Purpose We examined if data-driven food-patterns associate with weight change, incidence of type 2 diabetes (T2D), coro- nary events (CE) and stroke. Methods The study included 20,487 individuals (61% women) from the Malmö Diet and Cancer cohort, 45–74 years, with- out diabetes and CVD at baseline (1991–1996) and who did not report dietary changes. Diet was measured with a modified diet history method. During 15 years follow-up, 2206 T2D, 1571 CE and 1332 stroke cases were identified. Data on weight change after 16.7 years were available in 2627 individuals. Results From principal component analysis, we identified six food-patterns which were similar in women and men. The first pattern, explaining 7% of the variance, was characterized by high intake of fibre-rich bread, breakfast cereals, fruits, vegeta- bles, fish and low-fat yoghurt, and by low intake of low-fibre bread. This health conscious pattern was associated with lower T2D risk (HR comparing highest quintile with lowest: 0.75; 95% CI 0.61–0.92, 0.82; 95% CI 0.68–1.00 in women and men, respectively, P trends = 0.003, 0.01) and CE (HR 0.77; 95% CI 0.58–1.02, HR 0.83; 95% CI 0.68–1.01, P trends = 0.05, 0.07), and in men also with lower risk of ischemic stroke (HR 0.69; 95% CI 0.54–0.88; P trend = 0.001) and less pronounced weight gain (0.93 kg/10 years, P trend = 0.03). A low-fat product pattern was associated with increased T2D risk in gender combined analyses (P trend = 0.03) and a pattern characterized by dressing and vegetables with lower CE risk in men (P trend = 0.02). Conclusions Our main finding was that a dietary pattern indicating health conscious food choices was associated with lower risk of cardiometabolic diseases in both genders. Keywords Food intake · Epidemiology · Weight gain · Type 2 diabetes · Cardiovascular diseases Introduction Electronic supplementary material The online version of this article (https ://doi.org/10.1007/s0039 4-018-1727-9) contains The prevalence of obesity, type 2 diabetes (T2D) and car- supplementary material, which is available to authorized users. diovascular disease is increasing globally. Although many lifestyle-related risk factors have been identified, the great * Ulrika Ericson challenge of designing public health recommendations, email@example.com including dietary advice to stop this negative trend, still Department of Clinical Sciences in Malmö, Diabetes remains. and Cardiovascular Disease, Genetic Epidemiology, Lund Diet is an extremely complex exposure and during the University, Malmö, Sweden last decades, the value of examining intake of foods or food Department of Clinical Sciences in Malmö, Nutritional patterns, in addition to nutrients, has been highlighted . Epidemiology, Lund University, Malmö, Sweden Correlations and interactions between food components may Department of Internal Medicine, Skåne University Hospital, not be satisfactorily taken into account in observational stud- Lund University, Malmö, Sweden ies on single dietary components , which complicates the Clinical Research Centre, building 60, floor 13, SUS interpretation of findings in nutritional epidemiology. It is in Malmö, entrance 72, Jan Waldenströms gata 35, difficult to isolate the effect of specific nutrients or foods. 205 02 Malmö, Sweden Vol.:(0123456789) 1 3 European Journal of Nutrition In addition, cumulative effects of several food components eligible persons) completed all baseline examinations. In may be of greater magnitude and, therefore, easier to detect the current study, we included 20,487 participants (12,456 . Moreover, we do not eat single nutrients or foods, but women and 8031 men) free of diabetes and cardiovascular combinations of these, and it is, therefore, of great value to disease at baseline and who did not report previous dietary capture overall dietary patterns that could be translated into changes (Supplementary Fig. 1). We excluded 1193 individ- relevant food-based dietary guidelines. One option is to use uals, based on self-reported diabetes diagnosis, self-reported data-driven statistical methods to reduce intake data on a diabetes medication or information from medical data reg- number of single foods or food groups into meaningful eat- istries (see below) indicating a date of diagnosis preceding ing patterns reflecting how foods are commonly consumed baseline examination date and 855 individuals with a history together. of coronary events or stroke. Among those without history Several so-called “healthy/prudent” data-driven dietary of diabetes, coronary events and stroke, we excluded 5149 patterns have been found to associate with lower risk of T2D individuals who reported dietary change in the past based [3, 4] and cardiovascular disease , whereas “unhealthy/ on the question “Have you substantially changed your eating western” patterns seem to associate with increased risk of habits because of illness or some other reasons?”, because T2D . A few studies have also derived dietary patterns intake changes of specific foods may change the relative associated with weight change [6–9]. Still, it is important to intake of other foods and thereby the overall dietary patterns. examine dietary patterns in die ff rent populations, as the food Among the 20,487 individuals in this study, 2627 could be types, food preparation, as well as how foods are combined included in the analysis of longitudinal weight change, as differ and observed findings in one study may, therefore, not they participated in follow-up examinations between 2007 be applicable to food-based guidelines in other populations. and 2012. For the follow-up examination, all individuals In this population-based prospective study of men and that were alive and still living in Sweden (n = 4924) from women from the Swedish Malmö Diet and Cancer (MDC) a random sub-cohort of 6103 MDC participants, the MDC cohort, we aimed to identify food patterns using principal cardiovascular Cohort, were invited to participate. The ethi- component analysis and to examine associations with weight cal committee at Lund University has approved the study change, and incidence of T2D, coronary disease and stroke. (LU 51–90) and the participants have given their written informed consent. Subjects and methods Dietary data Study population and data collection Dietary data were collected once during the baseline period. The MDC study used an interview-based, modified diet his- The MDC study is a population-based prospective cohort tory method that combined (1) a 7-day menu book (food study in Malmö, a city in the south of Sweden. Baseline record of meals that vary from day to day) (usually lunch and examinations were conducted between 1991 and 1996. dinner meals), cold beverages and nutrient supplements, and All women born 1923–1950 and all men born 1923–1945, (2) a 168-item food frequency questionnaire for assessment living in the city of Malmö, were invited to participate of consumption frequencies and portion sizes of regularly (n = 74,138). Details of the cohort and the recruitment pro- eaten foods that were not covered by the 7-day menu book. cedures are described elsewhere . The only exclusion Finally, (3) a 45-min interview completed the dietary assess- criteria were mental incapacity and inadequate Swedish ment. The MDC method is described in detail elsewhere language skills (eligible persons = 68,905). The participants [11, 12]. filled out questionnaires covering socio-economic, lifestyle The diet analyses were adjusted for a variable called “diet and dietary factors, and recorded meals, and underwent a method version” because slightly altered coding routines of diet history interview. Anthropometric measurements were dietary data were introduced in September 1994 to shorten conducted by nurses. Weight was measured using a bal- the interview time (from 60 to 45 min). The altered coding ance-beam scale with subjects wearing light clothing and routines resulted in two slightly different method versions no shoes. Standing height was measured with a fixed stadi- (before or after September 1994) without any major influ- ometer calibrated in centimeters. Waist circumference was ence on the ranking of individuals . measured midway between the lowest rib margin and the The relative validity of the MDC method was evaluated in iliac crest. Body composition was estimated with a bioelec- the Malmö Food study 1984–1985, comparing the method trical impedance analyzer (BIA 103, RJL systems, single- with 18 days weighed food records [13, 14]. The Pearson frequency analyzer, Detroit, USA). Body fat percent was correlation coefficients, adjusted for total energy, between calculated using an algorithm provided by the manufacturer. the reference method and the MDC-method, were in women During the screening period, 28,098 participants (40% of the and men, respectively, for intakes of bread 0.58/0.50, cereals 1 3 European Journal of Nutrition 0.73/0.74, fruits 0.77/0.60, vegetables 0.53/0.65, low-fat and Welfare in Sweden: the Swedish National Inpatient Reg- milk 0.92/0.90, high-fat milk 0.75/0.76, cheese 0.59/0.47, istry, the Swedish Hospital-based outpatient care, the Cause- fish 0.70/0.35, low-fat meat 0.51/0.43 and for high-fat meat of-death Registry and the Swedish Prescribed Drug Registry. 