Id		116
Author		Hyyppä, M. T.; Mäki, J.; Impivaara, O.; Aromaa, A
Title		Leisure participation predicts survival: a population-based study in Finland.
Reference		Hyyppä, M. T.; Mäki, J.; Impivaara, O.; Aromaa, A. (2006). Leisure participation predicts survival: A population‑based study in Finland. Health Promotion International, 21(1): 5–12.
Keywords		leisure participation; survival; longitudinal survey
Link to article		https://doi.org/10.1093/heapro/dai027
Abstract		The authors study whether leisure participation is an independent predictor of survival over 20 years. Of the nationally representative sample of 8000 adult Finns (Mini-Finland Health Survey), aged >30 years, the cohort of 30–59 years (n 5087) was chosen for the Cox proportional survival analyses. The sum score of leisure participation was divided in quartiles (the lowest quartile = scarce = 0–6), two intermediate quartiles = 7–11 and the highest quartile = abundant = 12–21). Adjusted for statistically significant covariates (age, tobacco smoking, alcohol consumption, obesity, self-rated health and diagnosed chronic diseases), and with scarce participation as the reference, the hazard ratios and 95% confidence intervals (CIs) for the risk of death were 0.80, 0.67–0.95 (intermediate) and 0.66, 0.52–0.84 (abundant) for men. The association was insignificant in women with good health. The results show that leisure participation predicts survival in middle-aged Finnish men and its effect is independent of demographic features, of health status and of several other health-related factors. The beneficial effect emphasizes the significance of leisure activities for the promotion of men’s health.
Metodology		First, we divided the population in three groups according to the quartiles of sum score in leisure participation (the lowest quartile = scarce = 0–6), two intermediate quartiles = 7–11 and the highest quartile = abundant = 12–21). Secondly, we applied multivariate Cox proportional hazard models for survival to identify and to control for relevant health-related covariates and their interactions with leisure participation measures. Thirdly, controlling for the health-related variables that significantly predicted survival and/or confounded the relationship between leisure participation scores and survival, we calculated hazard ratios with 95% confidence intervals (CIs). The SAS procedures (PHREG, LIFETEST) were applied for statistical calculations.
Findings		At the end of the follow-up, 68.5% of the subjects with scarce leisure participation and 84.6% of the subjects with intermediate and abundant participation were alive. Compared with scarce leisure participation, the hazard ratios of the risk of death for intermediate or abundant. participation were significant, even after adjustments for the statistically significant covariates, i.e. age, (gender), tobacco smoking, alcohol use, obesity, self-rated health and diagnosed chronic disease. Inclusion of the death cases occurring only after a 3-year follow-up did not notably alter the association in men but reduced the association in women. In men and women, age, tobacco smoking, chronic disease and obesity were predictive of survival (or mortality). In men, but not in women, alcohol use and self-rated good health were also associated with survival. In addition, self-rated health was identified as an effectmodifying factor in women. Among the women with good health, the hazard ratios for the risk of death were insignificant: 0.83, 0.50–1.37 (intermediate) and 0.97, 0.57–1.64 (abundant). Entering native tongue, type of residence, migration, marital status, household income, employment status, socio-economic status, education level, number of close friends, trustful relations and mental health status score to the adjusted models, or deleting outdoor and productive activity and studying from the sum score of the leisure participation did not significantly alter the hazard ratios
Open Access		YES
DOI		doi.10.1093/heapro/dai027
Search Database		Researcher knowledge
Technique		Multivariate Cox proportional Hazard; PHREG; LIFETEST; statistical calculation; General Health Questionaire of Finland;

ARTICLE