Health behaviours as a predictor of quitting hazardous alcohol use in the Stockholm Public Health Cohort

Health behaviours as a predictor of quitting hazardous alcohol use in the Stockholm Public Health... Abstract Background Adopting healthy behaviours may facilitate the transition from hazardous to non-hazardous use of alcohol, yet, longitudinal studies of health behaviours in relation to the cessation of hazardous alcohol use are rare. We addressed this question using data from a large population-based cohort of adults in Sweden (Stockholm Public Health Cohort). Methods Participants from two sub-cohorts (inception in 2002 and 2010), with follow-up until the year 2014 were included. Health behaviours (tobacco use, diet and physical activity) and alcohol use were self-reported in questionnaire-based surveys. Hazardous alcohol use was defined as either usual weekly consumption (2002 sub-cohort) or heavy occasional alcohol consumption (2010 sub-cohort). Baseline hazardous drinkers with complete data constituted the analytical sample (n = 8946). Logistic regression was used to calculate the Odds Ratios and their 95% confidence intervals of quitting hazardous alcohol use, with tobacco use, diet and physical activity as predictors of change. Results In the 2002 sub-cohort, 28% reported non-hazardous use sustained through two consecutive follow-up points. In the 2010 sub-cohort, 36% of the participants reported non-hazardous use of alcohol at follow-up. Favourable health behaviours at baseline (e.g. no tobacco use, sufficient fruit intake and physical activity) were associated with a 19% to 75% higher of odds quitting hazardous alcohol use. Further, favourable changes in diet and tobacco cessation were associated with non-hazardous alcohol use at follow-up. Conclusions As many as one-third of hazardous alcohol users may quit this drinking pattern in a medium-long term. Holding or achieving a healthy lifestyle may facilitate this transition. Introduction Alcohol use is one of the leading preventable causes of premature death and disease. Worldwide, 2.3 million premature deaths and 3.4% of disability-adjusted life years (DALYs) were attributable to alcohol use in 2015.1 In Sweden, 2.6% of DALYs were due to alcohol,1 and the prevalence of hazardous and harmful alcohol use was estimated to 15% in 2015.2 Furthermore, hazardous alcohol use leads to a substantial economic burden to the society.3 Hazardous alcohol use refers to consumption patterns that increase the risk of adverse health consequences, while harmful use is defined as a behaviour already causing damage to health. Average consumption is one quantitative measure of hazardous and harmful use. Another aspect includes episodes of heavy drinking, which in itself convey important health consequences.4 To date, there is no universal consensus about which thresholds in these measures mark the transition between hazardous and non-hazardous alcohol use, with consequent heterogeneity in this definition between studies. During the last decade’s research has shifted its focus from alcohol dependence to the continuum from low risk to hazardous use.5 Along this pathway, the transition from hazardous and harmful to non-hazardous use is poorly understood. In fact, there is limited knowledge on the extent to which this behaviour naturally recedes over the lifetime.6,7 Spontaneous recovery seems frequent, but the success rate depends on the definition of the alcohol problem, and of recovery, as well as on the timespan.8,9 A better understanding of which factors promote a reduction of alcohol use is needed to design effective interventions, ultimately leading to a reduction of harm at the societal level. Previous studies mainly addressed demographics (e.g. sex, age, education) as predictors for change in hazardous alcohol use and recovery.9–11 A few studies described the co-occurrence of unfavourable health behaviours among hazardous drinkers,12 but whether and how changes in health behaviours are associated with changes in hazardous alcohol use longitudinally is poorly understood, particularly in Sweden. Unfavourable health behaviours tend to cluster,13 leading to elevated risks of comorbidities.14 Thus, intervention strategies targeting associated health behaviours may prove of utmost importance. The aim of this study was 2-fold: to describe cumulative incidence of the transition from hazardous to non-hazardous alcohol use in a population sample of Swedish adults, and to examine whether modifiable health behaviours, i.e. smoking, other tobacco use, physical activity and dietary habits predict quitting the hazardous use of alcohol. We hypothesized that holding favourable health behaviours or changing from unhealthy to healthier behaviours (e.g. smoking cessation, increasing physical activity), predicts the transition from hazardous to low-risk alcohol use. Methods Study design and participants Details on the study design and collection of information in the Stockholm Public Health Cohort (SPHC) have been reported by Svensson et al.15 Briefly, SPHC is a prospective study, which has enrolled sub-cohorts in 2002, 2006 and 2010. Each of the sub-cohorts were followed every 4th year until 2014. The initial samples were randomly selected from the general population of Stockholm County stratified by the municipality, and each included about 50 000 residents 18 years of age or older. The response rate was 62%, 61% and 56% in the sub-cohorts 2002, 2006 and 2010, respectively.15 Participants gave informed consent at inception. Additional consent to national register linkage and participation in longitudinal data collection for the 2002 sub-cohort was obtained in 2006, which led to the effective recruitment of 47% of the baseline sample (n = 23 771). Due to considerable modifications of the questionnaire in 2006, we did not include information from the follow-up or the sub-cohort in this year. Overall, SPHC responders were more likely to be female, over 45 years of age, born in Sweden and to have higher education and income than the average of the population reported in the Stockholm census data.15 We analyzed data from participants recruited in 2002 and 2010 who remained in the study until 2014, and who at baseline reported alcohol consumption above the cut-off for hazardous use. The response rate in 2014 was 64% in the 2002 sub-cohort, and 56% in the 2010 sub-cohort, corresponding to samples of 3856 and 5090 individuals, respectively, see figure 1. Ethical approval was obtained from the Ethical Review Board of Stockholm Region (DNR 2016/749-2). Figure 1 View largeDownload slide Flowchart of the derivation of the study samples, the SPHC 2002–14. UWAC usual weekly alcohol consumption and HOAC Heavy occasional alcohol consumption UWAC and HOAC. ∗Respondents among those who gave consent to register linkage in (total respondents in 2002 = 31 182) Figure 1 View largeDownload slide Flowchart of the derivation of the study samples, the SPHC 2002–14. UWAC usual weekly alcohol consumption and HOAC Heavy occasional alcohol consumption UWAC and HOAC. ∗Respondents among those who gave consent to register linkage in (total respondents in 2002 = 31 182) Measurements Usual weekly alcohol consumption In the 2002 sub-cohort, self-reported usual consumption of alcoholic beverages (spirits, fortified wines, wine, alcopop or cider, medium-strong beer and strong beer) was collected with reference to a standard week during the past year.16 Usual weekly alcohol consumption (UWAC) was measured in centilitres, exemplified as glasses and bottles or cans (2002) and in numbers of standard drinks (2010, 2014); after that, a measure of grams of pure ethanol per week was derived. UWAC was dichotomized into non-hazardous or hazardous use based on commonly used cut-offs in Sweden, i.e. a weekly consumption of >108 gram (>9 standard drinks) for women and >168 gram (>14 standard drinks) for men.17 In this sub-cohort, quitting hazardous alcohol use was defined as ‘reporting consumption below the cut-off in two consecutive follow-up surveys (2010 and 2014)’. Heavy occasional alcohol consumption In the 2010 sub-cohort, heavy occasional alcohol consumption (HOAC) was measured by events of drinking alcohol equivalent to five standard glasses (1 bottle of wine; 5 shots of spirits; 4 cans of strong beer; 6 cans of medium-strong beer), in the past year.18 HOAC was dichotomized to non-hazardous (less than monthly) and hazardous (at least monthly frequency of HOAC). In this sub-cohort, quitting hazardous alcohol use was defined as ‘reporting a frequency of heavy consumption less than monthly in 2014’. Health behaviours Two similar questions were used to measure current daily tobacco use, ‘Do you currently smoke daily or almost daily’ and ‘Do you currently use snus (the Swedish type of moist oral snuff) daily or almost daily?’19 Daily tobacco use was derived from these two separate questions, as any daily use of either cigarettes or snus. Quitting tobacco use was defined as a switch from daily to non-current daily use at follow-up. Usual frequency of fruit consumption during the past year was used to derive an indicator of favourable eating patterns. In the 2002 survey, sufficient consumption was defined by the response alternative ‘nearly daily or more frequently’. Following a modification of the survey question in 2010, sufficient consumption was defined by the alternative ‘6 times per week or more’. Following a modification of the survey question in 2010,20 daily consumption was defined by the alternative ‘6 times per week or more’. A switch from less than daily to daily consumption of fruit at follow-up was categorized as a favourable change. Physical activity was assessed as the usual level of leisure time physical activity during the past year. In the 2002 sub-cohort, sufficient leisure time physical activity was defined as ‘regular physical activity 1–2 times per week at a moderate level’. In the 2010 and 2014 surveys, the question and response alternatives were rephrased to include several domains.21 Therefore, in the 2010 sub-cohort, we chose the response alternative ‘2–3 h of exercise per week’ as the threshold for sufficient physical activity, in line with the current recommendations of a minimum of 150 min per week of moderate to intense physical activity.22 A switch from insufficient to sufficient levels of physical activity at follow-up was categorized as a favourable change. In the 2002 sub-cohort, behavioural changes were assessed in 2010, while in the 2010 sub-cohort they were assessed in 2014. Sociodemographic characteristics Employment and house-hold composition were based on self-reports while age, sex, education and country of birth were based on register data. Employment was derived from the question ‘What is your current employment status?’ Answers were coded as ‘employed’ (employed, self-employed, students and parental leave) or ‘not employed’ (long-term sick leave, age pension, disability pension, unemployed and others). House-hold composition was based on the question ‘Whom do you share a house-hold with?’ Answers were coded as ‘cohabitation’ (living with any other adult or child) or ‘living alone’. Education was grouped into three levels: ‘low’ (compulsory school and vocational training), ‘intermediate’ (secondary school) and ‘high’ (university studies). Self-rated health was assessed by the question ‘In general, how would you rate your health?’ Answers were coded as ‘good health’ (response alternatives good and excellent), ‘fair health’ (fair) and ‘poor health’ (bad or very bad).23 Statistical analysis Univariate analysis was conducted to compare baseline sociodemographic characteristics and health behaviours between hazardous alcohol consumers who did or did not change drinking pattern at follow-up. The departure of the observed from the expected proportions was explored using the chi-square statistic. Binary logistic regression was used to calculate Odds Ratios (OR) and the corresponding 95% Confidence Intervals (CI) of quitting hazardous alcohol use (outcome variable in this study). We conducted two sets of analyzes. First, baseline values of each health behaviour were regressed on the outcome. Second, favourable modifications of health behaviours, e.g. quitting tobacco use were modelled as predictors of quitting hazardous alcohol use. These latter analyzes were restricted to hazardous drinkers with unfavourable health behaviours at baseline. The results are presented with and without adjustments for potential confounders. These were selected based on prior explanatory models of behavioural change (i.e. age, sex, education, employment, cohabitation status and self-rated health at baseline), and were kept in the final model only if significantly associated with the outcome. The conventional significance level was set at P < 0.05. Data analysis was performed using STATA 14.1. Results Hazardous alcohol consumers retained at follow-up, were more likely to be female, born in Sweden, older, have higher education and be non-smokers at baseline than non-responders (data not shown). Also, participants tended to have a slightly lower intake of alcohol at baseline (UWAC) compared to non-participants, while no difference were seen for HOAC. Missing information did not exceed five percent for all variables used in the analysis. Table 1 shows the distribution of characteristics of the two sub-cohorts according to changes in drinking patterns. The probability of sustained change, i.e. record of non-hazardous alcohol use at 8 and 12 years follow-up was 28.3% (sub-cohort 2002). The probability of having quit hazardous use at 4 years follow-up was 35.9% (sub-cohort 2010). Likewise, the point prevalence of non-hazardous use in the 2002 sub-cohort was 36% after 8 years and 46% after 12 years follow-up (data not shown). Table 1 Characteristics of participants with hazardous alcohol use at baseline who were followed until 2014, using two definitions of hazardous alcohol use, UWAC and HOAC   2002 Sub-cohort  2010 Sub-cohort  Alcohol pattern  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  UWAC, n (%)  2764 (71.7)  1092 (28.3)  3856            HOAC, n (%)          3261 (64.1)  1829 (35.9)  5090    Characteristics      Age, m (sd)  49.0 (12.8)  43.8 (15.4)  47.5 (13.8)  <0.01  48.1 (15.7)  46.6 (16.8)  47.5 (16.1)  <0.01  Sex      Female  70.3  29.7  2140 (55.5)  <0.05  55.2  44.8  2021 (39.7)  <0.001      Male  73.4  26.6  1716 (44.5)    69.9  30.1  3069 (60.3)    Education      Low  70.7  29.3  492 (13.0)  0.59  68.1  31.9  865 (17.1)  <0.01      Intermediate  72.7  27.3  1588 (41.8)    65.1  34.9  2145 (42.5)        High  71.4  28.6  1718 (45.2)    61.3  38.7  2038 (40.4)    Employment      Yes  71.4  28.6  3109 (82.0)  0.74  64.4  35.6  3878 (77.3)  0.45      No  72.1  27.9  684 (18.0)    63.1  36.9  1140 (22.7)    Country of birth      Sweden  71.6  28.4  3527 (91.5)  0.32  64.5  35.5  4497 (88.4)  <0.01      Nordic  76.3  23.7  186 (4.8)    69.0  31.0  226 (4.4)        Europe  66.7  33.3  99 (2.6)    60.1  39.9  178 (3.5)        Outside of Europe  68.2  31.8  44 (1.1)    51.9  48.1  189 (3.7)    Cohabitation      Yes  73.8  26.2  3140 (81.7)  <0.001  64.1  35.9  4042 (79.9)  0.99      No  62.7  37.3  705 (18.3)    64.1  35.9  1016 (20.1)    Self-rated health      Good  71.4  28.6  3022 (79.2)  0.78  63.7  36.3  3876 (76.8)  0.23      Fair  72.7  27.3  677 (17.7)    66.0  34.0  1057 (21.0)        Poor  72.7  27.3  117 (3.1)    59.8  40.2  112 (2.2)    Daily smoking      Yes  76.8  23.2  699 (18.2)  <0.01  71.3  28.7  767 (15.3)  <0.001      No  70.5  29.5  3143 (81.8)    62.7  37.3  4244 (84.7)    Daily snus use      Yes  77.8  22.2  499 (13.7)  <0.01  46.3  23.7  842 (16.7)  <0.001      No  70.7  29.3  3149 (86.3)    61.6  38.4  4187 (83.3)    Daily tobacco use      Yes  77.0  23.0  1144 (29.7)  <0.001  73.3  26.7  1522 (30.1)  <0.001      No  69.4  30.6  2711 (70.3)    60.1  39.9  3539 (69.9)    Insufficient fruit intake      Yes  75.0  25.0  1594 (41.6)  <0.001  68.2  31.8  2330 (47.2)  <0.001      No  69.3  30.7  2241 (58.4)    60.7  39.3  2603 (52.8)    Insufficient physical activity      Yes  74.0  26.0  2289 (60.0)  <0.001  65.7  34.3  3206 (63.7)  <0.01      No  68.4  31.6  1524 (40.0)    61.2  38.8  1830 (36.3)      2002 Sub-cohort  2010 Sub-cohort  Alcohol pattern  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  UWAC, n (%)  2764 (71.7)  1092 (28.3)  3856            HOAC, n (%)          3261 (64.1)  1829 (35.9)  5090    Characteristics      Age, m (sd)  49.0 (12.8)  43.8 (15.4)  47.5 (13.8)  <0.01  48.1 (15.7)  46.6 (16.8)  47.5 (16.1)  <0.01  Sex      Female  70.3  29.7  2140 (55.5)  <0.05  55.2  44.8  2021 (39.7)  <0.001      Male  73.4  26.6  1716 (44.5)    69.9  30.1  3069 (60.3)    Education      Low  70.7  29.3  492 (13.0)  0.59  68.1  31.9  865 (17.1)  <0.01      Intermediate  72.7  27.3  1588 (41.8)    65.1  34.9  2145 (42.5)        High  71.4  28.6  1718 (45.2)    61.3  38.7  2038 (40.4)    Employment      Yes  71.4  28.6  3109 (82.0)  0.74  64.4  35.6  3878 (77.3)  0.45      No  72.1  27.9  684 (18.0)    63.1  36.9  1140 (22.7)    Country of birth      Sweden  71.6  28.4  3527 (91.5)  0.32  64.5  35.5  4497 (88.4)  <0.01      Nordic  76.3  23.7  186 (4.8)    69.0  31.0  226 (4.4)        Europe  66.7  33.3  99 (2.6)    60.1  39.9  178 (3.5)        Outside of Europe  68.2  31.8  44 (1.1)    51.9  48.1  189 (3.7)    Cohabitation      Yes  73.8  26.2  3140 (81.7)  <0.001  64.1  35.9  4042 (79.9)  0.99      No  62.7  37.3  705 (18.3)    64.1  35.9  1016 (20.1)    Self-rated health      Good  71.4  28.6  3022 (79.2)  0.78  63.7  36.3  3876 (76.8)  0.23      Fair  72.7  27.3  677 (17.7)    66.0  34.0  1057 (21.0)        Poor  72.7  27.3  117 (3.1)    59.8  40.2  112 (2.2)    Daily smoking      Yes  76.8  23.2  699 (18.2)  <0.01  71.3  28.7  767 (15.3)  <0.001      No  70.5  29.5  3143 (81.8)    62.7  37.3  4244 (84.7)    Daily snus use      Yes  77.8  22.2  499 (13.7)  <0.01  46.3  23.7  842 (16.7)  <0.001      No  70.7  29.3  3149 (86.3)    61.6  38.4  4187 (83.3)    Daily tobacco use      Yes  77.0  23.0  1144 (29.7)  <0.001  73.3  26.7  1522 (30.1)  <0.001      No  69.4  30.6  2711 (70.3)    60.1  39.9  3539 (69.9)    Insufficient fruit intake      Yes  75.0  25.0  1594 (41.6)  <0.001  68.2  31.8  2330 (47.2)  <0.001      No  69.3  30.7  2241 (58.4)    60.7  39.3  2603 (52.8)    Insufficient physical activity      Yes  74.0  26.0  2289 (60.0)  <0.001  65.7  34.3  3206 (63.7)  <0.01      No  68.4  31.6  1524 (40.0)    61.2  38.8  1830 (36.3)    Table 1 Characteristics of participants with hazardous alcohol use at baseline who were followed until 2014, using two definitions of hazardous alcohol use, UWAC and HOAC   2002 Sub-cohort  2010 Sub-cohort  Alcohol pattern  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  UWAC, n (%)  2764 (71.7)  1092 (28.3)  3856            HOAC, n (%)          3261 (64.1)  1829 (35.9)  5090    Characteristics      Age, m (sd)  49.0 (12.8)  43.8 (15.4)  47.5 (13.8)  <0.01  48.1 (15.7)  46.6 (16.8)  47.5 (16.1)  <0.01  Sex      Female  70.3  29.7  2140 (55.5)  <0.05  55.2  44.8  2021 (39.7)  <0.001      Male  73.4  26.6  1716 (44.5)    69.9  30.1  3069 (60.3)    Education      Low  70.7  29.3  492 (13.0)  0.59  68.1  31.9  865 (17.1)  <0.01      Intermediate  72.7  27.3  1588 (41.8)    65.1  34.9  2145 (42.5)        High  71.4  28.6  1718 (45.2)    61.3  38.7  2038 (40.4)    Employment      Yes  71.4  28.6  3109 (82.0)  0.74  64.4  35.6  3878 (77.3)  0.45      No  72.1  27.9  684 (18.0)    63.1  36.9  1140 (22.7)    Country of birth      Sweden  71.6  28.4  3527 (91.5)  0.32  64.5  35.5  4497 (88.4)  <0.01      Nordic  76.3  23.7  186 (4.8)    69.0  31.0  226 (4.4)        Europe  66.7  33.3  99 (2.6)    60.1  39.9  178 (3.5)        Outside of Europe  68.2  31.8  44 (1.1)    51.9  48.1  189 (3.7)    Cohabitation      Yes  73.8  26.2  3140 (81.7)  <0.001  64.1  35.9  4042 (79.9)  0.99      No  62.7  37.3  705 (18.3)    64.1  35.9  1016 (20.1)    Self-rated health      Good  71.4  28.6  3022 (79.2)  0.78  63.7  36.3  3876 (76.8)  0.23      Fair  72.7  27.3  677 (17.7)    66.0  34.0  1057 (21.0)        Poor  72.7  27.3  117 (3.1)    59.8  40.2  112 (2.2)    Daily smoking      Yes  76.8  23.2  699 (18.2)  <0.01  71.3  28.7  767 (15.3)  <0.001      No  70.5  29.5  3143 (81.8)    62.7  37.3  4244 (84.7)    Daily snus use      Yes  77.8  22.2  499 (13.7)  <0.01  46.3  23.7  842 (16.7)  <0.001      No  70.7  29.3  3149 (86.3)    61.6  38.4  4187 (83.3)    Daily tobacco use      Yes  77.0  23.0  1144 (29.7)  <0.001  73.3  26.7  1522 (30.1)  <0.001      No  69.4  30.6  2711 (70.3)    60.1  39.9  3539 (69.9)    Insufficient fruit intake      Yes  75.0  25.0  1594 (41.6)  <0.001  68.2  31.8  2330 (47.2)  <0.001      No  69.3  30.7  2241 (58.4)    60.7  39.3  2603 (52.8)    Insufficient physical activity      Yes  74.0  26.0  2289 (60.0)  <0.001  65.7  34.3  3206 (63.7)  <0.01      No  68.4  31.6  1524 (40.0)    61.2  38.8  1830 (36.3)      2002 Sub-cohort  2010 Sub-cohort  Alcohol pattern  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  UWAC, n (%)  2764 (71.7)  1092 (28.3)  3856            HOAC, n (%)          3261 (64.1)  1829 (35.9)  5090    Characteristics      Age, m (sd)  49.0 (12.8)  43.8 (15.4)  47.5 (13.8)  <0.01  48.1 (15.7)  46.6 (16.8)  47.5 (16.1)  <0.01  Sex      Female  70.3  29.7  2140 (55.5)  <0.05  55.2  44.8  2021 (39.7)  <0.001      Male  73.4  26.6  1716 (44.5)    69.9  30.1  3069 (60.3)    Education      Low  70.7  29.3  492 (13.0)  0.59  68.1  31.9  865 (17.1)  <0.01      Intermediate  72.7  27.3  1588 (41.8)    65.1  34.9  2145 (42.5)        High  71.4  28.6  1718 (45.2)    61.3  38.7  2038 (40.4)    Employment      Yes  71.4  28.6  3109 (82.0)  0.74  64.4  35.6  3878 (77.3)  0.45      No  72.1  27.9  684 (18.0)    63.1  36.9  1140 (22.7)    Country of birth      Sweden  71.6  28.4  3527 (91.5)  0.32  64.5  35.5  4497 (88.4)  <0.01      Nordic  76.3  23.7  186 (4.8)    69.0  31.0  226 (4.4)        Europe  66.7  33.3  99 (2.6)    60.1  39.9  178 (3.5)        Outside of Europe  68.2  31.8  44 (1.1)    51.9  48.1  189 (3.7)    Cohabitation      Yes  73.8  26.2  3140 (81.7)  <0.001  64.1  35.9  4042 (79.9)  0.99      No  62.7  37.3  705 (18.3)    64.1  35.9  1016 (20.1)    Self-rated health      Good  71.4  28.6  3022 (79.2)  0.78  63.7  36.3  3876 (76.8)  0.23      Fair  72.7  27.3  677 (17.7)    66.0  34.0  1057 (21.0)        Poor  72.7  27.3  117 (3.1)    59.8  40.2  112 (2.2)    Daily smoking      Yes  76.8  23.2  699 (18.2)  <0.01  71.3  28.7  767 (15.3)  <0.001      No  70.5  29.5  3143 (81.8)    62.7  37.3  4244 (84.7)    Daily snus use      Yes  77.8  22.2  499 (13.7)  <0.01  46.3  23.7  842 (16.7)  <0.001      No  70.7  29.3  3149 (86.3)    61.6  38.4  4187 (83.3)    Daily tobacco use      Yes  77.0  23.0  1144 (29.7)  <0.001  73.3  26.7  1522 (30.1)  <0.001      No  69.4  30.6  2711 (70.3)    60.1  39.9  3539 (69.9)    Insufficient fruit intake      Yes  75.0  25.0  1594 (41.6)  <0.001  68.2  31.8  2330 (47.2)  <0.001      No  69.3  30.7  2241 (58.4)    60.7  39.3  2603 (52.8)    Insufficient physical activity      Yes  74.0  26.0  2289 (60.0)  <0.001  65.7  34.3  3206 (63.7)  <0.01      No  68.4  31.6  1524 (40.0)    61.2  38.8  1830 (36.3)    Age and sex were associated with quitting hazardous use of alcohol defined by UWAC or HOAC. Cohabitation was associated with quitting hazardous UWAC while education was associated with quitting hazardous HOAC. Employment and self-rated health were not significantly associated with any outcome measure. Table 2 shows the association between health behaviours at baseline and quitting hazardous alcohol use at follow-up. A beneficial profile in each of the four health behaviours was significantly associated with quitting hazardous alcohol use at follow-up, both before and after adjustment for demographic factors. Table 2 ORs and 95% CIs of quitting hazardous alcohol use sustained at two time points (2002 sub-cohort) and of quitting hazardous alcohol use (2010 sub-cohort) according to health behaviours at baseline   2002 Sub-cohort  2010 Sub-cohort    Unadjusted  Model 1 adjusted for age, sex, cohabitation  Unadjusted  Model 1 adjusted for age, sex, education  Baseline health behaviour  OR  CI  OR  CI  OR  CI  OR  CI  Non-Smoking vs. daily smoking  1.39  1.15–1.68  1.44  1.19–1.76  1.48  1.25–1.75  1.58  1.33–1.89  No-Snus use vs. daily snus use  1.45  1.16–1.81  1.70  1.34–2.16  2.00  1.68–2.37  1.75  1.46–2.08  Non-Tobacco use vs. daily tobacco use  1.48  1.26–1.73  1.63  1.38–1.93  1.83  1.60–2.08  1.74  1.52–1.99  Sufficient vs. insufficient fruit intake  1.33  1.15–1.54  1.59  1.36–1.85  1.39  1.23–1.56  1.23  1.09–1.39  Sufficient vs. insufficient physical activity  1.31  1.13–1.51  1.19  1.03–1.38  1.21  1.08–1.37  1.21  1.07–1.37    2002 Sub-cohort  2010 Sub-cohort    Unadjusted  Model 1 adjusted for age, sex, cohabitation  Unadjusted  Model 1 adjusted for age, sex, education  Baseline health behaviour  OR  CI  OR  CI  OR  CI  OR  CI  Non-Smoking vs. daily smoking  1.39  1.15–1.68  1.44  1.19–1.76  1.48  1.25–1.75  1.58  1.33–1.89  No-Snus use vs. daily snus use  1.45  1.16–1.81  1.70  1.34–2.16  2.00  1.68–2.37  1.75  1.46–2.08  Non-Tobacco use vs. daily tobacco use  1.48  1.26–1.73  1.63  1.38–1.93  1.83  1.60–2.08  1.74  1.52–1.99  Sufficient vs. insufficient fruit intake  1.33  1.15–1.54  1.59  1.36–1.85  1.39  1.23–1.56  1.23  1.09–1.39  Sufficient vs. insufficient physical activity  1.31  1.13–1.51  1.19  1.03–1.38  1.21  1.08–1.37  1.21  1.07–1.37  Table 2 ORs and 95% CIs of quitting hazardous alcohol use sustained at two time points (2002 sub-cohort) and of quitting hazardous alcohol use (2010 sub-cohort) according to health behaviours at baseline   2002 Sub-cohort  2010 Sub-cohort    Unadjusted  Model 1 adjusted for age, sex, cohabitation  Unadjusted  Model 1 adjusted for age, sex, education  Baseline health behaviour  OR  CI  OR  CI  OR  CI  OR  CI  Non-Smoking vs. daily smoking  1.39  1.15–1.68  1.44  1.19–1.76  1.48  1.25–1.75  1.58  1.33–1.89  No-Snus use vs. daily snus use  1.45  1.16–1.81  1.70  1.34–2.16  2.00  1.68–2.37  1.75  1.46–2.08  Non-Tobacco use vs. daily tobacco use  1.48  1.26–1.73  1.63  1.38–1.93  1.83  1.60–2.08  1.74  1.52–1.99  Sufficient vs. insufficient fruit intake  1.33  1.15–1.54  1.59  1.36–1.85  1.39  1.23–1.56  1.23  1.09–1.39  Sufficient vs. insufficient physical activity  1.31  1.13–1.51  1.19  1.03–1.38  1.21  1.08–1.37  1.21  1.07–1.37    2002 Sub-cohort  2010 Sub-cohort    Unadjusted  Model 1 adjusted for age, sex, cohabitation  