0.80/0.40 . Type of diabetes was based on the glycaemic parameters, The mean daily intake of foods was calculated based on treatment/medication, age at diagnosis, GADA, C-peptide frequency and portion size estimates from the question- and BMI. naire and menu-book. The food intake was converted to Information about prevalent and incident coronary event energy and nutrient intakes using the MDC nutrient data- and ischemic stroke was taken from the national Swedish base where the majority of the nutrient information comes Hospital Discharge register, Cause-of-death register  and from PC-KOST2-93 from the National Food Agency in the local stroke register in Malmö (STROMA) . A coro- Uppsala, Sweden. The food intakes were aggregated into nary event was defined on the basis of codes 410–414 (fatal 33 groups to obtain food groups more frequently consumed or non-fatal myocardial infarction or death due to ischemic in the population, but to keep characteristics related to both heart disease) in the International Classification of Diseases, dietary behaviors and nutrient content. When aggregating 9th Revision (ICD-9). Ischemic stroke was den fi ed based on the foods, some special concern was taken regarding fat code 434 (ICD-9) and diagnosed when computed tomogra- and fibre contents, as the dietary assessment was especially phy, magnetic resonance imaging or autopsy could verify designed to capture those intakes  and as dietary fat and the infarction and/or exclude haemorrhage and non-vascular fibre are thought to be crucial in the development of cardio- disease. If neither imaging nor autopsy was performed, the metabolic disease [16, 17]. Energy-adjusted intakes of the stroke was classified as unspecified. Haemorrhagic or non- 33 food groups were obtained by regressing the intakes on specific stroke cases (ICD-9 code 430, 431 and 436) were non-alcohol energy intake. excluded since these subtypes of stroke do not have the same underlying risk factors as ischemic stroke. National Tax Ascertainment of diabetes, coronary events Board provided information on vital status and emigration. and stroke Weight change We identified 2206 incident cases of T2D during 304,182 person years of follow-up via at least one of seven registries Weight was measured both at baseline and at follow-up (90%) or at new screenings or examinations during follow- examinations in 2627 individuals, with on average 16.7 years up (10%). The mean follow-up time was 15 years (range between the examinations. The yearly weight change was 0–20). In total, 1571 coronary event cases and 1332 stroke calculated: weight at follow-up minus weight at baseline cases were identified during 312,262–312,303 person years divided by number of follow-up years. Mean 10-year weight of follow-up, respectively, with a mean follow-up time of change was obtained by multiplying the yearly weight 15 years (range 0–20). The subjects contributed person- change by 10. time from date of enrolment until date of diagnosis, death, migration from Sweden, or end of follow-up (December Other variables 2010), whichever occurred first. During follow-up 0.5% had migrated from Sweden. Information on age was obtained from the personal identi- If available, we used information on the date of T2D fication number. Body mass index (BMI; kg/m ) was calcu- diagnosis from two registries prioritized in the following lated from direct measurement of weight and height. Leisure order: (1) the regional Diabetes 2000 registry of Scania time physical activity was assessed by asking the partici-  and (2) the Swedish National Diabetes Registry . pants to estimate the number of minutes per week they spent These registries required a physician diagnosis according on 17 different activities. The duration was multiplied with to established diagnosis criteria (fasting plasma glucose an activity specific intensity coefficient and an overall lei - concentration ≥ 7.0 mmol/L or fasting whole blood concen- sure time physical activity score was created. The score was tration ≥ 6.1 mmol/L, measured at two different occasions). divided into gender-specific quintiles. The smoking status Individuals with at least two HbA1c values above 6.0% with of the participants was defined as current smokers (includ- the Swedish Mono-S standardization system (correspond- ing irregular smokers), ex-smokers and never-smokers. The ing to 6.9% in the US National Glycohemoglobin Stand- total consumption of alcohol was defined by a four-category ardization Program and 52 mmol/mol with the International variable. Participants reporting zero consumption in the Federation of Clinical Chemistry and Laboratory Medicine menu book, and indicating no consumption of any type of (IFCC) units) [20, 21] were categorized as diabetes cases in alcohol during the previous year, were categorized as zero- the Malmö HbA1c Registry. In addition, cases were identi- reporters. The other category ranges were < 15 g alcohol/day fied via four registries from the National Board of Health for women and < 20 g/day for men (low), 15–30 g/day for 1 3 European Journal of Nutrition women and 20–40 g/day for men (medium), and > 30 g/day energy intake (continuous). Our full multivariable model for women and > 40 g/day for men (high). Participants were additionally included adjustments for the following cat- divided into four categories according to their highest level egorical variables: leisure time physical activity, smoking, of education (≤ 8, 9–10, 11–13 years or university degree). alcohol intake, and education, and finally baseline BMI as Season was defined as season of diet data collection (winter, a continuous variable. Since associations between diet and spring, summer and fall). cardiometabolic disease may partly be mediated via BMI, we also performed analyses with an intermediate multi- Statistical analysis variable model without inclusion of BMI. The covariates were identified from the literature and indicated potential The SPSS statistical computer package (version 24.0; IBM confounding in the MDC cohort, due to associations with Corporation, Armonk, NY, USA) was used for all statistical incident cardiometabolic diseases and dietary intakes. We analyses. All food variables were log transformed (e-log) also made additional adjustments for waist circumference. to normalize the distribution before analysis. To handle In analyses of weight change, we also performed analyses log transformation of zero intakes, we added a very small with adjustment for baseline weight. Missing values for the amount (0.01 g). All food intakes were energy adjusted with variables were treated as separate categories. Tests for inter- the residual method. actions between the food patterns and obesity status or sex We used principal component analysis (eigenvalues > 1 with regard to the incident diseases were performed [e.g. and varimax rotation) to reduce 33 energy-adjusted food BMI (≤ 25 or > 25) × quintile of food patterns (treated as groups into factors