Physical activity in men and relation to prostate cancer

Abstract

Physical activity is hypothesized to reduce the risk of several chronic diseases and enhance longevity. Low levels of activity have been associated with some but not all types of cancers. Although an association with physical activity and prostate cancer the most common cancer among Swedish men and with an increasing incidence is biologically plausible, results from published studies are inconclusive.

Most of the questionnaires used measure only part of total physical activity leisure-time activity and/or occupational which might lead to misclassification of total physical activity level and to dilution of risk estimates. Moreover, despite a large public health interest in physical activity, and the fact that the incidence of many chronic diseases associated with physical inactivity, is increasing over time and with age, only few reports to date have addressed levels and trends of different types of physical activity by age and over time.

We evaluated the validity and reproducibility of a short self-administered physical activity questionnaire, intended to measure present total daily 24-hour physical activity (Paper I). The questionnaire included questions on level of physical activity at work, hours per day of walking/bicycling, home/household work, leisure-time activity/inactivity, and sleeping. Levels of total activity were quantitatively assessed by adding the activities together and measured as MET-hours/day (metabolic equivalents of resting metabolic rate).

One hundred eleven men in central Sweden, aged 44-78, completed the questionnaire and one or two 7-day activity records used as the reference method. Spearman correlation coefficient was 0.56 between total daily activity score estimated from the first questionnaire and activity records (validity) (r=0.69 when deattenuated for random error in the questionnaire) and 0.65 between the first and a second repeated questionnaire (reproducibility). Significantly higher validity correlations were observed in men with self-reported body mass index below 26 kg/m2 than in heavier men. We conclude that our questionnaire may be considered a valid instrument for assessing total physical activity.

Levels of total physical activity and different types of activities were studied in relation to (1) age in a cross-sectional setting (Paper II), (2) age in a longitudinal setting and (3) calendar-time (Paper III). In a population-based cohort of 33,466 men aged 45-79 years in central Sweden, we collected information on levels of total current physical activity (the validated self-administered questionnaire), and retrospectively at ages 15, 30 and 50 years. Total daily physical activity decreased with age, both in cross-sectional (-4% from age 45 to 79) and in longitudinal analyses (-4% from age 15 to 50). The decrease was greatest among obese men, current smokers, low-educated men, and men with low self-rated health in cross-sectional analyses. Total physical activity decreased over a period of 60 years in all age groups (15, 30, 50 years).

Effects of occupational physical activity on risk for prostate cancer were investigated in a nationwide cohort that was established through linkage between Swedish census data in 1960 and 1970, and the Cancer register 1971-1989 (Paper IV). Physical activity levels at work were classified based on the codes for occupational titles in the census. A total of 19,670 prostate cancers occurred during the 19 years of follow-up in the cohort of men with the same physical activity level in 1960 and 1970 (n=673,443). Sedentary men had a small but statistically significant increased risk of 11% as compared to men with a very high/high activity level.

In conclusion, we observed a slightly increased risk for prostate cancer in men with sedentary and light occupations. From a public health point of view, observed decreasing levels of total physical activity both with age and over time (last 60 years), require undertaking of effective preventive strategies since many chronic diseases associated with physical inactivity have been increasing over time.