Unadjusted  Model 1 adjusted for age, sex, education  Baseline health behaviour  OR  CI  OR  CI  OR  CI  OR  CI  Non-Smoking vs. daily smoking  1.39  1.15–1.68  1.44  1.19–1.76  1.48  1.25–1.75  1.58  1.33–1.89  No-Snus use vs. daily snus use  1.45  1.16–1.81  1.70  1.34–2.16  2.00  1.68–2.37  1.75  1.46–2.08  Non-Tobacco use vs. daily tobacco use  1.48  1.26–1.73  1.63  1.38–1.93  1.83  1.60–2.08  1.74  1.52–1.99  Sufficient vs. insufficient fruit intake  1.33  1.15–1.54  1.59  1.36–1.85  1.39  1.23–1.56  1.23  1.09–1.39  Sufficient vs. insufficient physical activity  1.31  1.13–1.51  1.19  1.03–1.38  1.21  1.08–1.37  1.21  1.07–1.37  Table 3 displays the association between changes from unfavourable to favourable health behaviours and quitting hazardous alcohol use. Favourable changes in health behaviours were associated with quitting hazardous alcohol use, as was the hypothesized direction, albeit statistical significance was only attained for increased consumption of fruit and for quitting tobacco (2010 sub-cohort). The results were practically unchanged after adjustment for potential confounders. Table 3 Characteristics of participants with hazardous alcohol use and unfavourable health behaviours at baseline   2002 Sub-cohort  2010 Sub-cohort    Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, cohabitation  Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, education  Health behaviour  n (%)  n (%)  n  OR  CI  OR  CI  n (%)  n (%)  n  OR  CI  OR  CI   Daily smoking  Continued  315 (77.4)  92 (22.6)  407  1.00    1.00    363 (72.9)  135 (27.1)  498  1.00    1.00    Quit  219 (75.8)  70 (24.2)  289  1.09  0.76–1.56  1.00  0.69–1.44  179 (68.6)  82 (31.4)  261  1.23  0.88–1.71  1.21  0.86–1.70   Daily snus use  Continued  269 (78.2)  75 (21.8)  344  1.00    1.00    528 (77.4)  154 (22.6)  682  1.00    1.00    Quit  114 (76.0)  36 (24.0)  150  1.13  0.72–1.78  1.09  0.68–1.73  110 (71.4)  44 (28.6)  154  1.37  0.93–2.03  1.27  0.85–1.89   Daily tobacco use  Continued  605 (78.0)  171 (22.0)  776  1.00    1.00    874 (75.0)  292 (25.0)  1166  1.00    1.00    Quit  276 (75.0)  92 (25.0)  368  1.18  0.88–1.58  1.16  0.86–1.56  237 (67.9)  112 (32.1)  349  1.41  1.09–1.84  1.31  1.01–1.71   Insufficient fruit intake  Continued  748 (77.1)  222 (22.9)  970  1.00    1.00    1218 (71.4)  487 (28.6)  1705  1.00    1.00    Favourable change  410 (70.9)  168 (29.1)  578  1.38  1.09–1.74  1.30  1.02–1.66  319 (59.0)  222 (41.0)  541  1.74  1.42–2.12  1.64  1.33–2.01   Insufficient physical activity  Continued  1336 (74.5)  457 (25.5)  1793  1.00    1.00    1626 (65.7)  850 (34.3)  2476  1.00    1.00    Favourable change  342 (71.7)  135 (28.3)  477  1.15  0.92–1.44  1.10  0.88–1.39  448 (66.2)  229 (33.8)  667  0.98  0.82–1.17  0.98  0.82–1.18    2002 Sub-cohort  2010 Sub-cohort    Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, cohabitation  Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, education  Health behaviour  n (%)  n (%)  n  OR  CI  OR  CI  n (%)  n (%)  n  OR  CI  OR  CI   Daily smoking  Continued  315 (77.4)  92 (22.6)  407  1.00    1.00    363 (72.9)  135 (27.1)  498  1.00    1.00    Quit  219 (75.8)  70 (24.2)  289  1.09  0.76–1.56  1.00  0.69–1.44  179 (68.6)  82 (31.4)  261  1.23  0.88–1.71  1.21  0.86–1.70   Daily snus use  Continued  269 (78.2)  75 (21.8)  344  1.00    1.00    528 (77.4)  154 (22.6)  682  1.00    1.00    Quit  114 (76.0)  36 (24.0)  150  1.13  0.72–1.78  1.09  0.68–1.73  110 (71.4)  44 (28.6)  154  1.37  0.93–2.03  1.27  0.85–1.89   Daily tobacco use  Continued  605 (78.0)  171 (22.0)  776  1.00    1.00    874 (75.0)  292 (25.0)  1166  1.00    1.00    Quit  276 (75.0)  92 (25.0)  368  1.18  0.88–1.58  1.16  0.86–1.56  237 (67.9)  112 (32.1)  349  1.41  1.09–1.84  1.31  1.01–1.71   Insufficient fruit intake  Continued  748 (77.1)  222 (22.9)  970  1.00    1.00    1218 (71.4)  487 (28.6)  1705  1.00    1.00    Favourable change  410 (70.9)  168 (29.1)  578  1.38  1.09–1.74  1.30  1.02–1.66  319 (59.0)  222 (41.0)  541  1.74  1.42–2.12  1.64  1.33–2.01   Insufficient physical activity  Continued  1336 (74.5)  457 (25.5)  1793  1.00    1.00    1626 (65.7)  850 (34.3)  2476  1.00    1.00    Favourable change  342 (71.7)  135 (28.3)  477  1.15  0.92–1.44  1.10  0.88–1.39  448 (66.2)  229 (33.8)  667  0.98  0.82–1.17  0.98  0.82–1.18  ORs and 95% CIs of quitting hazardous alcohol use according to shift in health behaviours between baseline and follow-up. Table 3 Characteristics of participants with hazardous alcohol use and unfavourable health behaviours at baseline   2002 Sub-cohort  2010 Sub-cohort    Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, cohabitation  Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, education  Health behaviour  n (%)  n (%)  n  OR  CI  OR  CI  n (%)  n (%)  n  OR  CI  OR  CI   Daily smoking  Continued  315 (77.4)  92 (22.6)  407  1.00    1.00    363 (72.9)  135 (27.1)  498  1.00    1.00    Quit  219 (75.8)  70 (24.2)  289  1.09  0.76–1.56  1.00  0.69–1.44  179 (68.6)  82 (31.4)  261  1.23  0.88–1.71  1.21  0.86–1.70   Daily snus use  Continued  269 (78.2)  75 (21.8)  344  1.00    1.00    528 (77.4)  154 (22.6)  682  1.00    1.00    Quit  114 (76.0)  36 (24.0)  150  1.13  0.72–1.78  1.09  0.68–1.73  110 (71.4)  44 (28.6)  154  1.37  0.93–2.03  1.27  0.85–1.89   Daily tobacco use  Continued  605 (78.0)  171 (22.0)  776  1.00    1.00    874 (75.0)  292 (25.0)  1166  1.00    1.00    Quit  276 (75.0)  92 (25.0)  368  1.18  0.88–1.58  1.16  0.86–1.56  237 (67.9)  112 (32.1)  349  1.41  1.09–1.84  1.31  1.01–1.71   Insufficient fruit intake  Continued  748 (77.1)  222 (22.9)  970  1.00    1.00    1218 (71.4)  487 (28.6)  1705  1.00    1.00    Favourable change  410 (70.9)  168 (29.1)  578  1.38  1.09–1.74  1.30  1.02–1.66  319 (59.0)  222 (41.0)  541  1.74  1.42–2.12  1.64  1.33–2.01   Insufficient physical activity  Continued  1336 (74.5)  457 (25.5)  1793  1.00    1.00    1626 (65.7)  850 (34.3)  2476  1.00    1.00    Favourable change  342 (71.7)  135 (28.3)  477  1.15  0.92–1.44  1.10  0.88–1.39  448 (66.2)  229 (33.8)  667  0.98  0.82–1.17  0.98  0.82–1.18    2002 Sub-cohort  2010 Sub-cohort    Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, cohabitation  Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, education  Health behaviour  n (%)  n (%)  n  OR  CI  OR  CI  n (%)  n (%)  n  OR  CI  OR  CI   Daily smoking  Continued  315 (77.4)  92 (22.6)  407  1.00    1.00    363 (72.9)  135 (27.1)  498  1.00    1.00    Quit  219 (75.8)  70 (24.2)  289  1.09  0.76–1.56  1.00  0.69–1.44  179 (68.6)  82 (31.4)  261  1.23  0.88–1.71  1.21  0.86–1.70   Daily snus use  Continued  269 (78.2)  75 (21.8)  344  1.00    1.00    528 (77.4)  154 (22.6)  682  1.00    1.00    Quit  114 (76.0)  36 (24.0)  150  1.13  0.72–1.78  1.09  0.68–1.73  110 (71.4)  44 (28.6)  154  1.37  0.93–2.03  1.27  0.85–1.89   Daily tobacco use  Continued  605 (78.0)  171 (22.0)  776  1.00    1.00    874 (75.0)  292 (25.0)  1166  1.00    1.00    Quit  276 (75.0)  92 (25.0)  368  1.18  0.88–1.58  1.16  0.86–1.56  237 (67.9)  112 (32.1)  349  1.41  1.09–1.84  1.31  1.01–1.71   Insufficient fruit intake  Continued  748 (77.1)  222 (22.9)  970  1.00    1.00    1218 (71.4)  487 (28.6)  1705  1.00    1.00    Favourable change  410 (70.9)  168 (29.1)  578  1.38  1.09–1.74  1.30  1.02–1.66  319 (59.0)  222 (41.0)  541  1.74  1.42–2.12  1.64  1.33–2.01   Insufficient physical activity  Continued  1336 (74.5)  457 (25.5)  1793  1.00    1.00    1626 (65.7)  850 (34.3)  2476  1.00    1.00    Favourable change  342 (71.7)  135 (28.3)  477  1.15  0.92–1.44  1.10  0.88–1.39  448 (66.2)  229 (33.8)  667  0.98  0.82–1.17  0.98  0.82–1.18  ORs and 95% CIs of quitting hazardous alcohol use according to shift in health behaviours between baseline and follow-up. Discussion In this longitudinal study with a relatively long follow-up, favourable health behaviours at baseline were associated with quitting hazardous alcohol use. These associations were robust to any definition of hazardous alcohol use as well as to the length of follow-up. In addition, sociodemographic characteristics and self-rated health of the hazardous drinkers did not appear to explain the associations. Moreover, favourable changes in dietary habits seemed to predict the transition from hazardous to non-hazardous alcohol use. The same pattern was discernible for quitting daily tobacco use, although the limited sample sizes hampered the precision of the results. The association between use of tobacco and quitting the hazardous use of alcohol is in line with previous findings of decreased alcohol use or abstention following smoking cessation.10,24 Smoking has also been related to increased alcohol use among Finnish adults in a prospective longitudinal study.25 Novel from the current study is the finding that use of the Swedish smokeless tobacco, i.e. snus, seems to be associated with hazardous alcohol use to the same extent as smoking. This is in line with previous cross-sectional analyzes of the SPHC, indicating average high alcohol consumption among snus users.26 Further, snus use was associated with an increased risk of alcohol dependence in a large prospective cohort study.27 Therefore, it seems likely that hazardous alcohol users with the concurrent daily use of tobacco constitute a risk group for the maintenance of the alcohol drinking profile. Daily fruit intake as an indicator of a healthy eating pattern was the only behaviour consistently associated with quitting hazardous alcohol use. Cross-sectional studies have suggested a decline in total dietary quality and fruit intake following increased alcohol consumption.28,29 In a Finnish study, being a male moderate drinker was associated with low intake of fruits while being a heavy drinker was associated with overall poor dietary quality.30 To our knowledge, change of diet and associated change of hazardous alcohol use have not been studied in longitudinal samples. Considering fruit intake as the only proxy for dietary habits is certainly a limitation of this study, but it sheds light on an association that warrants further investigation. At present, findings on the relationship between physical activity and alcohol use are contradictory and rest mainly on cross-sectional studies. Some studies found an association between increased alcohol consumption and increased physical activity,31,32 while others found a curvilinear pattern with increased physical activity among moderate drinkers33 or associations of both sedentary behaviour and high physical activity levels with higher levels of drinking.34 Despite the relatively large amount of publications, we were not able to identify studies relating physical activity to spontaneously quitting the hazardous use of alcohol. DeRuiter et al. found the change in physical activity and change in alcohol consumption not to be associated,35 but the sample in this study was not restricted to hazardous alcohol users. Our results on physical activity are somewhat conflicting. On the one side, we found that physically active individuals at baseline were more likely to quit hazardous alcohol use, although this association was weaker than for other health behaviours. On the other side, changes to favourable physical activity patterns suggested rather a negative impact on the change of hazardous alcohol use. Potential explanations include misclassification, selection, reverse causality or chance. Therefore, the relation of physical activity and hazardous alcohol use needs further analyzes that are more refined. In this study, adopting beneficial behaviour concerning diet and quitting daily tobacco use was predictive of quitting the hazardous use of alcohol. This pattern is in line with theories of behavioural change that posit that any positive lifestyle modification may foster new values and norms; raise self-efficacy, i.e. confidence in one‘s abilities to achieve behavioural goals; introduce skill training and resistance training, and finally promote intention to change and motivation enhancement.36 Accordingly, empirical findings support the possibility of favourable results in interventions targeting multiple behaviours. For instance, a meta-analysis reported that adding smoking cessation to alcohol treatment interventions enhanced the long-term effect on sobriety.37 Based on the results of the current study it can be hypothesized that interventions including multiple behavioural modifications may achieve better results even in the subgroup of hazardous drinkers. Additionally, greater impact on public health can be expected from actions targeting multiple-behaviours.38 Because no agreement exists on the sequence of interventions for coexisting health behaviours,12 this topic should be addressed in future studies. This study had several strengths. First, the prospective design enrolling a large population-based sample. Second, the follow-up period was long, extending over more than a decade, thus allowing the study of long-term modification of hazardous alcohol use. Third, the consistency of the results obtained with two different measures of hazardous use renders them more reliable, as these measures reflect different patterns of hazardous alcohol consumption.39 Heavy occasional drinking per se is