representing food patterns. We derived continuous variables)]. In sensitivity analyses, we excluded factors separately in women and men. From the obtained individuals above 60 years of age at baseline in analyses Scree plots (Supplementary Figs. 2a and 2b), we decided of T2D, coronary events and stroke, as these participants to retain and rotate the six factors (eigenvalues > 1.2) that could be regarded as elderly already at baseline. Although explained most of the variance in the data in both genders. they did not report to have made any major dietary changes, These factors were possible to interpret and translate into information about their diet, physical activity and weight food patterns based on their loadings for the initial food may not reflect that of their previous life. All statistical tests group variables. Reported characteristics of the patterns were two-sided and statistical significance was assumed at were based on food group loadings < − 0.25 and > + 0.25. P < 0.05. We examined baseline characteristics across quintiles of the factors representing food patterns with the general linear model for continuous variables (adjusted for age) and with Results Chi square test for categorical variables. We used Cox pro- portional hazards regression model to estimate hazard ratios Food patterns (HRs) of incident T2D, coronary events and stroke asso- ciated with quintiles of factors representing food patterns. We retained six factors, which were similar in both genders. The first quintile was used as the reference. Years of fol- Together, these factors explained 30% of the variance in the low-up was used as the underlying time variable. To assess food intake data. The first derived factor explained 7.2% of the proportional hazards assumption, we used graphs and the variance and was characterized by (i.e. loadings < − 0.25 tested interactions between the underlying time variable and or > 0.25) high intakes of fibre-rich bread, fruits, vegetables, examined covariates. The assumption was considered to be breakfast cereals, fish and low-fat yoghurt, and by low intake satisfied for all covariates except age with regard to T2D in of low-fibre bread, in both women and men. In men, the first women. We, therefore, additionally performed T2D-analyses factor was also characterized by high intake of cream, and with age-stratified cox models (per 1-year age interval) in in women by cottage cheese (Fig. 1a, b). In addition, rather women, but the results remained unchanged. The general lin- low intakes of red and processed meat and sugar-sweetened ear model was used to examine mean 10-year weight change beverages could be noted in both genders. This pattern was in quintiles of the food patterns. If non-significant tendencies named the “health conscious” food pattern. The other five of associations in the same direction were seen in both gen- patterns (Supplementary Fig. 3a–j) were named the “low-fat ders (no statistical interaction with gender), we performed products” pattern (characterized by high intakes of low-fat gender-combined analyses on factors representing similar margarines, low-fat milk and low-fat yoghurt, but by low food patterns in both genders. Significant observations from intake of butter, explained 5% of the variance), the “dress- gender-combined analyses are reported in the text. ing and vegetables” pattern (characterized by high intake We used covariates obtained from the baseline exami- of dressing/oils, vegetables, poultry, salty snacks, rice/ nations. A basic model included adjustments for age (con- pasta, fried potatoes and cheese, but by low intake of boiled tinuous), diet method version, season (categorical) and total potatoes and jam/sugar, explained 5% of the variance), the 1 3 European Journal of Nutrition Fig. 1 a The food pattern that Factor loadings - Health conscious food pattern-women explained most of the variance in women (7%). b The food pat- Fiber-rich bread Fruits tern that explained most of the Vegetables Yoghurt, low-fat variance in men (7%) Breakfast cereals Cottage cheese Fish/shellfish milk, low-fat Icecream Juice Cream Tea Cheese Pastry and biscuits Potatoes, boiled Poultry Oils and dressing Coffee Yoghurt, high-fat Margarine, low-fat Rice/pasta Eggs Butter Salty snacks/nuts Sweets Non-energy soft drinks Marmelade/sugar Margarine, high-fat Milk, high-fat Potatoes, fried Sugar sweetened beverages Meat, red/processed Low-fiber bread -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 Factor loadings - Health conscious food pattern-men Fiber-rich bread Fish/shellfish Fruits Yoghurt, low-fat Breakfast cereals Vegetables Cream milk, low-fat Juice Cottage cheese Cheese Potatoes, boiled Icecream Pastry and biscuits Oils and dressing Tea Coffee Butter Yoghurt, high-fat Poultry Sweets Rice/pasta Margarine, low-fat Salty snacks/nuts Eggs Non-energy soft drinks Margarine, high-fat Marmelade/sugar Potatoes, fried Meat, red/processed Milk, high-fat Sugar sweetened beverages Low-fiber bread -0.8 -0.6 -0.4 -0.2 0.00.2 0.40.6 0.8 “traditional meal” pattern (characterized by high intakes of Baseline characteristics egg, high-fat margarine, boiled potatoes, fish, red/processed meat, cream and high-fat milk, explained 5% of the variance Baseline characteristics in the whole study sample and in the in women and 4% in men), the “tea-breakfast” pattern (char- subsample with data on weight-change are shown in Supple- acterized by high intakes of tea and breakfast foods such as mentary Table 1. Women and men in the higher quintiles of cereals, jam/sugar, high-fat yoghurt and high-fat milk, but the first food pattern (“health conscious”), which explained by low intake of coffee and red/processed meat, explained most of the variation in the food intakes, had higher age and 4% of the variance in women and 5% in men) and the “sugar- reported higher intakes of energy, protein, carbohydrates, rich” pattern (characterized by high intakes of sweets, pastry, fibre and vitamin C, but lower intakes of fat and sucrose ice cream and sugar-sweetened beverages, explained 4% of compared to individuals in the lower quintiles (Table 1). the variance on both genders). They were also characterized by rather favourable lifestyle 1 3 European Journal of Nutrition Table 1 Baseline characteristics across quintiles (Q) of food patterns in men and women from the Malmö Diet and Cancer cohort Women (n = 12,456) Men (n = 8031) Q of health conscious food pattern Q of HEALTH CONSCIOUS food pattern a a Baseline characteristics Beta 1 2 3 4 5 P trend Beta 1 2 3 4 5 P trend Age (y) + 0.32 56.3 56.9 57.3 57.6 57.6 < 0.001 + 0.33 57.9 58.8 58.8 59.1 59.4 < 0.001 2) BMI (kg/m − 0.07 25.2 25.3 25.2 25.1 25.0 0.01 + 0.05 26.0 26.1 26.0 26.1 26.2 0.09 Energy (MJ/day) + 0.23 8.1 8.4 8.7 8.9 9.1 < 0.001 + 0.21 10.7 11.2 11.4 11.6 11.6 < 0.001 Protein (E%) + 0.30 14.9 15.2 14.4 15.6 16.2 < 0.001 + 0.26 14.6 14.8 14.9 15.2 15.7 < 0.001 Fat (E%) − 1.0 40.2 39.2 38.8 37.7 36.1 < 0.001 − 0.62 40.5 40.7 40.1 39.4 38.0 < 0.001 PUFA (E%) − 0.14 6.2 6.0 6.0 5.8 5.6 < 0.001 − 0.10 6.4 6.5 6.3 6.2 6.0 < 0.001 Carbohydrate (E%) + 0.66 44.9 45.6 45.8 46.6 47.7 < 0.001 + 0.36 44.9 44.5 45.0 45.4 46.2 < 0.001 Fibre (g/MJ) + 0.21 1.8 2.1 2.2 2.4 2.7 < 0.001 + 0.14 1.6 1.8 1.9 2.0 2.2 < 0.001 Sucrose (E%) − 0.16 9.1 8.9 8.5 8.6 8.4 < 0.001 − 0.25 8.6 8.2 8.2 7.8 7.6 < 0.001 Vitamin C (g/MJ) + 2.4 18 20 21 24 28 < 0.001 + 1.6 10.9 11.2 13.3 15.7 16.7 < 0.001 b b P value P value Alcohol intake, high (g/day) 2.7 2.7 2.6 2.2 2.1 0.39 7.7 8.3 7.7 7.7 6.9 0.71 Smokers, ex/current (%) 63.1 56.6 54.0 51.1 52.0 < 0.001 78.2 72.5 70.6 68.7 63.4 < 0.001 LTP activity, high (%) 13.8 17.5 19.4 20.8 25.2 < 0.001 15.1 18.2 18.2 21.4 24.6 < 0.001 Education, high (> 10 years) (%) 21.5 27.0 29.7 34.1 38.4 < 0.001 27.1 29.2 33.5 38.1 44.3 < 0.001 Beta Q of low-fat products pattern Beta Q of low-fat products pattern Age (y) − 0.004 57.4 56.7 57.1 57.1 57.2 0.94 − 0.02 59.0 58.5 58.9 59.1 58.6 0.77 2) BMI (kg/m + 0.40 24.2 25.0 25.2 25.5 25.9 < 0.001 + 0.21 25.5 26.0 26.1 26.3 26.4 < 0.001 Energy (MJ/day) − 0.11 9.0 8.7 8.5 8.3 8.7 < 0.001 − 0.12 11.8 11.3 11.0 10.8 11.5 < 