Leisure-time physical activity in youth as a predictor of adult leisure physical activity among Japanese workers: a cross-sectional study

Physical activity is associated with reduced risk of many chronic diseases [1] including cardiovascular diseases [2, 3], type 2 diabetes [4], and several cancers [5, 6], in addition to all-cause mortality [7]. Despite such apparent benefits of physical activity, a large proportion of adults in developed countries such as Japan do not participate in regular leisure-time physical activity. The National Health and Nutrition Surveys in Japan for 1997, 2004, and 2009 revealed that 70–80% of the working generation (aged 20–59 years) were not sufficiently active (i.e., were exercising less than 30 min twice a week) [8]. Because workers in Japan may not have enough leisure time or be in the habit of participating in regular leisure-time physical activity, it remains unclear how their leisure-time physical activity habits could be developed.

In Western countries, a non-negligible number of studies suggest the possibility that physical activity in childhood and youth is associated with future physical activity [9]. Such a “carryover” may also be effective for building workers’ leisure-time physical activity habits in Japan. However, previous studies have not reported consistent results [9‐14]; therefore, further investigation is needed. Moreover, epidemiological evidence that considers specific age and intensity of leisure-time physical activity in youth and the relationship of these variables to leisure-time physical activity habits in adults is lacking, particularly in Asian countries. The Japanese mentality is unique and differs in potentially important ways from that of Western societies. Workaholism, obedience, tolerance and selfless devotion to country are typical Japanese tendencies that may influence workers’ behavioral priorities and their responses to health promotion efforts. Because it is also unclear whether leisure-time physical activity in youth predicts leisure-time physical activity in adulthood among Japanese people, an epidemiological study is needed. In addition, the optimal intensity level of leisure-time physical activity at specific ages that is necessary to develop future leisure-time physical activity habits remains to be determined. For example, experience of strenuous leisure-time physical activity in youth may build an individual’s physical strength and enable them to resume leisure-time physical activity in adulthood.

The purpose of the present study was to investigate, in a sample of Japanese workers, the associations between age- and intensity-specific leisure-time physical activity in youth and participation as an adult in regular leisure-time physical activity. Whereas most previous studies have examined the association between total activity in youth and total activity in adulthood [10], this study focused on leisure-time physical activity. This will provide further insights for promoting a more desirable physical activity strategy, because leisure-time activity is modifiable and controllable. Furthermore, health effects of leisure-time physical activity may differ from those of work-related physical activity. In fact, higher occupational physical activity has recently been shown to have acute and chronic adverse effects, such as increased risk of cardiovascular disease and all-cause mortality, particularly among men [15]. Therefore, our independent evaluation of leisure-time physical activity will be informative. We used ages 12 and 20 years as representative time points for leisure-time physical activity in youth, based on the following precedents. In previous studies conducted in Australia [10] and Canada [16] that tracked physical activity, the mean age at baseline was 12 years. In another tracking study conducted in Belgium [17], the mean ages of the male and female follow-up samples at baseline was 20 and 22 years, respectively.

Leisure-time physical activity is defined as “physical activity performed during exercise, recreation, or any time other than that associated with one’s regular occupation, housework, or transportation” [18, 19], and as “primarily exercise or sports-related activities” [20]. “Exercise and other forms of physical training are types of recreational physical activity” [21]. Therefore, “leisure-time physical activity” may be synonymous with “exercise.” These two terms were treated impartially in the present study.

Table 1 shows participants’ basic characteristics according to current participation in leisure-time physical activity by sex. There were no significant differences by current participation in leisure-time physical activity, with the exception of the number of holidays. Women participating in current leisure-time physical activity had more holidays than women without current participation in leisure-time physical activity, and women had more holidays than men.

Figure 1 shows the percentages of workers with current participation in regular leisure-time physical activity by age group. There was a higher proportion of participation in older age groups (50s and 60s).

Table 2 shows the ORs and 95% CIs for workers’ current participation in regular leisure-time physical activity by intensity levels of age-specific leisure-time physical activity. We found that strenuous leisure-time physical activity at age 12 years was significantly positively associated with current participation in leisure-time physical activity among male workers [adjusted OR (95% CI) = 2.29 (1.02, 5.14)]. In contrast, female leisure-time physical activity at age 12 years was not significantly associated with adulthood participation in regular leisure-time physical activity. Moreover, strenuous and moderate leisure-time physical activity at age 20 years were also significantly positively associated with workers’ current participation in regular leisure-time physical activity, and this held true for both males and females.

Table 3 shows ORs and 95% CIs for workers’ current participation in regular leisure-time physical activity by history of participation in sports club activities at a young age. Female participation in such activities during high school or higher professional school was significantly positively associated with their current participation in regular leisure-time physical activity [adjusted OR (95% CI) = 3.21 (1.54, 6.68)]. In addition, female sports club activity in junior/technical college or university was also associated with participation in adulthood leisure-time physical activity. Point estimates for other ORs ranged from 1.16 to 2.18, although they were statistically insignificant.

The direction and significance of the ORs in Tables 2 and 3 were unchanged after additionally adjusting for type of business (railway, transport, driving school, and hospital), night shift work, and the number of holidays (data not shown), with the exception of male sports club activity during high school or higher professional school [adjusted OR (95% CI) = 2.54 (1.04, 6.21)].