associated with health risks4 and is usually neglected if subjects are asked to report their average consumption.40 Limitations of the study include a selection of respondents at baseline and attrition at follow-up. In fact, attrition was highest among the most socially disadvantaged groups that also are more likely to present unfavourable health behaviours. This selection may have resulted in a loss of statistical power and an attenuation of the associations under study, since we do not have reasons to assume that the direction of the association between health behaviours and alcohol use would be different among those retained and non-retained at follow-up. In total, this bias might impact the findings primarily by reduced study efficiency and limited generalizability. A second limitation is the different way questions on physical activity and diet were asked in 2002 and 2010. This alteration might have led to different classifications of change in physical activity and diet in the two sub-cohorts. The cut-off used to define sufficient physical activity was higher in 2010 than 2002, i.e. we strived to achieve high specificity for behavioural change at the expense of sensitivity. Again, these methodologic shortcomings possibly led to non-differential misclassification of the predictor and diluted the associations under investigation. However, since the magnitude of the association was similar in both samples, this bias is likely to be of a modest entity if any. A third limitation is the low reliability of self-reports, a common feature of studies including measures of health behaviours. However, questions on alcohol use were constructed according to good methodological practices including reference period, beverage-specific consumption, quantity and frequency and additionally a question of frequency of heavy drinking.39 Fourth, we might have underestimated hazardous use among women by using the common cut-off of five standard glasses on a single occasion for HOAC, while lately, sex-specific measures detecting corresponding blood concentration in men and women suggested a limit of four standard glasses for women.40 Conclusion A high proportion of hazardous alcohol users may spontaneously quit hazardous alcohol consumption, indicating favourable transitional patterns among these drinkers. Favourable health behaviours at baseline and their positive changes over time may predict changes of alcohol patterns in the desirable direction, thus suggesting the possibility to achieve concurrent behavioural modification in several domains. Acknowledgements The authors are grateful to the Centre for Epidemiology and Community Medicine, Stockholm County for making the SPHC data available for this study. Further, they wish to acknowledge Filip Andersson for statistical advice. M.R.G. designed the study, supervised the data analysis and critically revised the manuscript. E.S. participated in the study design, performed the statistical analyzes and drafted the manuscript. Y.F. and M.R. provided technical advice and participated in the interpretation and the discussion of the results. All authors read and approved the final version of the manuscript. Funding This work was partly supported by the Public Health Agency of Sweden (grant numbers 05576-2014-6.2, 03129-2015-6.2, 02332-2016-6.2). Partly by the, Stockholm County Council, CES (salary to M.R.G. and Y.F.) and The Swedish Council for Information on Alcohol and Other Drugs, CAN (salary M.R.) Conflicts of interest: None declared. Key points It is not known to what extent health behaviours predict change into less risky alcohol use among hazardous consumers. Our results suggest that having or achieving a healthy lifestyle predict desirable transitions in hazardous alcohol use. Co-occurring health behaviours may be considered when developing intervention strategies targeting this subgroup of drinkers. In particular, multiple behavioural changes should be addressed for hazardous drinkers hosting several unfavourable health behaviours. References 1 Institute for Health Metrics and Evaluation, University of Washington. GBD Compare Data Visualization [database on the Internet]. 2016. Seattle, WA, Available at: http://vizhub.healthdata.org/gbd-compare (2 June 2017, date last accessed). 2 Folkhälsomyndigheten [Public Health Agency of Sweden] Alkoholvanor—nationella resultat och tidsserier 2015 [Alcohol use—national results and time series 2015]. 2016. Available at: https://www.folkhalsomyndigheten.se/folkhalsorapportering-statistik/statistikdatabaser-och-visualisering/nationella-folkhalsoenkaten/levnadsvanor/alkoholvanor/ (24 March 2016, date last accessed). 3 Jarl J, Johansson P, Eriksson A, et al.   The societal cost of alcohol consumption: an estimation of the economic and human cost including health effects in Sweden, 2002. Eur J Health Econ  2008; 9: 351– 60. Google Scholar CrossRef Search ADS PubMed  4 Rehm J, Baliunas D, Borges GL, et al.   The relation between different dimensions of alcohol consumption and burden of disease: an overview. Addiction  2010; 105: 817– 43. Google Scholar CrossRef Search ADS PubMed  5 Rehm J, Marmet S, Anderson P, et al.   Defining substance use disorders: do we really need more than heavy use? Alcohol Alcohol  2013; 48: 633– 40. Google Scholar CrossRef Search ADS PubMed  6 Klingemann H, Sobell MB, Sobell LC. Continuities and changes in self‐change research. Addiction  2010; 105: 1510– 8. Google Scholar CrossRef Search ADS PubMed  7 Sobell LC, Ellingstad TP, Sobell MB. Natural recovery from alcohol and drug problems: methodological review of the research with suggestions for future directions. Addiction  2000; 95: 749– 64. Google Scholar CrossRef Search ADS PubMed  8 Cunningham JA. Resolving alcohol-related problems with and without treatment: the effects of different problem criteria. J Stud Alcohol  1999; 60: 463– 6. Google Scholar CrossRef Search ADS PubMed  9 Blomqvist J, Cunningham J, Wallander L, Collin L. Att förbättra sina dryckesvanor: om olika mönster för förändring och om vad vården betyder. [Improving One's Drinking Habits–on different patterns of change and on the role of alcohol treatment]. Stockholm: Stockholm University, Faculty of Social Sciences Centre for Social Research on Alcohol and Drugs; 2007. Report No.: 978-91-85707-04-01. 10 Karlamangla A, Zhou K, Reuben D, et al.   Longitudinal trajectories of heavy drinking in adults in the United States of America. Addiction  2006; 101: 91– 9. Google Scholar CrossRef Search ADS PubMed  11 Moore AA, Gould R, Reuben DB, et al.   Longitudinal patterns and predictors of alcohol consumption in the United States. Am J Public Health  2005; 95: 458– 65. Google Scholar CrossRef Search ADS PubMed  12 Meader N, King K, Moe-Byrne T, et al.   A systematic review on the clustering and co-occurrence of multiple risk behaviours. BMC Public Health  2016; 16: 657. Google Scholar CrossRef Search ADS PubMed  13 de Vries H, van 't Riet J, Spigt M, et al.   Clusters of lifestyle behaviors: results from the Dutch SMILE study. Prev Med  2008; 46: 203– 8. Google Scholar CrossRef Search ADS PubMed  14 Khaw KT, Wareham N, Bingham S, et al.   Combined impact of health behaviours and mortality in men and women: the EPIC-Norfolk prospective population study. PLoS Med  2008; 5: e70. Google Scholar CrossRef Search ADS   15 Svensson AC, Fredlund P, Laflamme L, et al.   Cohort profile: the Stockholm public health cohort. Int J Epidemiol  2013; 42: 1263– 72. Google Scholar CrossRef Search ADS PubMed  16 Romelsjö A, Leifman H, Nyström S. A comparative study of two methods for the measurement of alcohol consumption in the general population. Int J Epidemiol  1995; 24: 929– 36. Google Scholar CrossRef Search ADS PubMed  17 Leifman H. Alkohol och hälsa: en kunskapsöversikt om alkoholens positiva och negativa effekter på vår hälsa . [Alcohol and health a review]. Stockholm: Statens folkhälsoinstitut [Public Health Agency of Sweden], 2005. 18 Naimi TS, Nelson DE, Brewer RD. The intensity of binge alcohol consumption among U.S. adults. Am J Prev Med  2010; 38: 201– 7. Google Scholar CrossRef Search ADS PubMed  19 Connor Gorber S, Schofield-Hurwitz S, Hardt J, et al.   The accuracy of self-reported smoking: a systematic review of the relationship between self-reported and cotinine-assessed smoking status. Nicotine Tob Res  2009; 11: 12– 24. Google Scholar CrossRef Search ADS PubMed  20 Sepp H, Ekelund U, Becker W. Enkätfrågor om kost och fysisk aktivitet bland vuxna—underlag till urval av frågor i befolkningsinriktade enkäter [Survey questions for diet and physical activity—basis for the selection of questions in populationsbased surveys]: Swedish Food Agency, 2004. Report No.: 21. 21 Orsini N, Bellocco R, Bottai M, et al.   Reproducibility of the past year and historical self-administered total physical activity questionnaire among older women. Eur J Epidemiol  2007; 22: 363– 8. Google Scholar CrossRef Search ADS PubMed  22 World health organization (WHO). Global Recommendations on Physical Activity for Health . Geneva, 2010, Available at: http://www.who.int/dietphysicalactivity/publications/9789241599979/en/ (7 July 2017, date last accessed). 23 Burström B, Fredlund P. Self rated health: is it as good a predictor of subsequent mortality among adults in lower as well as in higher social classes? J Epidemiol Commun Health  2001; 55: 836– 40. Google Scholar CrossRef Search ADS   24 Dawson DA, Goldstein RB, Grant BF. Prospective correlates of drinking cessation: variation across the life-course. Addiction  2013; 108: 712– 22. Google Scholar CrossRef Search ADS PubMed  25 Laaksonen M, Luoto R, Helakorpi S, Uutela A. Associations between health-related behaviors: a 7-year follow-up of adults. Prev Med  2002; 34: 162– 70. Google Scholar CrossRef Search ADS PubMed  26 Engström K, Magnusson C, Galanti MR. Socio-demographic, lifestyle and health characteristics among snus users and dual tobacco users in Stockholm County, Sweden. BMC Public Health  2010; 10: 619. Google Scholar CrossRef Search ADS PubMed  27 Norberg M, Malmberg G, Ng N, Broström G. Use of moist smokeless tobacco (snus) and the risk of development of alcohol dependence: a cohort study in a middle-aged population in Sweden. Drug Alcohol Depend  2015; 149: 151– 7. Google Scholar CrossRef Search ADS PubMed  28 Breslow RA, Guenther PM, Juan W, Graubard BI. Alcoholic beverage consumption, nutrient intakes, and diet quality in the US adult population, 1999-2006. J Am Diet Assoc  2010; 110: 551– 62. Google Scholar CrossRef Search ADS PubMed  29 Ruidavets JB, Bataille V, Dallongeville J, et al.   Alcohol intake and diet in France, the prominent role of lifestyle. Eur Heart J  2004; 25: 1153– 62. Google Scholar CrossRef Search ADS PubMed  30 Fawehinmi TO, Ilomaki J, Voutilainen S, Kauhanen J. Alcohol consumption and dietary patterns: the FinDrink study. PLoS One  2012; 7: e38607. Google Scholar CrossRef Search ADS PubMed  31 Burger M, Mensink GBM, Bergmann E, Pietrzik K. Characteristics associated with alcohol consumption in Germany. J Stud Alcohol  2003; 64: 262– 9. Google Scholar CrossRef Search ADS PubMed  32 French MT, Popovici I, Maclean JC. Do alcohol consumers exercise more? Findings from a national survey. Am J Health Promot  2009; 24: 2– 10. Google Scholar CrossRef Search ADS PubMed  33 Smothers B, Bertolucci D. Alcohol consumption and health-promoting behavior in a U.S. household sample: leisure-time physical activity. J Stud Alcohol  2001; 62: 467– 76. Google Scholar CrossRef Search ADS PubMed  34 Kopp M, Burtscher M, Kopp-Wilfling P, et al.   Is there a link between physical activity and alcohol use? Subst Use Misuse  2015; 50: 546– 51. Google Scholar CrossRef Search ADS PubMed  35 deRuiter WK, Cairney J, Leatherdale ST, Faulkner GE. A longitudinal examination of the interrelationship of multiple health behaviors. Am J Prev Med  2014; 47: 283– 9. Google Scholar CrossRef Search ADS PubMed  36 Glanz K, Bishop DB. The role of behavioral science theory in development and implementation of public health interventions. Annu Rev Public Health  2010; 31: 399– 418. Google Scholar CrossRef Search ADS PubMed  37 Prochaska JJ, Delucchi K, Hall SM. A meta-analysis of smoking cessation interventions with individuals in substance abuse treatment or recovery. J Consult Clin Psychol  2004; 72: 1144– 56. Google Scholar CrossRef Search ADS PubMed  38 Prochaska JJ, Spring B, Nigg CR. Multiple health behavior change research: an introduction and overview. Prev Med  2008; 46: 181– 8. Google Scholar CrossRef Search ADS PubMed  39 Greenfield TK, Kerr WC. Alcohol measurement methodology in epidemiology: recent advances and opportunities. Addiction  2008; 103: 1082– 99. Google Scholar CrossRef Search ADS PubMed  40 Stahre M, Naimi T, Brewer R, Holt J. Measuring average alcohol consumption: the impact of including binge drinks in quantity-frequency calculations. Addiction  2006; 101: 1711– 8. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The European Journal of Public Health Oxford University Press

Health behaviours as a predictor of quitting hazardous alcohol use in the Stockholm Public Health Cohort

Loading next page...
 
/lp/ou_press/health-behaviours-as-a-predictor-of-quitting-hazardous-alcohol-use-in-rXGnToQMVa
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved.