0.001 Protein (E%) + 0.41 14.4 15.3 15.6 15.8 16.2 < 0.001 + 0.41 14.0 14.8 15.2 15.3 15.8 < 0.001 Fat (E%) − 1.60 42.3 39.4 37.8 37.1 35.4 < 0.001 − 1.71 43.8 41.0 39.1 38.3 36.6 < 0.001 PUFA (E%) + 0.18 5.4 5.7 6.2 6.2 6.1 < 0.001 + 0.27 5.5 6.0 6.7 6.7 6.5 < 0.001 Carbohydrate (E%) + 1.2 43.3 45.3 46.6 47.1 48.3 < 0.001 + 1.3 42.2 44.2 45.7 46.3 47.6 < 0.001 Fibre (g/MJ) + 0.07 2.0 2.1 2.2 2.2 2.3 < 0.001 + 0.08 1.7 1.8 2.0 2.0 2.0 < 0.001 Sucrose (E%) + 0.05 8.6 8.7 8.7 8.8 8.8 0.02 − 0.03 8.2 7.9 8.0 8.3 7.9 0.25 Vitamin C (g/MJ) − 0.14 22.0 22.5 23.1 22.5 21.1 0.49 + 0.02 13.2 13.2 14.7 14.1 12.9 0.91 b b P value P value Alcohol intake, high (g/day) 2.8 3.5 2.6 2.0 1.4 < 0.001 11.1 9.7 7.0 5.8 4.8 < 0.001 Smokers, ex/current (%) 61.1 55.4 53.9 54.3 52.2 < 0.001 75.8 72.0 67.6 69.7 68.4 < 0.001 LTP activity, high (%) 20.2 20.4 19.1 18.1 19.0 0.25 17.4 19.1 20.6 21.4 19.1 0.04 Education, high (> 10 years) (%) 32.8 33.4 31.7 27.9 24.7 < 0.001 35.5 38.3 38.0 31.4 29.4 < 0.001 Beta Q of dressing/vegetables pattern Beta Q of dressing/vegetables pattern Age (years) − 2.1 61.3 59.1 57.2 55.5 52.6 < 0.001 − 1.5 61.7 60.2 59.0 57.4 55.8 < 0.001 2) BMI (kg/m + 0.10 25.0 25.1 25.1 25.1 25.5 0.001 + 0.12 25.8 26.0 25.9 26.2 26.3 < 0.001 Energy (MJ/day) + 0.03 8.5 8.6 8.7 8.7 8.6 0.06 − 0.06 11.4 11.3 11.4 11.3 11.1 0.01 Protein (E%) + 0.20 15.2 15.2 15.4 15.6 16.0 < 0.001 + 0.26 14.5 14.8 15.1 15.2 15.6 < 0.001 Fat (E%) + 0.47 37.4 37.9 38.5 39.1 39.1 < 0.001 + 0.35 38.9 39.5 39.9 40.2 40.3 < 0.001 PUFA (E%) + 0.29 5.2 5.7 6.0 6.2 6.5 < 0.001 + 0.30 5.6 6.0 6.4 6.5 6.9 < 0.001 Carbohydrate (E%) − 0.66 47.4 46.8 46.1 45.3 44.9 < 0.001 − 0.61 46.6 45.7 45.0 44.6 44.1 < 0.001 Fibre (g/MJ) + 0.02 2.1 2.2 2.2 2.2 2.3 < 0.001 + 0.03 1.8 1.9 1.9 1.9 2.0 < 0.001 Sucrose (E%) − 0.31 9.2 9.0 8.8 8.4 8.0 < 0.001 − 0.40 8.9 8.4 8.0 7.9 7.2 < 0.001 Vitamin C (g/MJ) + 1.3 18.9 21.0 23.2 24.1 23.8 < 0.001 + 1.2 11.0 12.8 13.7 14.4 16.1 < 0.001 b b P value P value Alcohol intake, high (g/day) 0.8 1.3 2.0 3.2 5.0 < 0.001 3.5 4.8 6.9 9.5 13.6 < 0.001 Smokers, ex/current (%) 49.5 51.2 56.1 57.4 62.6 < 0.001 71.4 71.7 70.0 69.4 71.0 0.58 LTP activity, high (%) 20.1 18.1 20.1 19.5 18.9 0.34 20.0 19.7 19.5 20.2 18.1 0.57 Education, high (> 10 years) (%) 15.6 23.8 28.1 37.2 45.9 < 0.001 19.4 27.9 34.4 43.0 47.6 < 0.001 1 3 European Journal of Nutrition Table 1 (continued) Calculated with the general linear model. Adjusted for age (continuous) when appropriate Chi-square test Leisure time physical activity, high = 5th quintile characteristics including higher levels of leisure time physi- T2D (P for trend = 0.07) in men, and a tendency of inverse cal activity and education, and a lower prevalence of smok- association between the “sugar-rich” pattern and T2D in ers compared to those in the lower quintiles. Women adher- women (P for trend = 0.06) (Supplementary Table 4). ing to this pattern had also higher BMI. In addition, some gender differences were observed regarding absolute intakes Coronary events of foods in the lowest and highest quintile of the “health con- scious” food pattern (Supplementary Table 2). Individuals The “health conscious” pattern was also inversely associ- adhering to the “low-fat products” pattern reported lower ated with incidence of coronary events (HR comparing the intakes of energy and fat, but higher BMI. Fewer among highest quintile with the lowest: 0.77; 95% CI 0.58–1.02; P them had high education and they tended to smoke less for trend = 0.05 and HR 0.83; 95% CI 0.68–1.01; P for trend and drink less alcohol. Those adhering to the “dressing and across quintiles = 0.07 in women and men, respectively, P vegetables” pattern were younger, had higher BMI, diets for trend in gender-combined analysis = 0.006, full multivar- with more polyunsaturated fat and protein, but less sugar. iable model) (Table 3). There seemed to be a threshold effect They tended to have higher education, smoke more and in men, indicating that it mainly were those with very low drink more alcohol. The other observed food patterns were adherence to the health conscious food pattern who were at also associated with baseline characteristics, although in a higher risk of coronary events. We observed an inverse asso- less consistent manner resulting in a less clear picture with ciation between the “dressing and vegetables” pattern and regard to overall lifestyle (Supplementary Table 3). risk of coronary events (P for trend = 0.02) in men, but no similar tendency in women. In contrast, tendency of inverse Food patterns and incidence of type 2 diabetes, association between the “sugar-rich” pattern and coronary coronary events, stroke and weight change events (P for trend = 0.053) was restricted to women (Sup- plementary Table 5). The “health conscious”, “low-fat products” and “dressing and vegetables” patterns showed significant associations Stroke with cardiometabolic diseases and the results regarding those patterns are presented in the main tables, whereas The “health conscious” pattern was associated with results regarding the other three patterns are presented as decreased incidence of stroke (HR comparing the highest supplementary data. quintile with the lowest: 0.69; 95% CI 0.54–0.88; P for trend across quintiles = 0.001, full multivariable model) (Table 4) Type 2 diabetes in men, but no such tendencies were seen in women. None of the other obtained food patterns was associated with stroke In both genders, the “health conscious” pattern was associ- (Table 4, Supplementary Table 6). ated with a significantly decreased incidence of T2D (HR comparing the highest quintile with the lowest: 0.75; 95% Weight change CI 0.61–0.92; P for trend across quintiles = 0.003 and HR 0.82; 95% CI 0.68–1.00; P for trend = 0.01 in women and The women gained on average 2.5 kg during follow-up, men, respectively, full multivariable model) (Table 2). whereas men gained 1.7 kg. Concerning the “low-fat products” pattern, no significant In men, the “health conscious” food pattern was associ- associations were observed with incidence of T2D in the ated with less pronounced weight gain during the 17-year gender-specific analysis. However, as the “low-fat products” follow-up (0.93 kg less/10 years in the highest compared pattern tended to be associated with a somewhat higher risk with the lowest quintile; P for trend across the quin- in both genders, we also performed gender-combined analy- tiles = 0.03) (Table 5). The “tea-breakfast” pattern tended sis and observed that adherence to the “low-fat products” to associate with less pronounced weight gain in women pattern was associated with increased risk of T2D (P for (Supplementary Table 7). trend = 0.03). The other patterns did not show significant Neither exclusion of BMI from the full multivariable associations with T2D, but we observed a tendency of model, nor inclusion of baseline weight or waist circumfer- inverse association between the “tea-breakfast” pattern and ence had any major influence on any of the results. Finally, 1 3 European Journal of Nutrition Table 2 Hazard ratios of type 2 diabetes across quintiles of dietary patterns in 12,456 women and 8031 men from the Malmö Diet and Cancer cohort Quintiles of dietary patterns Women Men Cases/person years HR with 95% CIs Cases/person years HR with 95% CIs Health conscious 1 228/36,371 1.00 240/22,107 1.00 2 252/37,305 1.09 (0.91, 1.31) 240/22,879 0.98 (0.81, 1.17) 3 207/37,738 0.92 (0.76, 1.11) 234/23,070 0.95 (0.79–1.15) 4 216/38,289 0.96 (0.79, 1.16) 208/23,404 0.83 (0.68-1.00) 5 167/37,756 0.75 (0.61–0.92) 214/23,931 0.82 (0.68-1.00) P trend across quintiles 0.003 0.01 ab P trend, continuous score 0.003 0.01 Low-fat products 1 173/37,574 1.00 196/22,663 1.00 2 198/37,726 1.03 (0.84, 1.26) 225/22,876 1.07 (0.88, 1.30) 3 218/38,000 0.75 (0.61, 0.92) 213/23,584 0.98 (0.80, 1.19) 4 208/37,648 1.02 (0.84, 1.26) 235/22,898 1.10 (0.90, 1.33) 5 273/37,845 1.19 (0.98, 1.45) 267/23,368 1.15 (0.96, 1.40) P trend across quintiles 0.10 0.12 ab P trend, continuous score 0.24 0.10 Dressing and vegetables 1 247/36,617 1.00 214/21,924 1.00 2 229/37,666 1.02 (0.85, 1.22) 