10. 2147/ TACG. S1676 40. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : 80. Vyas CM, Ogata S, Reynolds CF 3rd, Mischoulon D, Chang G, Cook NR, et al. Lifestyle and behavioral factors and mitochondrial DNA copy fast, convenient online submission number in a diverse cohort of mid-life and older adults. PLoS ONE. thorough peer review by experienced researchers in your field 2020;15(8):e0237235. https:// doi. org/ 10. 1371/ journ al. pone. 02372 35. 81. Verhoeven JE, Révész D, van Oppen P, Epel ES, Wolkowitz OM, Penninx rapid publication on acceptance BW. Anxiety disorders and accelerated cellular ageing. Br J Psychiatry. support for research data, including large and complex data types 2015;206(5):371–8. https:// doi. org/ 10. 1192/ bjp. bp. 114. 151027. • gold Open Access which fosters wider collaboration and increased citations 82. Pisanu C, Vitali E, Meloni A, Congiu D, Severino G, Ardau R, et al. Inves- tigating the role of leukocyte telomere length in treatment-resistant maximum visibility for your research: over 100M website views per year depression and in response to electroconvulsive therapy. J Pers Med. 2021;11(11):1100. https:// doi. org/ 10. 3390/ jpm11 111100. At BMC, research is always in progress. 83. da Silva RS, de Moraes LS, da Rocha CAM, Ferreira-Fernandes H, Yoshioka Learn more biomedcentral.com/submissions FKN, Rey JA, et al. Telomere length and telomerase activity of leukocytes as biomarkers of selective serotonin reuptake inhibitor responses in

Journal

BMC Medicine – Springer Journals

Published: Sep 29, 2022

Keywords: Mental health; Lifestyle; Biological ageing; Mitochondrial DNA content; Telomere length

There are no references for this article.