ISSN
1101-1262
eISSN
1464-360X
D.O.I.
10.1093/eurpub/ckx193
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Adopting healthy behaviours may facilitate the transition from hazardous to non-hazardous use of alcohol, yet, longitudinal studies of health behaviours in relation to the cessation of hazardous alcohol use are rare. We addressed this question using data from a large population-based cohort of adults in Sweden (Stockholm Public Health Cohort). Methods Participants from two sub-cohorts (inception in 2002 and 2010), with follow-up until the year 2014 were included. Health behaviours (tobacco use, diet and physical activity) and alcohol use were self-reported in questionnaire-based surveys. Hazardous alcohol use was defined as either usual weekly consumption (2002 sub-cohort) or heavy occasional alcohol consumption (2010 sub-cohort). Baseline hazardous drinkers with complete data constituted the analytical sample (n = 8946). Logistic regression was used to calculate the Odds Ratios and their 95% confidence intervals of quitting hazardous alcohol use, with tobacco use, diet and physical activity as predictors of change. Results In the 2002 sub-cohort, 28% reported non-hazardous use sustained through two consecutive follow-up points. In the 2010 sub-cohort, 36% of the participants reported non-hazardous use of alcohol at follow-up. Favourable health behaviours at baseline (e.g. no tobacco use, sufficient fruit intake and physical activity) were associated with a 19% to 75% higher of odds quitting hazardous alcohol use. Further, favourable changes in diet and tobacco cessation were associated with non-hazardous alcohol use at follow-up. Conclusions As many as one-third of hazardous alcohol users may quit this drinking pattern in a medium-long term. Holding or achieving a healthy lifestyle may facilitate this transition. Introduction Alcohol use is one of the leading preventable causes of premature death and disease. Worldwide, 2.3 million premature deaths and 3.4% of disability-adjusted life years (DALYs) were attributable to alcohol use in 2015.1 In Sweden, 2.6% of DALYs were due to alcohol,1 and the prevalence of hazardous and harmful alcohol use was estimated to 15% in 2015.2 Furthermore, hazardous alcohol use leads to a substantial economic burden to the society.3 Hazardous alcohol use refers to consumption patterns that increase the risk of adverse health consequences, while harmful use is defined as a behaviour already causing damage to health. Average consumption is one quantitative measure of hazardous and harmful use. Another aspect includes episodes of heavy drinking, which in itself convey important health consequences.4 To date, there is no universal consensus about which thresholds in these measures mark the transition between hazardous and non-hazardous alcohol use, with consequent heterogeneity in this definition between studies. During the last decade’s research has shifted its focus from alcohol dependence to the continuum from low risk to hazardous use.5 Along this pathway, the transition from hazardous and harmful to non-hazardous use is poorly understood. In fact, there is limited knowledge on the extent to which this behaviour naturally recedes over the lifetime.6,7 Spontaneous recovery seems frequent, but the success rate depends on the definition of the alcohol problem, and of recovery, as well as on the timespan.8,9 A better understanding of which factors promote a reduction of alcohol use is needed to design effective interventions, ultimately leading to a reduction of harm at the societal level. Previous studies mainly addressed demographics (e.g. sex, age, education) as predictors for change in hazardous alcohol use and recovery.9–11 A few studies described the co-occurrence of unfavourable health behaviours among hazardous drinkers,12 but whether and how changes in health behaviours are associated with changes in hazardous alcohol use longitudinally is poorly understood, particularly in Sweden. Unfavourable health behaviours tend to cluster,13 leading to elevated risks of comorbidities.14 Thus, intervention strategies targeting associated health behaviours may prove of utmost importance. The aim of this study was 2-fold: to describe cumulative incidence of the transition from hazardous to non-hazardous alcohol use in a population sample of Swedish adults, and to examine whether modifiable health behaviours, i.e. smoking, other tobacco use, physical activity and dietary habits predict quitting the hazardous use of alcohol. We hypothesized that holding favourable health behaviours or changing from unhealthy to healthier behaviours (e.g. smoking cessation, increasing physical activity), predicts the transition from hazardous to low-risk alcohol use. Methods Study design and participants Details on the study design and collection of information in the Stockholm Public Health Cohort (SPHC) have been reported by Svensson et al.15 Briefly, SPHC is a prospective study, which has enrolled sub-cohorts in 2002, 2006 and 2010. Each of the sub-cohorts were followed every 4th year until 2014. The initial samples were randomly selected from the general population of Stockholm County stratified by the municipality, and each included about 50 000 residents 18 years of age or older. The response rate was 62%, 61% and 56% in the sub-cohorts 2002, 2006 and 2010, respectively.15 Participants gave informed consent at inception. Additional consent to national register linkage and participation in longitudinal data collection for the 2002 sub-cohort was obtained in 2006, which led to the effective recruitment of 47% of the baseline sample (n = 23 771). Due to considerable modifications of the questionnaire in 2006, we did not include information from the follow-up or the sub-cohort in this year. Overall, SPHC responders were more likely to be female, over 45 years of age, born in Sweden and to have higher education and income than the average of the population reported in the Stockholm census data.15 We analyzed data from participants recruited in 2002 and 2010 who remained in the study until 2014, and who at baseline reported alcohol consumption above the cut-off for hazardous use. The response rate in 2014 was 64% in the 2002 sub-cohort, and 56% in the 2010 sub-cohort, corresponding to samples of 3856 and 5090 individuals, respectively, see figure 1. Ethical approval was obtained from the Ethical Review Board of Stockholm Region (DNR 2016/749-2). Figure 1 View largeDownload slide Flowchart of the derivation of the study samples, the SPHC 2002–14. UWAC usual weekly alcohol consumption and HOAC Heavy occasional alcohol consumption UWAC and HOAC. ∗Respondents among those who gave consent to register linkage in (total respondents in 2002 = 31 182) Figure 1 View largeDownload slide Flowchart of the derivation of the study samples, the SPHC 2002–14. UWAC usual weekly alcohol consumption and HOAC Heavy occasional alcohol consumption UWAC and HOAC. ∗Respondents among those who gave consent to register linkage in (total respondents in 2002 = 31 182) Measurements Usual weekly alcohol consumption In the 2002 sub-cohort, self-reported usual consumption of alcoholic beverages (spirits, fortified wines, wine, alcopop or cider, medium-strong beer and strong beer) was collected with reference to a standard week during the past year.16 Usual weekly alcohol consumption (UWAC) was measured in centilitres, exemplified as glasses and bottles or cans (2002) and in numbers of standard drinks (2010, 2014); after that, a measure of grams of pure ethanol per week was derived. UWAC was dichotomized into non-hazardous or hazardous use based on commonly used cut-offs in Sweden, i.e. a weekly consumption of >108 gram (>9 standard drinks) for women and >168 gram (>14 standard drinks) for men.17 In this sub-cohort, quitting hazardous alcohol use was defined as ‘reporting consumption below the cut-off in two consecutive follow-up surveys (2010 and 2014)’. Heavy occasional alcohol consumption In the 2010 sub-cohort, heavy occasional alcohol consumption (HOAC) was measured by events of drinking alcohol equivalent to five standard glasses (1 bottle of wine; 5 shots of spirits; 4 cans of strong beer; 6 cans of medium-strong beer), in the past year.18 HOAC was dichotomized to non-hazardous (less than monthly) and hazardous (at least monthly frequency of HOAC). In this sub-cohort, quitting hazardous alcohol use was defined as ‘reporting a frequency of heavy consumption less than monthly in 2014’. Health behaviours Two similar questions were used to measure current daily tobacco use, ‘Do you currently smoke daily or almost daily’ and ‘Do you currently use snus (the Swedish type of moist oral snuff) daily or almost daily?’19 Daily tobacco use was derived from these two separate questions, as any daily use of either cigarettes or snus. Quitting tobacco use was defined as a switch from daily to non-current daily use at follow-up. Usual frequency of fruit consumption during the past year was used to derive an indicator of favourable eating patterns. In the 2002 survey, sufficient consumption was defined by the response alternative ‘nearly daily or more frequently’. Following a modification of the survey question in 2010, sufficient consumption was defined by the alternative ‘6 times per week or more’. Following a modification of the survey question in 2010,20 daily consumption was defined by the alternative ‘6 times per week or more’. A switch from less than daily to daily consumption of fruit at follow-up was categorized as a favourable change. Physical activity was assessed as the usual level of leisure time physical activity during the past year. In the 2002 sub-cohort, sufficient leisure time physical activity was defined as ‘regular physical activity 1–2 times per week at a moderate level’. In the 2010 and 2014 surveys, the question and response alternatives were rephrased to include several domains.21 Therefore, in the 2010 sub-cohort, we chose the response alternative ‘2–3 h of exercise per week’ as the threshold for sufficient physical activity, in line with the current recommendations of a minimum of 150 min per week of moderate to intense physical activity.22 A switch from insufficient to sufficient levels of physical activity at follow-up was categorized as a favourable change. In the 2002 sub-cohort, behavioural changes were assessed in 2010, while in the 2010 sub-cohort they were assessed in 2014. Sociodemographic characteristics Employment and house-hold composition were based on self-reports while age, sex, education and country of birth were based on register data. Employment was derived from the question ‘What is your current employment status?’ Answers were coded as ‘employed’ (employed, self-employed, students and parental leave) or ‘not employed’ (long-term sick leave, age pension, disability pension, unemployed and others). House-hold composition was based on the question ‘Whom do you share a house-hold with?’ Answers were coded as ‘cohabitation’ (living with any other adult or child) or ‘living alone’. Education was grouped into three levels: ‘low’ (compulsory school and vocational training), ‘intermediate’ (secondary school) and ‘high’ (university studies). Self-rated health was assessed by the question ‘In general, how would you rate your health?’ Answers were coded as ‘good health’ (response alternatives good and excellent), ‘fair health’ (fair) and ‘poor health’ (bad or very bad).23 Statistical analysis Univariate analysis was conducted to compare baseline sociodemographic characteristics and health behaviours between hazardous alcohol consumers who did or did not change drinking pattern at follow-up. The departure of the observed from the expected proportions was explored using the chi-square statistic. Binary logistic regression was used to calculate Odds Ratios (OR) and the corresponding 95% Confidence Intervals (CI) of quitting hazardous alcohol use (outcome variable in this study). We conducted two sets of analyzes. First, baseline values of each health behaviour were regressed on the outcome. Second, favourable modifications of health behaviours, e.g. quitting tobacco use were modelled as predictors of quitting hazardous alcohol use. These latter analyzes were restricted to hazardous drinkers with unfavourable health behaviours at baseline. The results are presented with and without adjustments for potential confounders. These were selected based on prior explanatory models of behavioural change (i.e. age, sex, education, employment, cohabitation status and self-rated health at baseline), and were kept in the final model only if significantly associated with the outcome. The conventional significance level was set at P < 0.05. Data analysis was performed using STATA 14.1. Results Hazardous alcohol consumers retained at follow-up, were more likely to be female, born in Sweden, older, have higher education and be non-smokers at baseline than non-responders (data not shown). Also, participants tended to have a slightly lower intake of alcohol at baseline (UWAC) compared to non-participants, while no difference were seen for HOAC. Missing information did not exceed five percent for all variables used in the analysis. Table 1 shows the distribution of characteristics of the two sub-cohorts according to changes in drinking patterns. The probability of sustained change, i.e. record of non-hazardous alcohol use at 8 and 12 years follow-up was 28.3% (sub-cohort 2002). The probability of having quit hazardous use at 4 years follow-up was 35.9% (sub-cohort 2010). Likewise, the point prevalence of non-hazardous use in the 2002 sub-cohort was 36% after 8 years and 46% after 12 years follow-up (data not shown). Table 1 Characteristics of participants with hazardous alcohol use at baseline who were followed until 2014, using two definitions of hazardous alcohol use, UWAC and HOAC   2002 Sub-cohort  2010 Sub-cohort  Alcohol pattern  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  UWAC, n (%)  2764 (71.7)  1092 (28.3)  3856            HOAC, n (%)          3261 (64.1)  1829 (35.9)  5090    Characteristics      Age, m (sd)  49.0 (12.8)  43.8 (15.4)  47.5 (13.8)  <0.01  48.1 (15.7)  46.6 (16.8)  47.5 (16.1)  <0.01  Sex      Female  70.3  29.7  2140 (55.5)  <0.05  55.2  44.8  2021 (39.7)  <0.001      Male  73.4  26.6  1716 (44.5)    69.9  30.1  3069 (60.3)    Education      Low  70.7  29.3  492 (13.0)  0.59  68.1  31.9  865 (17.1)  <0.01      Intermediate  72.7  27.3  1588 (41.8)    65.1  34.9  2145 (42.5)        High  71.4  28.6  1718 (45.2)    61.3  38.7  2038 (40.4)    Employment      Yes  71.4  28.6  3109 (82.0)  0.74  64.4  35.6  3878 (77.3)  0.45      No  72.1  27.9  684 (18.0)    63.1  36.9  1140 (22.7)    Country of birth      Sweden  71.6  28.4  3527 (91.5)  0.32  64.5  35.5  4497 (88.4)  <0.01      Nordic  76.3  23.7  186 (4.8)    69.0  31.0  226 (4.4)        Europe  66.7  33.3  99 (2.6)    60.1  39.9  178 (3.5)        Outside of Europe  68.2  31.8  44 (1.1)    51.9  48.1  189 (3.7)    Cohabitation      Yes  73.8  26.2  3140 (81.7)  <0.001  64.1  35.9  4042 (79.9)  0.99      No  62.7  37.3  705 (18.3)    64.1  35.9  1016 (20.1)    Self-rated health      Good  71.4  28.6  3022 (79.2)  0.78  63.7  36.3  3876 (76.8)  0.23      Fair  72.7  27.3  677 (17.7)    66.0  34.0  1057 (21.0)        Poor  72.7  27.3  117 (3.1)    59.8  40.2  112 (2.2)    Daily smoking      Yes  76.8  23.2  699 (18.2)  <0.01  71.3  28.7  767 (15.3)  <0.001      No  70.5  29.5  3143 (81.8)    62.7  37.3  4244 (84.7)    Daily snus use      Yes  77.8  22.2  499 (13.7)  <0.01  46.3  23.7  842 (16.7)  <0.001      No  70.7  29.3  3149 (86.3)    61.6  38.4  4187 (83.3)    Daily tobacco use      Yes  77.0  23.0  1144 (29.7)  <0.001  73.3  26.7  1522 (30.1)  <0.001      No  69.4  30.6  2711 (70.3)    60.1  39.9  3539 (69.9)    Insufficient fruit intake      Yes  75.0  25.0  1594 (41.6)  <0.001  68.2  31.8  2330 (47.2)  <0.001      No  69.3  30.7  2241 (58.4)    60.7  39.3  2603 (52.8)    Insufficient physical activity      Yes  74.0  26.0  2289 (60.0)  <0.001  65.7  34.3  3206 (63.7)  <0.01      No  68.4  31.6  1524 (40.0)    61.2  38.8  1830 (36.3)      2002 Sub-cohort  2010 Sub-cohort  Alcohol pattern  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  UWAC, n (%)  2764 (71.7)  1092 (28.3)  3856            HOAC, n (%)          3261 (64.1)  1829 (35.9)  5090    Characteristics      Age, m (sd)  49.0 (12.8)  43.8 (15.4)  47.5 (13.8)  <0.01  48.1 (15.7)  46.6 (16.8)  47.5 (16.1)  <0.01  Sex      Female  70.3  29.7  2140 (55.5)  <0.05  55.2  44.8  2021 (39.7)  <0.001      Male  73.4  26.6  1716 (44.5)    69.9  30.1  3069 (60.3)    Education      Low  70.7  29.3  492 (13.0)  0.59  68.1  31.9  865 (17.1)  <0.01      Intermediate  72.7  27.3  1588 (41.8)    65.1  34.9  2145 (42.5)        High  71.4  28.6  1718 (45.2)    61.3  38.7  2038 (40.4)    Employment      Yes  71.4  28.6  3109 (82.0)  0.74  64.4  35.6  3878 (77.3)  0.45      No  72.1  27.9  684 (18.0)    63.1  36.9  1140 (22.7)    Country of birth      Sweden  71.6  28.4  3527 (91.5)  0.32  64.5  35.5  4497 (88.4)  <0.01      Nordic  76.3  23.7  186 (4.8)    69.0  31.0  226 (4.4)        Europe  66.7  33.3  99 (2.6)    60.1  39.9  178 (3.5)        Outside of Europe  68.2  31.8  44 (1.1)    51.9  48.1  189 (3.7)    Cohabitation      Yes  73.8  26.2  3140 (81.7)  <0.001  64.1  35.9  4042 (79.9)  0.99      No  62.7  37.3  705 (18.3)    64.1  35.9  1016 (20.1)    Self-rated health      Good  71.4  28.6  3022 (79.2)  0.78  63.7  36.3  3876 (76.8)  0.23      Fair  72.7  27.3  677 (17.7)    66.0  34.0  1057 (21.0)        Poor  72.7  27.3  117 (3.1)    59.8  40.2  112 (2.2)    Daily smoking      Yes  76.8  23.2  699 (18.2)  <0.01  71.3  28.7  767 (15.3)  <0.001      No  70.5  29.5  3143 (81.8)    62.7  37.3  4244 (84.7)    Daily snus use      Yes  77.8  22.2  499 (13.7)  <0.01  46.3  23.7  842 (16.7)  <0.001      No  70.7  29.3  3149 (86.3)    61.6  38.4  4187 (83.3)    Daily tobacco use      Yes  77.0  23.0  1144 (29.7)  <0.001  73.3  26.7  1522 (30.1)  <0.001      No  69.4  30.6  2711 (70.3)    60.1  39.9  3539 (69.9)    Insufficient fruit intake      Yes  75.0  25.0  1594 (41.6)  <0.001  68.2  31.8  2330 (47.2)  <0.001      No  69.3  30.7  2241 (58.4)    60.7  39.3  2603 (52.8)    Insufficient physical activity      Yes  74.0  26.0  2289 (60.0)  <0.001  65.7  34.3  3206 (63.7)  <0.01      No  68.4  31.6  1524 (40.0)    61.2  38.8  1830 (36.3)    Table 1 Characteristics of participants with hazardous alcohol use at baseline who were followed until 2014, using two definitions of hazardous alcohol use, UWAC and HOAC   2002 Sub-cohort  2010 Sub-cohort  Alcohol pattern  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  UWAC, n (%)  2764 (71.7)  1092 (28.3)  3856            HOAC, n (%)          3261 (64.1)  1829 (35.9)  5090    Characteristics      Age, m (sd)  49.0 (12.8)  43.8 (15.4)  47.5 (13.8)  <0.01  48.1 (15.7)  46.6 (16.8)  47.5 (16.1)  <0.01  Sex      Female  70.3  29.7  2140 (55.5)  <0.05  55.2  44.8  2021 (39.7)  <0.001      Male  73.4  26.6  1716 (44.5)    69.9  30.1  3069 (60.3)    Education      Low  70.7  29.3  492 (13.0)  0.59  68.1  31.9  865 (17.1)  <0.01      Intermediate  72.7  27.3  1588 (41.8)    65.1  34.9  2145 (42.5)        High  71.4  28.6  1718 (45.2)    61.3  38.7  2038 (40.4)    Employment      Yes  71.4  28.6  3109 (82.0)  0.74  64.4  35.6  3878 (77.3)  0.45      No  72.1  27.9  684 (18.0)    63.1  36.9  1140 (22.7)    Country of birth      Sweden  71.6  28.4  3527 (91.5)  0.32  64.5  35.5  4497 (88.4)  <0.01      Nordic  76.3  23.7  186 (4.8)    69.0  31.0  226 (4.4)        Europe  66.7  33.3  99 (2.6)    60.1  39.9  178 (3.5)        Outside of Europe  68.2  31.8  44 (1.1)    51.9  48.1  189 (3.7)    Cohabitation      Yes  73.8  26.2  3140 (81.7)  <0.001  64.1  35.9  4042 (79.9)  0.99      No  62.7  37.3  705 (18.3)    64.1  35.9  1016 (20.1)    Self-rated health      Good  71.4  28.6  3022 (79.2)  0.78  63.7  36.3  3876 (76.8)  0.23      Fair  72.7  27.3  677 (17.7)    66.0  34.0  1057 (21.0)        Poor  72.7  27.3  117 (3.1)    59.8  40.2  112 (2.2)    Daily smoking      Yes  76.8  23.2  699 (18.2)  <0.01  71.3  28.7  767 (15.3)  <0.001      No  70.5  29.5  3143 (81.8)    62.7  37.3  4244 (84.7)    Daily snus use      Yes  77.8  22.2  499 (13.7)  <0.01  46.3  23.7  842 (16.7)  <0.001      No  70.7  29.3  3149 (86.3)    61.6  38.4  4187 (83.3)    Daily tobacco use      Yes  77.0  23.0  1144 (29.7)  <0.001  73.3  26.7  1522 (30.1)  <0.001      No  69.4  30.6  2711 (70.3)    60.1  39.9  3539 (69.9)    Insufficient fruit intake      Yes  75.0  25.0  1594 (41.6)  <0.001  68.2  31.8  2330 (47.2)  <0.001      No  69.3  30.7  2241 (58.4)    60.7  39.3  2603 (52.8)    Insufficient physical activity      Yes  74.0  26.0  2289 (60.0)  <0.001  65.7  34.3  3206 (63.7)  <0.01      No  68.4  31.6  1524 (40.0)    61.2  38.8  1830 (36.3)      2002 Sub-cohort  2010 Sub-cohort  Alcohol pattern  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  Continued hazardous %  Quit hazardous %  Total n (%)  P-value  UWAC, n (%)  2764 (71.7)  1092 (28.3)  3856            HOAC, n (%)          3261 (64.1)  1829 (35.9)  5090    Characteristics      Age, m (sd)  49.0 (12.8)  43.8 (15.4)  47.5 (13.8)  <0.01  48.1 (15.7)  46.6 (16.8)  47.5 (16.1)  <0.01  Sex      Female  70.3  29.7  2140 (55.5)  <0.05  55.2  44.8  2021 (39.7)  <0.001      Male  73.4  26.6  1716 (44.5)    69.9  30.1  3069 (60.3)    Education      Low  70.7  29.3  492 (13.0)  0.59  68.1  31.9  865 (17.1)  <0.01      Intermediate  72.7  27.3  1588 (41.8)    65.1  34.9  2145 (42.5)        High  71.4  28.6  1718 (45.2)    61.3  38.7  2038 (40.4)    Employment      Yes  71.4  28.6  3109 (82.0)  0.74  64.4  35.6  3878 (77.3)  0.45      No  72.1  27.9  684 (18.0)    63.1  36.9  1140 (22.7)    Country of birth      Sweden  71.6  28.4  3527 (91.5)  0.32  64.5  35.5  4497 (88.4)  <0.01      Nordic  76.3  23.7  186 (4.8)    69.0  31.0  226 (4.4)        Europe  66.7  33.3  99 (2.6)    60.1  39.9  178 (3.5)        Outside of Europe  68.2  31.8  44 (1.1)    51.9  48.1  189 (3.7)    Cohabitation      Yes  73.8  26.2  3140 (81.7)  <0.001  64.1  35.9  4042 (79.9)  0.99      No  62.7  37.3  705 (18.3)    64.1  35.9  1016 (20.1)    Self-rated health      Good  71.4  28.6  3022 (79.2)  0.78  63.7  36.3  3876 (76.8)  0.23      Fair  72.7  27.3  677 (17.7)    66.0  34.0  1057 (21.0)        Poor  72.7  27.3  117 (3.1)    59.8  40.2  112 (2.2)    Daily smoking      Yes  76.8  23.2  699 (18.2)  <0.01  71.3  28.7  767 (15.3)  <0.001      No  70.5  29.5  3143 (81.8)    62.7  37.3  4244 (84.7)    Daily snus use      Yes  77.8  22.2  499 (13.7)  <0.01  46.3  23.7  842 (16.7)  <0.001      No  70.7  29.3  3149 (86.3)    61.6  38.4  4187 (83.3)    Daily tobacco use      Yes  77.0  23.0  1144 (29.7)  <0.001  73.3  26.7  1522 (30.1)  <0.001      No  69.4  30.6  2711 (70.3)    60.1  39.9  3539 (69.9)    Insufficient fruit intake      Yes  75.0  25.0  1594 (41.6)  <0.001  68.2  31.8  2330 (47.2)  <0.001      No  69.3  30.7  2241 (58.4)    60.7  39.3  2603 (52.8)    Insufficient physical activity      Yes  74.0  26.0  2289 (60.0)  <0.001  65.7  34.3  3206 (63.7)  <0.01      No  68.4  31.6  1524 (40.0)    61.2  38.8  1830 (36.3)    Age and sex were associated with quitting hazardous use of alcohol defined by UWAC or HOAC. Cohabitation was associated with quitting hazardous UWAC while education was associated with quitting hazardous HOAC. Employment and self-rated health were not significantly associated with any outcome measure. Table 2 shows the association between health behaviours at baseline and quitting hazardous alcohol use at follow-up. A beneficial profile in each of the four health behaviours was significantly associated with quitting hazardous alcohol use at follow-up, both before and after adjustment for demographic factors. Table 2 ORs and 95% CIs of quitting hazardous alcohol use sustained at two time points (2002 sub-cohort) and of quitting hazardous alcohol use (2010 sub-cohort) according to health behaviours at baseline   2002 Sub-cohort  2010 Sub-cohort    Unadjusted  Model 1 adjusted for age, sex, cohabitation  Unadjusted  Model 1 adjusted for age, sex, education  Baseline health behaviour  OR  CI  OR  CI  OR  CI  OR  CI  Non-Smoking vs. daily smoking  1.39  1.15–1.68  1.44  1.19–1.76  1.48  1.25–1.75  1.58  1.33–1.89  No-Snus use vs. daily snus use  1.45  1.16–1.81  1.70  1.34–2.16  2.00  1.68–2.37  1.75  1.46–2.08  Non-Tobacco use vs. daily tobacco use  1.48  1.26–1.73  1.63  1.38–1.93  1.83  1.60–2.08  1.74  1.52–1.99  Sufficient vs. insufficient fruit intake  1.33  1.15–1.54  1.59  1.36–1.85  1.39  1.23–1.56  1.23  1.09–1.39  Sufficient vs. insufficient physical activity  1.31  1.13–1.51  1.19  1.03–1.38  1.21  1.08–1.37  1.21  1.07–1.37    2002 Sub-cohort  2010 Sub-cohort    Unadjusted  Model 1 adjusted for age, sex, cohabitation  Unadjusted  Model 1 adjusted for age, sex, education  Baseline health behaviour  OR  CI  OR  CI  OR  CI  OR  CI  Non-Smoking vs. daily smoking  1.39  1.15–1.68  1.44  1.19–1.76  1.48  1.25–1.75  1.58  1.33–1.89  No-Snus use vs. daily snus use  1.45  1.16–1.81  1.70  1.34–2.16  2.00  1.68–2.37  1.75  1.46–2.08  Non-Tobacco use vs. daily tobacco use  1.48  1.26–1.73  1.63  1.38–1.93  1.83  1.60–2.08  1.74  1.52–1.99  Sufficient vs. insufficient fruit intake  1.33  1.15–1.54  1.59  1.36–1.85  1.39  1.23–1.56  1.23  1.09–1.39  Sufficient vs. insufficient physical activity  1.31  1.13–1.51  1.19  1.03–1.38  1.21  1.08–1.37  1.21  1.07–1.37  Table 2 ORs and 95% CIs of quitting hazardous alcohol use sustained at two time points (2002 sub-cohort) and of quitting hazardous alcohol use (2010 sub-cohort) according to health behaviours at baseline   2002 Sub-cohort  2010 Sub-cohort    Unadjusted  Model 1 adjusted for age, sex, cohabitation  Unadjusted  Model 1 adjusted for age, sex, education  Baseline health behaviour  OR  CI  OR  CI  OR  CI  OR  CI  Non-Smoking vs. daily smoking  1.39  1.15–1.68  1.44  1.19–1.76  1.48  1.25–1.75  1.58  1.33–1.89  No-Snus use vs. daily snus use  1.45  1.16–1.81  1.70  1.34–2.16  2.00  1.68–2.37  1.75  1.46–2.08  Non-Tobacco use vs. daily tobacco use  1.48  1.26–1.73  1.63  1.38–1.93  1.83  1.60–2.08  1.74  1.52–1.99  Sufficient vs. insufficient fruit intake  1.33  1.15–1.54  1.59  1.36–1.85  1.39  1.23–1.56  1.23  1.09–1.39  Sufficient vs. insufficient physical activity  1.31  1.13–1.51  1.19  1.03–1.38  1.21  1.08–1.37  1.21  1.07–1.37    2002 Sub-cohort  2010 Sub-cohort    Unadjusted  Model 1 adjusted for age, sex, cohabitation  Unadjusted  Model 1 adjusted for age, sex, education  