211/22,826 0.92 (0.76, 1.12) 3 199/38,219 0.92 (0.76, 1.11) 228/23,073 1.06 (0.87, 1.28) 4 187/38,149 0.96 (0.79, 1.18) 232/23,622 1.00 (0.82, 1.21) 5 208/38,144 1.12 (0.91, 1.37) 251/23,947 1.09 (0.90, 1.33) P trend across quintiles 0.54 0.25 ab P trend, continuous score 0.33 0.24 Adjusted for age, season, diet method version, total energy intake, leisure time physical activity, smoking, alcohol intake, education, and base- line BMI P trend per unit of the pattern factor we did not observe any significant interactions with gender 1.02; 95% CI 0.86, 1.22; P trend = 0.53 in gender-combined or BMI status. analysis). In men, adherence to the “dressing and vegetable” pat- Sensitivity analyses tern was in men more strongly associated with lower risk of coronary events after excluding those above 60 years of age After exclusion of individuals above 60 years of age at base- (HR comparing the highest quintile with the lowest: 0.63; line (38% of the women and 43% of the men), the “health 95% CI 0.45, 0.80; P trend = 0.01). conscious” pattern remained significantly associated with lower risk of T2D in both women and men, and with lower risk of stroke in men. Regarding the “health conscious” pat- Discussion tern and coronary events, the risk estimates were found to be similar to those in the main analysis although non-sig- In this large study, using principal component analysis nificant. However, to gain power we also performed gender to derive food patterns from the Malmö Diet and Cancer combined analysis and then observed a clear tendency of cohort, we observed similar patterns in both women and inverse association (HR comparing the highest quintile with men, suggesting that the patterns are fairly robust. A dietary the lowest: 0.79; 95% CI 0.61, 1.02; P trend = 0.07). pattern characterized by health-conscious food choices, such The “low-fat-products” pattern was no longer associated as plant foods, fish and low-fat yoghurt, was associated with with higher risk of T2D after excluding those above 60 years decreased incidence of T2D and coronary events in both of age (HR comparing the highest quintile with the lowest: genders. In men, the “health conscious” food pattern was 1 3 European Journal of Nutrition Table 3 Hazard ratios of coronary events across quintiles of data driven dietary patterns in 12,456 women and 8031 men from the Malmö Diet and Cancer cohort Quintiles of dietary patterns Women Men Cases/person years HR with 95% CIs Cases/person years HR with 95% CIs Health conscious 1 141/37,355 1.00 229/22,762 1.00 2 121/38,277 0.90 (0.70, 1.16) 198/23,642 0.82 (0.68, 1.00) 3 96/38,843 0.72 (0.54, 0.94) 190/23,964 0.79 (0.65, 0.96) 4 108/39,233 0.83 (0.64, 1.09) 191/24,021 0.80 (0.66, 0.98) 5 101/39,789 0.77 (0.58, 1.02) 196/24,365 0.83 (0.68, 1.01) P trend across quintiles 0.054 0.07 ab P trend, continuous score 0.03 0.02 Low-fat products 1 111/38,107 1.00 191/23,285 1.00 2 97/38,633 0.91 (0.69, 1.19) 198/23,653 1.11 (0.91, 1.35) 3 119/39,065 1.11 (0.86, 1.44 180/24,496 1.00 (0.81, 1.22) 4 107/38,610 0.98 (0.75, 1.28 236/23,334 1.32 (1.09, 1.61) 5 133/39,084 1.18 (0.91, 1.52))) 199/24,288 1.06 (0.87, 1.30 P trend across quintiles 0.17 0.19 ab P trend, continuous score 0.10 0.12 Dressing and vegetables 1 165/37,711 1.00 249/22,319 1.00 2 140/38,473 1.07 (0.85, 1.34) 247/23,253 1.07 (0.90, 1.28) 3 104/39,144 0.92 (0.72, 1.18) 193/23.762 0.91 (0.75, 1.10) 4 82/39,019 0.92 (0.69, 1.21) 172/24,495 0.87 (0.71, 1.07) 5 76/39,151 1.13 (0.84, 1.53) 143/24,925 0.83 (0.66, 1.03) P trend across quintiles 0.99 0.02 ab P trend, continuous score 0.66 0.01 Adjusted for age, season, diet method version, total energy intake, leisure time physical activity, smoking, alcohol intake, education, and base- line BMI P trend per unit of the pattern factor also associated with decreased risk of stroke and with less indicated protective associations with stroke in several pronounced weight gain during 17-year follow-up. A pat- Western populations [24–26], whereas the findings in Asian tern mainly characterized by dressing and vegetables was populations are less convincing [27–29]. Retained healthy associated with lower risk of coronary events in men. The food patterns in Asia are in line with other healthy food pat- other retained food patterns did not show any significant terns defined by high intakes of various fruits and vegeta- associations with cardiometabolic diseases in women or bles, but in contrast to a lesser degree by whole grain foods. men, although adherence to the “low-fat products” pattern It is thereby possible that the value of capturing intake of was associated with higher risk of T2D in gender-combined several healthy foods, with various health-promoting prop- analysis. erties, will partly be missed. In contrast to previous reports In line with our observation, previous studies have also from Western cohorts, our finding of an inverse association observed data-driven healthy dietary patterns, mainly char- between the “health conscious” food pattern and decreased acterized by high intakes of plant foods such as fruits, veg- risk of stroke was restricted to men and no such tendencies etables and whole grain, to associate with lower incidence were seen in women. Gender differences in intake levels of of both T2D  and coronary events . Nevertheless, a various foods could at least partially explain the differing need for additional studies to confirm the findings has been observations. Intakes of fruits and vegetables in the MDC declared and associations with stroke have been less con- cohort seem, for example, to be higher in women than in sistent . The differing observations regarding stroke may men , whereas intake of refined bread seem to be higher be explained by dissimilarities in foods characterizing the in men , and it is possible that intake of healthy foods healthy patterns; healthy/prudent food patterns have indeed may lie above potential critical threshold levels with regard 1 3 European Journal of Nutrition Table 4 Hazard ratios of stroke across quintiles of data driven dietary patterns in 12,456 women and 8031 men from the Malmö Diet and Cancer cohort Quintiles of dietary patterns Women Men Cases/person years HR with 95% CIs Cases/person years HR with 95% CIs Health conscious 1 138/37,128 1.00 156/23,009 1.00 2 145/38,187 1.02 (0.81, 1.30) 142/23,807 0.83 (0.66, 1.04) 3 130/38,539 0.92 (0.72, 1.17) 125/24,032 0.72 (0.56, 0.91) 4 121/39,111 0.88 (0.68, 1.27) 123/24,247 0.70 (0.55, 0.89) 5 131/39,594 0.96 (0.75, 1.24) 121/24,647 0.69 (0.54, 0.88) P trend across quintiles 0.42 0.001 ab P trend, continuous score 0.63 0.001 Low-fat products 1 146/37,925 1.00 145/23,329 1.00 2 132/38,359 0.94 (0.74, 1.19) 138/23,733 1.02 (0.80, 1.29) 3 128/38,853 0.88 (0.70, 1.12) 125/24,352 0.89 (0.70, 1.14) 4 130/38,361 0.91 (0.72, 1.16) 136/23,725 0.99 (0.78, 1.26) 5 129/39,056 0.87 (0.68, 1.11) 123/24,602 0.87 (0.68, 1.11) P trend across quintiles 0.26 0.26 ab P trend, continuous score 0.20 0.37 Dressing and vegetables 1 192/37,412 1.00 168/22,622 1.00 2 165/38,242 1.03 (0.74, 1.27) 143/23,497 0.96 (0.77, 1.21) 3 130/38,910 0.96 (0.76, 1.20) 136/23,880 0.98 (0.78, 1.24) 4 100/38,929 0.88 (0.68, 1.14) 121/24,633 0.97 (0.76, 1.25) 5 78/39,066 0.95 (0.71, 1.27) 99/25,109 0.91 (0.70, 1.19) P trend across quintiles 0.38 0.59 ab P trend, continuous score 0.62 0.33 Adjusted for age, season, diet method version, total energy intake, leisure time physical activity, smoking, alcohol intake, education, and base- line BMI P trend per unit of the pattern factor to stroke in most women. Absolute intake ranges of healthy effects on satiety, glucose and lipid metabolism, and oxida- foods between the lowest and highest quintile of the “health tive stress, due to higher intake of plant food components conscious” food pattern may also be of importance; in men, such as fibre and phytochemicals, but lower sugar intake mean estimated daily intake of fibre-rich bread in the quin- [34–36]. Effects on gut microbiota by fibre and yoghurt may tiles ranged, for example, from 7 to 70 g, whereas it ranged also play a role [37, 38]. from 10 to 60 g in