Among Japanese workers, some leisure-time physical activities in youth were significantly positively associated with participation in regular leisure-time physical activity in adulthood. In particular, male strenuous leisure-time physical activity at 12 years old was positively associated with adulthood leisure-time physical activity. Strenuous and moderate leisure-time physical activities at age 20 years were also predictive of workers’ current leisure-time physical activity behavior. Associations were evident and most were relatively strong in magnitude and consistent in direction. To our knowledge, this is the first study to investigate the association between leisure-time physical activity in youth and in adulthood among workers in an Asian country, also taking age- and intensity-specific activity into account. Although we only considered leisure-time physical activity, exposure in youth was a predictor of adult leisure-time physical activity in our Japanese sample. There is supportive evidence from a study in Australia reporting that childhood sport participation predicted adult physical activity, while total physical activity in childhood did not predict adult activity [10].

In this study, workers were distributed over a wide age range. Workers in older age groups tended to participate in regular leisure-time physical activity. This tendency is consistent with that observed in Japan’s National Health and Nutrition Survey [8]. Such change in behavior might be attributed in part to health consciousness and increased leisure time, given that those in older age groups no longer needed to care for young children.

Different intensities of leisure-time physical activity in youth were predictive of future leisure-time physical activity behavior. For males, strenuous leisure-time physical activity at age 12 years had a greater lingering effect than did moderate activity. Plausibly, children who engaged in strenuous leisure-time physical activity would gain physical strength more easily, enabling them to participate in future leisure-time physical activity. Additionally, because moderately intense leisure-time physical activity is often more accessible, safer, and more convenient, this level of activity at age 20 years may easily lead to lifelong participation. Our age- and intensity-specific analyses revealed that both strenuous and moderate intensity leisure-time physical activity at age 20 years was effective in developing future leisure-time physical activity habits. Thus, an age- and intensity-specific analysis can provide useful insights for leisure-time physical activity promotion strategies.

A sex difference in the results at age 12 was clearly observed. Such weaker associations in females may be attributed in part to major transitions in their life course, such as the transition from schooling to employment, from singlehood to marriage and having children, or experiencing unemployment [9]. Such events usually have different influences, particularly on continuous employment, for males and females in Japan. These findings regarding sex difference are supported by evidence from the majority of previous studies in Western countries. These studies also tend to show that childhood physical activity positively predicts adulthood physical activity in males, whereas about half of the associations are non-significant in females [9]. This tendency may be common in developed countries beyond the sphere of their residential area and local mindset.

Females’ sports club activity during high school or above was significantly positively associated with regular participation in adult leisure-time physical activity. This was not the case for males, although the point estimate of the OR was high. The small sample size of the subgroup might have led to the null result. Furthermore, with regard to sports club activities prior to high school in both male and female participants, no significant associations were seen. It is known that physical activity in the distant past generally shows a weaker correlation with adulthood physical activity than does activity the near past [11]. Another possible explanation for the lack of a significant association between sports club activities at a younger age and participation in leisure-time physical activity in adulthood is that some students who did not participate in organized school sports might have participated in other leisure-time physical activities and physical education. Although a previous study conducted in the United States [13] reported that the frequencies of forced and encouraged exercise during the preteen years were inversely associated with adult physical activity, such an association was not observed in the present study. Therefore, sports club activity may also be recommended for young people in Japan.

From a descriptive perspective, Tables 2 and 3 show that most workers who had participated in leisure-time physical activity in youth were not currently participating in regular leisure-time physical activity. For example, of the 127 male workers who had participated in strenuous leisure-time physical activity at age 12 years, 109 did not exercise regularly in adulthood. This suggests that continuing or resuming leisure-time physical activity in adulthood might not have been easy for the participating workers. Further studies are needed to investigate the kind of work environment that supports their participation in leisure-time physical activity.

Major strengths of the present study were as follows. Our questionnaire for assessing exposure to leisure-time physical activity at a young age took into consideration the intensity, frequency and timing of leisure-time physical activity. Workers’ current participation in regular leisure-time physical activity was clearly defined in terms of frequency and duration. Whereas most previous studies have examined the association between total activity in youth and in adulthood, the present study focused on leisure-time physical activity.

Several potential limitations of the study also warrant mention. First, we carried out multiple comparisons, which might have led to some false-positive results. However, we did not apply the Bonferroni correction because we considered this method too conservative. Second, owing to the cross-sectional nature of this investigation and self-reported measures of leisure-time physical activity habits in youth, we cannot completely rule out the possibility of error owing to recall bias. However, the reliability of long-term recall of participation in physical activity has been investigated in the United States and modest correlations have been reported [25]. Third, the response rate was suboptimal. However, the evidence for the relationship between low response rate and non-response bias is generally weak [26]; therefore, our moderate response rate may not necessarily indicate non-response bias. Fourth, the sample in the present study may not be representative of the general working population in Japan; therefore, our findings may not be applicable to a wide range of Japanese workers, although the observed associations were similar to those in previous studies. Although a disproportionate number of female participants (74.8%) were night-shift workers, this type of work is common in modern society. In developed countries, about 20% of the working population is engaged in shift work [27]. More detailed information regarding the type of work may provide further insights when assessing these associations. In addition, our sample size was not large enough to allow dose–response analysis using the frequency of leisure-time physical activity in youth, as well as subgroup analyses by age groups. A larger study sample that enables such analyses should be considered for future studies. In addition, timing of assessment of leisure-time physical activity in youth at ages other than 12 and 20 years may be useful, as the effects of leisure-time physical activity at other ages might differ from those in transitional years such as ages 12 and 20 years.

In conclusion, our results suggest that certain aspects of leisure-time physical activity in youth may predict current participation in regular leisure-time physical activity in Japanese workers. Nurturing leisure-time physical activity in youth may be an effective measure to develop the future leisure-time physical activity habits of adult workers.