Baseline health behaviour  OR  CI  OR  CI  OR  CI  OR  CI  Non-Smoking vs. daily smoking  1.39  1.15–1.68  1.44  1.19–1.76  1.48  1.25–1.75  1.58  1.33–1.89  No-Snus use vs. daily snus use  1.45  1.16–1.81  1.70  1.34–2.16  2.00  1.68–2.37  1.75  1.46–2.08  Non-Tobacco use vs. daily tobacco use  1.48  1.26–1.73  1.63  1.38–1.93  1.83  1.60–2.08  1.74  1.52–1.99  Sufficient vs. insufficient fruit intake  1.33  1.15–1.54  1.59  1.36–1.85  1.39  1.23–1.56  1.23  1.09–1.39  Sufficient vs. insufficient physical activity  1.31  1.13–1.51  1.19  1.03–1.38  1.21  1.08–1.37  1.21  1.07–1.37  Table 3 displays the association between changes from unfavourable to favourable health behaviours and quitting hazardous alcohol use. Favourable changes in health behaviours were associated with quitting hazardous alcohol use, as was the hypothesized direction, albeit statistical significance was only attained for increased consumption of fruit and for quitting tobacco (2010 sub-cohort). The results were practically unchanged after adjustment for potential confounders. Table 3 Characteristics of participants with hazardous alcohol use and unfavourable health behaviours at baseline   2002 Sub-cohort  2010 Sub-cohort    Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, cohabitation  Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, education  Health behaviour  n (%)  n (%)  n  OR  CI  OR  CI  n (%)  n (%)  n  OR  CI  OR  CI   Daily smoking  Continued  315 (77.4)  92 (22.6)  407  1.00    1.00    363 (72.9)  135 (27.1)  498  1.00    1.00    Quit  219 (75.8)  70 (24.2)  289  1.09  0.76–1.56  1.00  0.69–1.44  179 (68.6)  82 (31.4)  261  1.23  0.88–1.71  1.21  0.86–1.70   Daily snus use  Continued  269 (78.2)  75 (21.8)  344  1.00    1.00    528 (77.4)  154 (22.6)  682  1.00    1.00    Quit  114 (76.0)  36 (24.0)  150  1.13  0.72–1.78  1.09  0.68–1.73  110 (71.4)  44 (28.6)  154  1.37  0.93–2.03  1.27  0.85–1.89   Daily tobacco use  Continued  605 (78.0)  171 (22.0)  776  1.00    1.00    874 (75.0)  292 (25.0)  1166  1.00    1.00    Quit  276 (75.0)  92 (25.0)  368  1.18  0.88–1.58  1.16  0.86–1.56  237 (67.9)  112 (32.1)  349  1.41  1.09–1.84  1.31  1.01–1.71   Insufficient fruit intake  Continued  748 (77.1)  222 (22.9)  970  1.00    1.00    1218 (71.4)  487 (28.6)  1705  1.00    1.00    Favourable change  410 (70.9)  168 (29.1)  578  1.38  1.09–1.74  1.30  1.02–1.66  319 (59.0)  222 (41.0)  541  1.74  1.42–2.12  1.64  1.33–2.01   Insufficient physical activity  Continued  1336 (74.5)  457 (25.5)  1793  1.00    1.00    1626 (65.7)  850 (34.3)  2476  1.00    1.00    Favourable change  342 (71.7)  135 (28.3)  477  1.15  0.92–1.44  1.10  0.88–1.39  448 (66.2)  229 (33.8)  667  0.98  0.82–1.17  0.98  0.82–1.18    2002 Sub-cohort  2010 Sub-cohort    Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, cohabitation  Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, education  Health behaviour  n (%)  n (%)  n  OR  CI  OR  CI  n (%)  n (%)  n  OR  CI  OR  CI   Daily smoking  Continued  315 (77.4)  92 (22.6)  407  1.00    1.00    363 (72.9)  135 (27.1)  498  1.00    1.00    Quit  219 (75.8)  70 (24.2)  289  1.09  0.76–1.56  1.00  0.69–1.44  179 (68.6)  82 (31.4)  261  1.23  0.88–1.71  1.21  0.86–1.70   Daily snus use  Continued  269 (78.2)  75 (21.8)  344  1.00    1.00    528 (77.4)  154 (22.6)  682  1.00    1.00    Quit  114 (76.0)  36 (24.0)  150  1.13  0.72–1.78  1.09  0.68–1.73  110 (71.4)  44 (28.6)  154  1.37  0.93–2.03  1.27  0.85–1.89   Daily tobacco use  Continued  605 (78.0)  171 (22.0)  776  1.00    1.00    874 (75.0)  292 (25.0)  1166  1.00    1.00    Quit  276 (75.0)  92 (25.0)  368  1.18  0.88–1.58  1.16  0.86–1.56  237 (67.9)  112 (32.1)  349  1.41  1.09–1.84  1.31  1.01–1.71   Insufficient fruit intake  Continued  748 (77.1)  222 (22.9)  970  1.00    1.00    1218 (71.4)  487 (28.6)  1705  1.00    1.00    Favourable change  410 (70.9)  168 (29.1)  578  1.38  1.09–1.74  1.30  1.02–1.66  319 (59.0)  222 (41.0)  541  1.74  1.42–2.12  1.64  1.33–2.01   Insufficient physical activity  Continued  1336 (74.5)  457 (25.5)  1793  1.00    1.00    1626 (65.7)  850 (34.3)  2476  1.00    1.00    Favourable change  342 (71.7)  135 (28.3)  477  1.15  0.92–1.44  1.10  0.88–1.39  448 (66.2)  229 (33.8)  667  0.98  0.82–1.17  0.98  0.82–1.18  ORs and 95% CIs of quitting hazardous alcohol use according to shift in health behaviours between baseline and follow-up. Table 3 Characteristics of participants with hazardous alcohol use and unfavourable health behaviours at baseline   2002 Sub-cohort  2010 Sub-cohort    Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, cohabitation  Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, education  Health behaviour  n (%)  n (%)  n  OR  CI  OR  CI  n (%)  n (%)  n  OR  CI  OR  CI   Daily smoking  Continued  315 (77.4)  92 (22.6)  407  1.00    1.00    363 (72.9)  135 (27.1)  498  1.00    1.00    Quit  219 (75.8)  70 (24.2)  289  1.09  0.76–1.56  1.00  0.69–1.44  179 (68.6)  82 (31.4)  261  1.23  0.88–1.71  1.21  0.86–1.70   Daily snus use  Continued  269 (78.2)  75 (21.8)  344  1.00    1.00    528 (77.4)  154 (22.6)  682  1.00    1.00    Quit  114 (76.0)  36 (24.0)  150  1.13  0.72–1.78  1.09  0.68–1.73  110 (71.4)  44 (28.6)  154  1.37  0.93–2.03  1.27  0.85–1.89   Daily tobacco use  Continued  605 (78.0)  171 (22.0)  776  1.00    1.00    874 (75.0)  292 (25.0)  1166  1.00    1.00    Quit  276 (75.0)  92 (25.0)  368  1.18  0.88–1.58  1.16  0.86–1.56  237 (67.9)  112 (32.1)  349  1.41  1.09–1.84  1.31  1.01–1.71   Insufficient fruit intake  Continued  748 (77.1)  222 (22.9)  970  1.00    1.00    1218 (71.4)  487 (28.6)  1705  1.00    1.00    Favourable change  410 (70.9)  168 (29.1)  578  1.38  1.09–1.74  1.30  1.02–1.66  319 (59.0)  222 (41.0)  541  1.74  1.42–2.12  1.64  1.33–2.01   Insufficient physical activity  Continued  1336 (74.5)  457 (25.5)  1793  1.00    1.00    1626 (65.7)  850 (34.3)  2476  1.00    1.00    Favourable change  342 (71.7)  135 (28.3)  477  1.15  0.92–1.44  1.10  0.88–1.39  448 (66.2)  229 (33.8)  667  0.98  0.82–1.17  0.98  0.82–1.18    2002 Sub-cohort  2010 Sub-cohort    Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, cohabitation  Continued hazardous  Quit hazardous  Total  Unadjusted model  Model 1 adjusted for age, sex, education  Health behaviour  n (%)  n (%)  n  OR  CI  OR  CI  n (%)  n (%)  n  OR  CI  OR  CI   Daily smoking  Continued  315 (77.4)  92 (22.6)  407  1.00    1.00    363 (72.9)  135 (27.1)  498  1.00    1.00    Quit  219 (75.8)  70 (24.2)  289  1.09  0.76–1.56  1.00  0.69–1.44  179 (68.6)  82 (31.4)  261  1.23  0.88–1.71  1.21  0.86–1.70   Daily snus use  Continued  269 (78.2)  75 (21.8)  344  1.00    1.00    528 (77.4)  154 (22.6)  682  1.00    1.00    Quit  114 (76.0)  36 (24.0)  150  1.13  0.72–1.78  1.09  0.68–1.73  110 (71.4)  44 (28.6)  154  1.37  0.93–2.03  1.27  0.85–1.89   Daily tobacco use  Continued  605 (78.0)  171 (22.0)  776  1.00    1.00    874 (75.0)  292 (25.0)  1166  1.00    1.00    Quit  276 (75.0)  92 (25.0)  368  1.18  0.88–1.58  1.16  0.86–1.56  237 (67.9)  112 (32.1)  349  1.41  1.09–1.84  1.31  1.01–1.71   Insufficient fruit intake  Continued  748 (77.1)  222 (22.9)  970  1.00    1.00    1218 (71.4)  487 (28.6)  1705  1.00    1.00    Favourable change  410 (70.9)  168 (29.1)  578  1.38  1.09–1.74  1.30  1.02–1.66  319 (59.0)  222 (41.0)  541  1.74  1.42–2.12  1.64  1.33–2.01   Insufficient physical activity  Continued  1336 (74.5)  457 (25.5)  1793  1.00    1.00    1626 (65.7)  850 (34.3)  2476  1.00    1.00    Favourable change  342 (71.7)  135 (28.3)  477  1.15  0.92–1.44  1.10  0.88–1.39  448 (66.2)  229 (33.8)  667  0.98  0.82–1.17  0.98  0.82–1.18  ORs and 95% CIs of quitting hazardous alcohol use according to shift in health behaviours between baseline and follow-up. Discussion In this longitudinal study with a relatively long follow-up, favourable health behaviours at baseline were associated with quitting hazardous alcohol use. These associations were robust to any definition of hazardous alcohol use as well as to the length of follow-up. In addition, sociodemographic characteristics and self-rated health of the hazardous drinkers did not appear to explain the associations. Moreover, favourable changes in dietary habits seemed to predict the transition from hazardous to non-hazardous alcohol use. The same pattern was discernible for quitting daily tobacco use, although the limited sample sizes hampered the precision of the results. The association between use of tobacco and quitting the hazardous use of alcohol is in line with previous findings of decreased alcohol use or abstention following smoking cessation.10,24 Smoking has also been related to increased alcohol use among Finnish adults in a prospective longitudinal study.25 Novel from the current study is the finding that use of the Swedish smokeless tobacco, i.e. snus, seems to be associated with hazardous alcohol use to the same extent as smoking. This is in line with previous cross-sectional analyzes of the SPHC, indicating average high alcohol consumption among snus users.26 Further, snus use was associated with an increased risk of alcohol dependence in a large prospective cohort study.27 Therefore, it seems likely that hazardous alcohol users with the concurrent daily use of tobacco constitute a risk group for the maintenance of the alcohol drinking profile. Daily fruit intake as an indicator of a healthy eating pattern was the only behaviour consistently associated with quitting hazardous alcohol use. Cross-sectional studies have suggested a decline in total dietary quality and fruit intake following increased alcohol consumption.28,29 In a Finnish study, being a male moderate drinker was associated with low intake of fruits while being a heavy drinker was associated with overall poor dietary quality.30 To our knowledge, change of diet and associated change of hazardous alcohol use have not been studied in longitudinal samples. Considering fruit intake as the only proxy for dietary habits is certainly a limitation of this study, but it sheds light on an association that warrants further investigation. At present, findings on the relationship between physical activity and alcohol use are contradictory and rest mainly on cross-sectional studies. Some studies found an association between increased alcohol consumption and increased physical activity,31,32 while others found a curvilinear pattern with increased physical activity among moderate drinkers33 or associations of both sedentary behaviour and high physical activity levels with higher levels of drinking.34 Despite the relatively large amount of publications, we were not able to identify studies relating physical activity to spontaneously quitting the hazardous use of alcohol. DeRuiter et al. found the change in physical activity and change in alcohol consumption not to be associated,35 but the sample in this study was not restricted to hazardous alcohol users. Our results on physical activity are somewhat conflicting. On the one side, we found that physically active individuals at baseline were more likely to quit hazardous alcohol use, although this association was weaker than for other health behaviours. On the other side, changes to favourable physical activity patterns suggested rather a negative impact on the change of hazardous alcohol use. Potential explanations include misclassification, selection, reverse causality or chance. Therefore, the relation of physical activity and hazardous alcohol use needs further analyzes that are more refined. In this study, adopting beneficial behaviour concerning diet and quitting daily tobacco use was predictive of quitting the hazardous use of alcohol. This pattern is in line with theories of behavioural change that posit that any positive lifestyle modification may foster new values and norms; raise self-efficacy, i.e. confidence in one‘s abilities to achieve behavioural goals; introduce skill training and resistance training, and finally promote intention to change and motivation enhancement.36 Accordingly, empirical findings support the possibility of favourable results in interventions targeting multiple behaviours. For instance, a meta-analysis reported that adding smoking cessation to alcohol treatment interventions enhanced the long-term effect on sobriety.37 Based on the results of the current study it can be hypothesized that interventions including multiple behavioural modifications may achieve better results even in the subgroup of hazardous drinkers. Additionally, greater impact on public health can be expected from actions targeting multiple-behaviours.38 Because no agreement exists on the sequence of interventions for coexisting health behaviours,12 this topic should be addressed in future studies. This study had several strengths. First, the prospective design enrolling a large population-based sample. Second, the follow-up period was long, extending over more than a decade, thus allowing the study of long-term modification of hazardous alcohol use. Third, the consistency of the results obtained with two different measures of hazardous use renders them more reliable, as these measures reflect different patterns of hazardous alcohol consumption.39 Heavy occasional drinking per se is associated with health