women. In agreement with our obser- Consistent with our findings regarding the “health con- vations of that men with highest adherence to the health scious” food pattern, an a priori-defined high diet quality conscious food pattern gained almost 1 kg less weight per index, based on Swedish nutrition recommendations (SNR- 10 years compared to men with lowest adherence, a dietary DQI), is associated with lower incidence of cardiovascular pattern loading high in fruits and vegetables, whole grain disease in the MDC cohort . Nevertheless, the SNR-DQI products and fish associated with less weight gain in Aus- did not associate with risk of T2D . Similar to charac- tralian men, but not women . Furthermore, no overall teristics of our data-driven “health conscious” food pattern, association was seen between a vegetable/fruit pattern and the SNR-DQI was defined by intake levels of fibre, fruits, weight-change in African-American women . On the vegetables and fish, but in contrast also by intakes of sucrose other hand, in other studies, patterns similar to our “health and different types of fat, which may explain the differing conscious” pattern have been associated with smaller weight associations. The evidence regarding the importance of fat gain also in women [8, 9, 33]. Biological mechanisms under- quality is indeed more convincing with regard to the risk of lying our observed associations between the “health con- cardiovascular disease compared to that of T2D . In this scious” pattern and cardiometabolic disease may include study, the “dressing and vegetables” pattern, which includes 1 3 European Journal of Nutrition Table 5 Weight change during follow-up in quintiles of data driven example, strong evidence for that the role of LDL choles- dietary patterns in women and men from the Malmö Diet and Cancer terol differs, as Mendelian randomization studies indicate cohort that low LDL cholesterol is associated with hyperglycaemia 10-year weight change and increased risk of T2D . Our study has several strengths. It is a large study with Quintiles of dietary patterns Women (n = 1533) Men (n = 1094) long follow-up time. Moreover, it is a population-based Health conscious prospective study, which reduces the risk of selection bias a a 1 1.93 ± 1.26 2.04 ± 1.17 and reverse causation (although as discussed above it can- a ab 2 2.10 ± 1.27 1.65 ± 1.17 not be ruled out). The relative validity of food intakes of a b 3 2.30 ± 1.25 1.12 ± 1.17 importance in this study indicates dietary data of high qual- a b 4 1.96 ± 1.26 1.40 ± 1.16 ity [13, 14]. Further strengths are the extensive information a b 5 1.97 ± 1.25 1.11 ± 1.15 on potential confounders, and that diet was measured with a P trend across quintiles 0.81 0.03 modified diet history method including a 7-day food record P trend, continuous score 0.89 0.09 for cooked meals. In other studies reporting associations Low-fat products between data-driven dietary patterns and cardiometabolic a a 1 1.77 ± 1.25 1.11 ± 1.18 disease, diet has almost exclusively been measured using a a 2 2.45 ± 1.25 1.44 ± 1.16 food frequency questionnaires, and it may be valuable to a a 3 2.07 ± 1.26 1.42 ± 1.16 also report findings from studies using other assessment a a 4 2.12 ± 1.26 1.14 ± 1.17 methods, as the method used may affect obtained food pat- a a 5 2.15 ± 1.25 1.14 ± 1.17 terns, especially if predefined food lists are used. Different P trend across quintiles 0.68 0.76 measurement methods are also associated with differing P trend, continuous score 0.91 0.78 measurement errors that could affect the results. Finally, we Dressing and vegetables observed similar food patterns in women and men, indicat- a a 1 1.96 ± 1.25 1.31 ± 1.18 ing robust patterns. a a 2 1.84 ± 1.26 1.56 ± 1.17 A limitation of the study is that diet and other lifestyle a a 3 1.81 ± 1.25 1.28 ± 1.16 factors were only measured at baseline. Another drawback a a 4 1.90 ± 1.26 1.08 ± 1.16 may be potentially subjective decisions regarding the group- b a 5 2.46 ± 1.25 1.55 ±1.16 ing of food variables to include in the principal component P trend across quintiles 0.20 0.99 analysis, as it could influence the obtained dietary pat- P trend, continuous score 0.16 0.93 terns. Our aim was to cover as many parts of the overall diet as possible, but to have a less detailed level on foods Homogenous subsets are indicated by letters. Adjusted for age, sea- son, diet method version, total energy intake, leisure time physical reported during the record period and known to be con- activity, smoking, alcohol intake, education, and baseline BMI sumed irregularly, as they may be unsatisfactorily captured P trend per unit of the pattern factor on a 7-day basis if they are not aggregated. This was, for example, the reason to why we grouped fatty and lean fish together. Although it has been indicated that the number of oil-based dressings, was characterized by a markedly higher input variables do not have any major effect on the derived dietary content of polyunsaturated fat. Men adhering to this food patterns , we cannot exclude that a different detail pattern had indeed a lower risk for coronary events. In con- level regarding the food classification would have revealed trast, the “low-fat products” food pattern tended to associ- somewhat altered patterns. Another concern is that dietary ate with higher risk of T2D. However, we cannot exclude patterns could represent overall lifestyle, and despite adjust- that these findings could be a result of reverse causation and ment for several confounders, we cannot completely rule out the association did not persist when excluding individuals residual confounding. Finally, findings from dietary pattern above 60 years of age. It is possible that many of the indi- analyses may be explained by specific foods or nutrients viduals adhering to this pattern choose low-fat products to related to the pattern, and it can be challenging to reveal lose weight and dietary habits earlier in life may explain whether certain components lie behind the results or whether the high BMI as well as their higher risk of T2D, because the whole pattern is crucial . BMI at baseline was actually higher among those adhering The “health conscious” pattern, which explained most to this pattern. On the other hand, the findings regarding the of the variance in food intakes in this study, was strong- “low-fat product” pattern are in line with those of a previous est related to several foods that per se have been found to findings, suggesting that some dairy fats may contribute to associate with cardiometabolic disease in other studies lowering the risk of T2D [41, 42]. Risk factors may indeed [34–36, 46–50]. Furthermore, these foods have previously differ between T2D and coronary disease and there is, for been included in predefined healthy diet indexes, suggesting 1 3 European Journal of Nutrition 5. Rodriguez-Monforte M, Flores-Mateo G, Sanchez E (2015) Die- that the “health conscious” pattern may also be useful when tary patterns and CVD: a systematic review and meta-analysis of examining associations with other chronic disease such as observational studies. Br J Nutr 114(9):1341–1359. https ://doi. different types of cancer. Future studies will reveal the rel- org/10.1017/S0007 11451 50031 77 evance of our observations indicating that patterns charac- 6. Arabshahi S, Ibiebele TI, Hughes MCB, Lahmann PH, Wil- liams GM, van der Pols JC (2017) Dietary patterns and weight terized low-fat products and fat quality may be differently change: 15-year longitudinal study in Australian adults. Eur J Nutr related to T2D compared to coronary disease. 56(4):1455–1465. https ://doi.org/10.1007/s0039 4-016-1191-3 To conclude, our findings indicate that adhering to a 7. Newby PK, Weismayer C, Akesson A, Tucker KL, Wolk A (2006) “health conscious” food pattern, characterized by high Longitudinal changes in food patterns predict changes in weight and body mass index and the effects are greatest in obese women. intake of fibre-rich plant foods, fish and low-fat yoghurt, but J Nutr 136(10):2580–2587 low intake of low-fibre bread, sugar-sweetened beverages, 8. Schulz M, Nothlings U, Hoffmann K, Bergmann MM, Boeing H and red and processed meat, which is in line with European (2005) Identification of a food pattern characterized by high-fiber dietary guidelines , may contribute to decreased risk of and low-fat food choices associated with low prospective weight change in the EPIC-Potsdam cohort. J Nutr 135(5):1183–1189 cardiometabolic disease and especially to decreased risk of 9. Schulze MB, Fung TT, Manson JE, Willett WC, Hu FB (2006) T2D. Dietary patterns and changes in body weight in women. Obe- sity (Silver Spring) 14(8):1444–1453. https ://doi.org/10.1038/ Acknowledgements We thank the staff and participants of the Malmö oby.2006.164 Diet and Cancer Study. 10. Manjer J, Carlsson S, Elmstahl S, Gullberg B, Janzon L, Lind- strom M, Mattisson I, Berglund G (2001) The Malmo Diet Author contributions UE designed the research, performed statistical and Cancer Study: representativity, cancer incidence and mor- analysis and wrote the paper; UE, LB, JAD, ID, SH, BG, ES, PMN, EW tality in participants and non-participants. Eur J Cancer Prev and MOM contributed to the interpretation of results and revision of 10(6):489–499 the manuscript; BG gave statistical advice. UE had primary responsi- 11. Callmer E, Riboli E, Saracci R, Akesson B, Lindgarde F (1993) bility for final content. All authors read and approved the final version. Dietary assessment methods evaluated in the Malmo food study. J Intern Med 233(1):53–57 12. Wirfalt E, Mattisson I, Johansson U, Gullberg B, Wallstrom P, Funding The Swedish Research Council, the Region Skåne, the Skåne Berglund G (2002) A methodological report from the Malmo Diet University Hospital, the Novo Nordic Foundation, the Albert Påhlsson and Cancer study: development and evaluation of altered routines Research Foundation. in dietary data processing. Nutr J 1:3 13. Elmstahl S, Riboli E, Lindgarde F, Gullberg B, Saracci R (1996) Compliance with ethical standards The Malmo Food Study: the relative validity of a modified diet history method and an extensive food frequency questionnaire for Conflict of interest On behalf of all authors, the corresponding author measuring food intake. Eur J Clin Nutr 50(3):143–151 states that there is no conflict of interest. 14. Riboli E, Elmstahl S, Saracci R, Gullberg B, Lindgarde F (1997) The Malmo Food Study: validity of two dietary assessment methods for measuring nutrient intake. Int J Epidemiol 26(Suppl Open Access This article is distributed under the terms of the Crea- 1):S161–S173 tive Commons Attribution 4.0 International License (http://creat iveco 15. Berglund G, Elmstahl S, Janzon L, Larsson SA (1993) The mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- Malmo Diet and Cancer Study. Design and feasibility. J Intern tion, and reproduction in any medium, provided you give appropriate Med 233(1):45–51 credit to the original author(s) and the source, provide a link to the 16. Schwab U, Lauritzen L, Tholstrup T, Haldorssoni T, Riserus U, Creative Commons license, and indicate if changes were made. Uusitupa M, Becker W (2014) Effect of the amount and type of dietary fat on cardiometabolic risk factors and risk of developing type 2 diabetes, cardiovascular diseases, and cancer: a systematic review. Food Nutr Res. https ://doi.org/10.3402/fnr.v58.25145 17. Micha R, Shulkin ML, Penalvo JL, Khatibzadeh S, Singh GM, References Rao M, Fahimi S, Powles J, Mozaffarian D (2017) Etiologic effects and optimal intakes of foods and nutrients for risk of car - 1. Jacobs DR Jr, Tapsell LC (2007) Food, not nutrients, is the fun- diovascular diseases and diabetes: systematic reviews and meta- damental unit in nutrition. Nutr Rev 65(10):439–450 analyses from the Nutrition and Chronic Diseases Expert Group 2. Newby PK, Tucker KL (2004) Empirically derived eating patterns (NutriCoDE). Plos One 12(4):e0175149. https://doi.or g/10.1371/ using factor or cluster analysis: a review. Nutr Rev 62(5):177–203 journ al.pone.01751 49 3. Maghsoudi Z, Ghiasvand R, Salehi-Abargouei A (2016) Empiri- 18. Lindholm E, Agardh E, Tuomi T, Groop L, Agardh CD (2001) cally derived dietary patterns and incident type 2 diabetes mel- Classifying diabetes according to the new WHO clinical stages. litus: a systematic review and meta-analysis on prospective obser- Eur J Epidemiol 17(11):983–989 vational studies. Public Health Nutr 19(2):230–241. https ://doi. 19. Cederholm J, Eeg-Olofsson K, Eliasson B, Zethelius B, Nilsson org/10.1017/S1368 98001 50012 51 PM, Gudbjornsdottir S (2008) Risk prediction of cardiovascu- 4. McEvoy CT, Cardwell CR, Woodside JV, Young IS, Hunter SJ, lar disease in type 2 diabetes: a risk equation from the Swedish McKinley MC (2014) A posteriori dietary patterns are related National Diabetes Register. Diabetes Care 31(10):2038–2043. to risk of type 2 diabetes: findings from a systematic review and https ://doi.org/10.2337/dc08-0662 meta-analysis. J Acad Nutr Diet 114(11):1759–1775 e1754. https 20. Hanas R, John G (2010) 2010 consensus statement on the world- ://doi.org/10.1016/j.jand.2014.05.001 wide standardization of the hemoglobin A1C measurement. Dia- betes Care 33(8):1903–1904. https ://doi.org/10.2337/dc10-0953 1 3 European Journal of Nutrition 21. Hoelzel W, Weykamp C, Jeppsson JO, Miedema K, Barr JR, cardiovascular disease, and weight gain. J Nutr 142(7):1304–1313. Goodall I, Hoshino T, John WG, Kobold U, Little R, Mosca A, https ://doi.org/10.3945/jn.111.15532 5 Mauri P, Paroni R, Susanto F, Takei I, Thienpont L, Umemoto 36. Alissa EM, Ferns GA (2017) Dietary fruits and vegetables and M, Wiedmeyer HM (2004) IFCC reference system for measure- cardiovascular diseases risk. Crit Rev Food Sci Nutr 57(9):1950– ment of hemoglobin A1c in human blood and the national stand- 1962. https ://doi.org/10.1080/10408 398.2015.10404 87 ardization schemes in the United States, Japan, and Sweden: a 37. Brunkwall L, Orho-Melander M (2017) The gut microbiome as method-comparison study. Clin Chem 50(1):166–174. https://doi. a target for prevention and treatment of hyperglycaemia in type org/10.1373/clinc hem.2003.02480 2 2 diabetes: from current human evidence to future possibili- 22. Socialstyrelsen. (Stockholm 2000) Värdering av diagnoskvak- ties. Diabetologia 60(6):943–951. https ://doi.org/10.1007/s0012 iteten för akut hjärtinfarkt i patientregistret 1987 och 1995 5-017-4278-3 23. Zia E, Pessah-Rasmussen H, Khan FA, Norrving B, Janzon L, 38. Jonsson AL, Backhed F (2017) Role of gut microbiota in athero- Berglund G, Engstrom G (2006) Risk factors for primary intrac- sclerosis. Nat Rev Cardiol 14(2):79–87. https ://doi.org/10.1038/ erebral hemorrhage: a population-based nested case–control study. nrcar dio.2016.183 Cerebrovasc Dis 21(1–2):18–25. https ://doi.org/10.1159/00008 39. Hlebowicz J, Drake I, Gullberg B, Sonestedt E, Wallstrom P, Pers- 9589 son M, Nilsson J, Hedblad B, Wirfalt E (2013) A high diet quality 24. Fung TT, Stampfer MJ, Manson JE, Rexrode KM, Willett WC, is associated with lower incidence of cardiovascular events in the Hu FB (2004) Prospective study of major dietary patterns and Malmo diet and cancer cohort. PLoS One 8(8):e71095. https :// stroke risk in women. Stroke 35(9):2014–2019. h ttp s :/ /do i.doi.org/10.1371/journ al.pone.00710 95 org/10.1161/01.STR.00001 35762 .89154 .92 40. Mandalazi E, Drake I, Wirfalt E, Orho-Melander M, Sonestedt 25. Judd SE, Gutierrez OM, Newby PK, Howard G, Howard VJ, E (2016) A high diet quality based on dietary recommenda- Locher JL, Kissela BM, Shikany JM (2013) Dietary patterns are tions is not associated with lower incidence of type 2 diabetes associated with incident stroke and contribute to excess risk of in the Malmo diet and cancer cohort. Int J Mol Sci. https ://doi. stroke in black Americans. Stroke 44(12):3305–3311. https://doi. org/10.3390/ijms1 70609 01 org/10.1161/STROK EAHA.113.00263 6 41. Forouhi NG, Koulman A, Sharp SJ, Imamura F, Kroger J, 26. Stricker MD, Onland-Moret NC, Boer JM, van der Schouw YT, Schulze MB, Crowe FL, Huerta JM, Guevara M, Beulens JW, Verschuren WM, May AM, Peeters PH, Beulens JW (2013) van Woudenbergh GJ, Wang