risks4 and is usually neglected if subjects are asked to report their average consumption.40 Limitations of the study include a selection of respondents at baseline and attrition at follow-up. In fact, attrition was highest among the most socially disadvantaged groups that also are more likely to present unfavourable health behaviours. This selection may have resulted in a loss of statistical power and an attenuation of the associations under study, since we do not have reasons to assume that the direction of the association between health behaviours and alcohol use would be different among those retained and non-retained at follow-up. In total, this bias might impact the findings primarily by reduced study efficiency and limited generalizability. A second limitation is the different way questions on physical activity and diet were asked in 2002 and 2010. This alteration might have led to different classifications of change in physical activity and diet in the two sub-cohorts. The cut-off used to define sufficient physical activity was higher in 2010 than 2002, i.e. we strived to achieve high specificity for behavioural change at the expense of sensitivity. Again, these methodologic shortcomings possibly led to non-differential misclassification of the predictor and diluted the associations under investigation. However, since the magnitude of the association was similar in both samples, this bias is likely to be of a modest entity if any. A third limitation is the low reliability of self-reports, a common feature of studies including measures of health behaviours. However, questions on alcohol use were constructed according to good methodological practices including reference period, beverage-specific consumption, quantity and frequency and additionally a question of frequency of heavy drinking.39 Fourth, we might have underestimated hazardous use among women by using the common cut-off of five standard glasses on a single occasion for HOAC, while lately, sex-specific measures detecting corresponding blood concentration in men and women suggested a limit of four standard glasses for women.40 Conclusion A high proportion of hazardous alcohol users may spontaneously quit hazardous alcohol consumption, indicating favourable transitional patterns among these drinkers. Favourable health behaviours at baseline and their positive changes over time may predict changes of alcohol patterns in the desirable direction, thus suggesting the possibility to achieve concurrent behavioural modification in several domains. Acknowledgements The authors are grateful to the Centre for Epidemiology and Community Medicine, Stockholm County for making the SPHC data available for this study. Further, they wish to acknowledge Filip Andersson for statistical advice. M.R.G. designed the study, supervised the data analysis and critically revised the manuscript. E.S. participated in the study design, performed the statistical analyzes and drafted the manuscript. Y.F. and M.R. provided technical advice and participated in the interpretation and the discussion of the results. All authors read and approved the final version of the manuscript. Funding This work was partly supported by the Public Health Agency of Sweden (grant numbers 05576-2014-6.2, 03129-2015-6.2, 02332-2016-6.2). Partly by the, Stockholm County Council, CES (salary to M.R.G. and Y.F.) and The Swedish Council for Information on Alcohol and Other Drugs, CAN (salary M.R.) Conflicts of interest: None declared. Key points It is not known to what extent health behaviours predict change into less risky alcohol use among hazardous consumers. Our results suggest that having or achieving a healthy lifestyle predict desirable transitions in hazardous alcohol use. Co-occurring health behaviours may be considered when developing intervention strategies targeting this subgroup of drinkers. In particular, multiple behavioural changes should be addressed for hazardous drinkers hosting several unfavourable health behaviours. References 1 Institute for Health Metrics and Evaluation, University of Washington. GBD Compare Data Visualization [database on the Internet]. 2016. Seattle, WA, Available at: http://vizhub.healthdata.org/gbd-compare (2 June 2017, date last accessed). 2 Folkhälsomyndigheten [Public Health Agency of Sweden] Alkoholvanor—nationella resultat och tidsserier 2015 [Alcohol use—national results and time series 2015]. 2016. Available at: https://www.folkhalsomyndigheten.se/folkhalsorapportering-statistik/statistikdatabaser-och-visualisering/nationella-folkhalsoenkaten/levnadsvanor/alkoholvanor/ (24 March 2016, date last accessed). 3 Jarl J, Johansson P, Eriksson A, et al.   The societal cost of alcohol consumption: an estimation of the economic and human cost including health effects in Sweden, 2002. Eur J Health Econ  2008; 9: 351– 60. Google Scholar CrossRef Search ADS PubMed  4 Rehm J, Baliunas D, Borges GL, et al.   The relation between different dimensions of alcohol consumption and burden of disease: an overview. Addiction  2010; 105: 817– 43. Google Scholar CrossRef Search ADS PubMed  5 Rehm J, Marmet S, Anderson P, et al.   Defining substance use disorders: do we really need more than heavy use? Alcohol Alcohol  2013; 48: 633– 40. Google Scholar CrossRef Search ADS PubMed  6 Klingemann H, Sobell MB, Sobell LC. Continuities and changes in self‐change research. Addiction  2010; 105: 1510– 8. Google Scholar CrossRef Search ADS PubMed  7 Sobell LC, Ellingstad TP, Sobell MB. Natural recovery from alcohol and drug problems: methodological review of the research with suggestions for future directions. Addiction  2000; 95: 749– 64. Google Scholar CrossRef Search ADS PubMed  8 Cunningham JA. Resolving alcohol-related problems with and without treatment: the effects of different problem criteria. J Stud Alcohol  1999; 60: 463– 6. Google Scholar CrossRef Search ADS PubMed  9 Blomqvist J, Cunningham J, Wallander L, Collin L. Att förbättra sina dryckesvanor: om olika mönster för förändring och om vad vården betyder. [Improving One's Drinking Habits–on different patterns of change and on the role of alcohol treatment]. Stockholm: Stockholm University, Faculty of Social Sciences Centre for Social Research on Alcohol and Drugs; 2007. Report No.: 978-91-85707-04-01. 10 Karlamangla A, Zhou K, Reuben D, et al.   Longitudinal trajectories of heavy drinking in adults in the United States of America. Addiction  2006; 101: 91– 9. Google Scholar CrossRef Search ADS PubMed  11 Moore AA, Gould R, Reuben DB, et al.   Longitudinal patterns and predictors of alcohol consumption in the United States. Am J Public Health  2005; 95: 458– 65. Google Scholar CrossRef Search ADS PubMed  12 Meader N, King K, Moe-Byrne T, et al.   A systematic review on the clustering and co-occurrence of multiple risk behaviours. BMC Public Health  2016; 16: 657. Google Scholar CrossRef Search ADS PubMed  13 de Vries H, van 't Riet J, Spigt M, et al.   Clusters of lifestyle behaviors: results from the Dutch SMILE study. Prev Med  2008; 46: 203– 8. Google Scholar CrossRef Search ADS PubMed  14 Khaw KT, Wareham N, Bingham S, et al.   Combined impact of health behaviours and mortality in men and women: the EPIC-Norfolk prospective population study. PLoS Med  2008; 5: e70. Google Scholar CrossRef Search ADS   15 Svensson AC, Fredlund P, Laflamme L, et al.   Cohort profile: the Stockholm public health cohort. Int J Epidemiol  2013; 42: 1263– 72. Google Scholar CrossRef Search ADS PubMed  16 Romelsjö A, Leifman H, Nyström S. A comparative study of two methods for the measurement of alcohol consumption in the general population. Int J Epidemiol  1995; 24: 929– 36. Google Scholar CrossRef Search ADS PubMed  17 Leifman H. Alkohol och hälsa: en kunskapsöversikt om alkoholens positiva och negativa effekter på vår hälsa . [Alcohol and health a review]. Stockholm: Statens folkhälsoinstitut [Public Health Agency of Sweden], 2005. 18 Naimi TS, Nelson DE, Brewer RD. The intensity of binge alcohol consumption among U.S. adults. Am J Prev Med  2010; 38: 201– 7. Google Scholar CrossRef Search ADS PubMed  19 Connor Gorber S, Schofield-Hurwitz S, Hardt J, et al.   The accuracy of self-reported smoking: a systematic review of the relationship between self-reported and cotinine-assessed smoking status. Nicotine Tob Res  2009; 11: 12– 24. Google Scholar CrossRef Search ADS PubMed  20 Sepp H, Ekelund U, Becker W. Enkätfrågor om kost och fysisk aktivitet bland vuxna—underlag till urval av frågor i befolkningsinriktade enkäter [Survey questions for diet and physical activity—basis for the selection of questions in populationsbased surveys]: Swedish Food Agency, 2004. Report No.: 21. 21 Orsini N, Bellocco R, Bottai M, et al.   Reproducibility of the past year and historical self-administered total physical activity questionnaire among older women. Eur J Epidemiol  2007; 22: 363– 8. Google Scholar CrossRef Search ADS PubMed  22 World health organization (WHO). Global Recommendations on Physical Activity for Health . Geneva, 2010, Available at: http://www.who.int/dietphysicalactivity/publications/9789241599979/en/ (7 July 2017, date last accessed). 23 Burström B, Fredlund P. Self rated health: is it as good a predictor of subsequent mortality among adults in lower as well as in higher social classes? J Epidemiol Commun Health  2001; 55: 836– 40. Google Scholar CrossRef Search ADS   24 Dawson DA, Goldstein RB, Grant BF. Prospective correlates of drinking cessation: variation across the life-course. Addiction  2013; 108: 712– 22. Google Scholar CrossRef Search ADS PubMed  25 Laaksonen M, Luoto R, Helakorpi S, Uutela A. Associations between health-related behaviors: a 7-year follow-up of adults. Prev Med  2002; 34: 162– 70. Google Scholar CrossRef Search ADS PubMed  26 Engström K, Magnusson C, Galanti MR. Socio-demographic, lifestyle and health characteristics among snus users and dual tobacco users in Stockholm County, Sweden. BMC Public Health  2010; 10: 619. Google Scholar CrossRef Search ADS PubMed  27 Norberg M, Malmberg G, Ng N, Broström G. Use of moist smokeless tobacco (snus) and the risk of development of alcohol dependence: a cohort study in a middle-aged population in Sweden. Drug Alcohol Depend  2015; 149: 151– 7. Google Scholar CrossRef Search ADS PubMed  28 Breslow RA, Guenther PM, Juan W, Graubard BI. Alcoholic beverage consumption, nutrient intakes, and diet quality in the US adult population, 1999-2006. J Am Diet Assoc  2010; 110: 551– 62. Google Scholar CrossRef Search ADS PubMed  29 Ruidavets JB, Bataille V, Dallongeville J, et al.   Alcohol intake and diet in France, the prominent role of lifestyle. Eur Heart J  2004; 25: 1153– 62. Google Scholar CrossRef Search ADS PubMed  30 Fawehinmi TO, Ilomaki J, Voutilainen S, Kauhanen J. Alcohol consumption and dietary patterns: the FinDrink study. PLoS One  2012; 7: e38607. Google Scholar CrossRef Search ADS PubMed  31 Burger M, Mensink GBM, Bergmann E, Pietrzik K. Characteristics associated with alcohol consumption in Germany. J Stud Alcohol  2003; 64: 262– 9. Google Scholar CrossRef Search ADS PubMed  32 French MT, Popovici I, Maclean JC. Do alcohol consumers exercise more? Findings from a national survey. Am J Health Promot  2009; 24: 2– 10. Google Scholar CrossRef Search ADS PubMed  33 Smothers B, Bertolucci D. Alcohol consumption and health-promoting behavior in a U.S. household sample: leisure-time physical activity. J Stud Alcohol  2001; 62: 467– 76. Google Scholar CrossRef Search ADS PubMed  34 Kopp M, Burtscher M, Kopp-Wilfling P, et al.   Is there a link between physical activity and alcohol use? Subst Use Misuse  2015; 50: 546– 51. Google Scholar CrossRef Search ADS PubMed  35 deRuiter WK, Cairney J, Leatherdale ST, Faulkner GE. A longitudinal examination of the interrelationship of multiple health behaviors. Am J Prev Med  2014; 47: 283– 9. Google Scholar CrossRef Search ADS PubMed  36 Glanz K, Bishop DB. The role of behavioral science theory in development and implementation of public health interventions. Annu Rev Public Health  2010; 31: 399– 418. Google Scholar CrossRef Search ADS PubMed  37 Prochaska JJ, Delucchi K, Hall SM. A meta-analysis of smoking cessation interventions with individuals in substance abuse treatment or recovery. J Consult Clin Psychol  2004; 72: 1144– 56. Google Scholar CrossRef Search ADS PubMed  38 Prochaska JJ, Spring B, Nigg CR. Multiple health behavior change research: an introduction and overview. Prev Med  2008; 46: 181– 8. Google Scholar CrossRef Search ADS PubMed  39 Greenfield TK, Kerr WC. Alcohol measurement methodology in epidemiology: recent advances and opportunities. Addiction  2008; 103: 1082– 99. Google Scholar CrossRef Search ADS PubMed  40 Stahre M, Naimi T, Brewer R, Holt J. Measuring average alcohol consumption: the impact of including binge drinks in quantity-frequency calculations. Addiction  2006; 101: 1711– 8. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

The European Journal of Public HealthOxford University Press

Published: Nov 17, 2017

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

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

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

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.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off