L, Summerhill K, Griffin JL, Feskens Dietary patterns derived from principal component- and k-means EJ, Amiano P, Boeing H, Clavel-Chapelon F, Dartois L, Fagher- cluster analysis: long-term association with coronary heart disease azzi G, Franks PW, Gonzalez C, Jakobsen MU, Kaaks R, Key TJ, and stroke. Nutr Metab Cardiovasc Dis 23 (3):250–256. https :// Khaw KT, Kuhn T, Mattiello A, Nilsson PM, Overvad K, Pala doi.org/10.1016/j.numec d.2012.02.006 V, Palli D, Quiros JR, Rolandsson O, Roswall N, Sacerdote C, 27. Chan R, Chan D, Woo J (2013) The association of a priori Sanchez MJ, Slimani N, Spijkerman AM, Tjonneland A, Tormo and a posterior dietary patterns with the risk of incident stroke MJ, Tumino R, van der AD, van der Schouw YT, Langenberg in Chinese older people in Hong Kong. J Nutr Health Aging C, Riboli E, Wareham NJ (2014) Differences in the prospective 17(10):866–874. https ://doi.org/10.1007/s1260 3-013-0334-y association between individual plasma phospholipid saturated 28. Maruyama K, Iso H, Date C, Kikuchi S, Watanabe Y, Wada Y, fatty acids and incident type 2 diabetes: the EPIC-InterAct case- Inaba Y, Tamakoshi A, Group JS (2013) Dietary patterns and cohort study. Lancet Diabetes Endocrinol 2(10):810–818. https:// risk of cardiovascular deaths among middle-aged Japanese: JACC doi.org/10.1016/S2213 -8587(14)70146 -9 Study. Nutr Metab Cardiovasc Dis 23 (6):519–527. https ://doi. 42. Ericson U, Hellstrand S, Brunkwall L, Schulz CA, Sonestedt E, org/10.1016/j.numec d.2011.10.007 Wallstrom P, Gullberg B, Wirfalt E, Orho-Melander M (2015) 29. Chen Y, McClintock TR, Segers S, Parvez F, Islam T, Ahmed A, Food sources of fat may clarify the inconsistent role of dietary Rakibuz-Zaman M, Hasan R, Sarwar G, Ahsan H (2013) Prospec- fat intake for incidence of type 2 diabetes. Am J Clin Nutr tive investigation of major dietary patterns and risk of cardiovas- 101(5):1065–1080. https ://doi.org/10.3945/ajcn.114.10301 0 cular mortality in Bangladesh. Int J Cardiol 167(4):1495–1501. 43. Schmidt AF, Swerdlow DI, Holmes MV, Patel RS, Fairhurst- https ://doi.org/10.1016/j.ijcar d.2012.04.041 Hunter Z, Lyall DM, Hartwig FP, Horta BL, Hypponen E, Power 30. Wallstrom P, Wirfalt E, Janzon L, Mattisson I, Elmstahl S, Johans- C, Moldovan M, van Iperen E, Hovingh GK, Demuth I, Norman son U, Berglund G (2000) Fruit and vegetable consumption in K, Steinhagen-Thiessen E, Demuth J, Bertram L, Liu T, Coassin relation to risk factors for cancer: a report from the Malmo Diet S, Willeit J, Kiechl S, Willeit K, Mason D, Wright J, Morris R, and Cancer Study. Public Health Nutr 3(3):263–271 Wanamethee G, Whincup P, Ben-Shlomo Y, McLachlan S, Price 31. Ericson U, Sonestedt E, Gullberg B, Hellstrand S, Hindy G, Wir- JF, Kivimaki M, Welch C, Sanchez-Galvez A, Marques-Vidal P, falt E, Orho-Melander M (2013) High intakes of protein and pro- Nicolaides A, Panayiotou AG, Onland-Moret NC, van der Schouw cessed meat associate with increased incidence of type 2 diabetes. YT, Matullo G, Fiorito G, Guarrera S, Sacerdote C, Wareham NJ, Br J Nutr 109(6):1143–1153. https://doi.or g/10.1017/S000711451 Langenberg C, Scott R, Luan J, Bobak M, Malyutina S, Pajak A, 20030 17 Kubinova R, Tamosiunas A, Pikhart H, Husemoen LL, Grarup 32. Boggs DA, Palmer JR, Spiegelman D, Stampfer MJ, Adams- N, Pedersen O, Hansen T, Linneberg A, Simonsen KS, Cooper Campbell LL, Rosenberg L (2011) Dietary patterns and 14-y J, Humphries SE, Brilliant M, Kitchner T, Hakonarson H, Car- weight gain in African American women. Am J Clin Nutr rell DS, McCarty CA, Kirchner HL, Larson EB, Crosslin DR, de 94(1):86–94. https ://doi.org/10.3945/ajcn.111.01348 2 Andrade M, Roden DM, Denny JC, Carty C, Hancock S, Attia J, 33. Newby PK, Muller D, Hallfrisch J, Andres R, Tucker KL (2004) Holliday E, O’Donnell M, Yusuf S, Chong M, Pare G, van der Food patterns measured by factor analysis and anthropometric Harst P, Said MA, Eppinga RN, Verweij N, Snieder H, LifeLines changes in adults. Am J Clin Nutr 80(2):504–513 Cohort study g, Christen T, Mook-Kanamori DO, Gustafsson S, 34. Stanhope KL (2016) Sugar consumption, metabolic disease Lind L, Ingelsson E, Pazoki R, Franco O, Hofman A, Uitterlinden and obesity: the state of the controversy. Crit Rev Clin Lab Sci A, Dehghan A, Teumer A, Baumeister S, Dorr M, Lerch MM, 53(1):52–67. https ://doi.org/10.3109/10408 363.2015.10849 90 Volker U, Volzke H, Ward J, Pell JP, Smith DJ, Meade T, Mait- 35. Ye EQ, Chacko SA, Chou EL, Kugizaki M, Liu S (2012) Greater land-van der Zee AH, Baranova EV, Young R, Ford I, Campbell whole-grain intake is associated with lower risk of type 2 diabetes, A, Padmanabhan S, Bots ML, Grobbee DE, Froguel P, Thuillier 1 3 European Journal of Nutrition D, Balkau B, Bonnefond A, Cariou B, Smart M, Bao Y, Kumari (2012) The prospective association between total and type of fish M, Mahajan A, Ridker PM, Chasman DI, Reiner AP, Lange LA, intake and type 2 diabetes in 8 European countries: EPIC-InterAct Ritchie MD, Asselbergs FW, Casas JP, Keating BJ, Preiss D, Hin- Study. Am J Clin Nutr 95(6):1445–1453. https://doi.or g/10.3945/ gorani AD, Consortium U, Sattar N (2017) PCSK9 genetic vari-ajcn.111.02931 4 ants and risk of type 2 diabetes: a mendelian randomisation study. 48. Aune D, Norat T, Romundstad P, Vatten LJ (2013) Whole grain Lancet Diabetes Endocrinol 5(2):97–105. https://doi.or g/10.1016/ and refined grain consumption and the risk of type 2 diabetes: S2213 -8587(16)30396 -5 a systematic review and dose-response meta-analysis of cohort 44. McCann SE, Marshall JR, Brasure JR, Graham S, Freudenheim studies. Eur J Epidemiol 28(11):845–858. https://doi.or g/10.1007/ JL (2001) Analysis of patterns of food intake in nutritional epi-s1065 4-013-9852-5 demiology: food classification in principal components analysis 49. Wang M, Yu M, Fang L, Hu RY (2015) Association between and the subsequent impact on estimates for endometrial cancer. sugar-sweetened beverages and type 2 diabetes: a meta-analysis. J Public Health Nutr 4(5):989–997 Diabetes Investig 6(3):360–366. https://doi.or g/10.1111/jdi.12309 45. Willett W (2013) Nutritional epidemiology. Monographs in epi- 50. Sluijs I, Forouhi NG, Beulens JW, van der Schouw YT, Agnoli demiology and biostatistics, vol 40, 3rd edn. Oxford University C, Arriola L, Balkau B, Barricarte A, Boeing H, Bueno-de-Mes- Press, New York quita HB, Clavel-Chapelon F, Crowe FL, de Lauzon-Guillain B, 46. Micha R, Wallace SK, Mozaffarian D (2010) Red and processed Drogan D, Franks PW, Gavrila D, Gonzalez C, Halkjaer J, Kaaks meat consumption and risk of incident coronary heart disease, R, Moskal A, Nilsson P, Overvad K, Palli D, Panico S, Quiros JR, stroke, and diabetes mellitus: a systematic review and meta-anal- Ricceri F, Rinaldi S, Rolandsson O, Sacerdote C, Sanchez MJ, ysis. Circulation 121(21):2271–2283. https ://doi.org/10.1161/ Slimani N, Spijkerman AM, Teucher B, Tjonneland A, Tormo CIRCU LATIO NAHA.109.92497 7 MJ, Tumino R, van der AD, Sharp SJ, Langenberg C, Feskens EJ, 47. Patel PS, Forouhi NG, Kuijsten A, Schulze MB, van Woudenbergh Riboli E, Wareham NJ, InterAct C (2012) The amount and type GJ, Ardanaz E, Amiano P, Arriola L, Balkau B, Barricarte A, of dairy product intake and incident type 2 diabetes: results from Beulens JW, Boeing H, Buijsse B, Crowe FL, de Lauzon-Guillan the EPIC-InterAct Study. Am J Clin Nutr 96(2):382–390. https:// B, Fagherazzi G, Franks PW, Gonzalez C, Grioni S, Halkjaer J, doi.org/10.3945/ajcn.111.02190 7 Huerta JM, Key TJ, Kuhn T, Masala G, Nilsson P, Overvad K, 51. Montagnese C, Santarpia L, Buonifacio M, Nardelli A, Caldara Panico S, Quiros JR, Rolandsson O, Sacerdote C, Sanchez MJ, AR, Silvestri E, Contaldo F, Pasanisi F (2015) European food- Schmidt EB, Slimani N, Spijkerman AM, Teucher B, Tjonneland based dietary guidelines: a comparison and update. Nutrition A, Tormo MJ, Tumino R, van der AD, van der Schouw YT, Sharp 31(7–8):908–915. https ://doi.org/10.1016/j.nut.2015.01.002 SJ, Langenberg C, Feskens EJ, Riboli E, Wareham NJ, InterAct C 1 3
European Journal of Nutrition – Springer Journals
Published: May 31, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera