Background
Regular exercise is widely known as one of the most important health behaviors for maintaining quality of life and preventing falls, hospitalization and functional impairment among older adults [
1]. Despite these benefits, around 50% of Japanese older adults do not participate in regular exercise [
2], and approximately 50% of older adults fail to continue exercising within 6 months [
3]. Therefore, strategies to maintain good exercise adherence for a long period are the key to a successful exercise program [
4].
Although there are some exercise programs contrived to promote exercise adherence, such as telephone [
5] and mailing [
6] services, spousal-pair-based exercise programs are expected to be an innovative and gratis approach to improving exercise participation among older adults [
7]. A cross-sectional study found that the husband’s physical activity level was associated with the wife’s physical activity level for both structured and unstructured physical activities [
7]. Wallance et al. retrospectively compared exercise adherence over the 12-month intervention between middle-aged couples and non-couples, and reported that attendance and drop-out rates were significantly better among the couples compared to non-couples [
8]. Our previous study also identified that attending an exercise class as older married couples significantly enhanced exercise adherence during a 3-month intervention period [
9]. These previous studies indicate that involvement of one’s spouse may be an effective strategy to improve exercise adherence among older adults. Social support, such as support from a spouse, influenced exercise adherence via exercise self-efficacy [
10]. As such, we hypothesized that individuals who participated in an exercise class with their spouse would maintain higher exercise adherence compared to individuals who participated in the program alone, since attending the exercise class with a spouse would enhance self-efficacy. However, to our knowledge, there were no prospective follow-up studies that examined the effect of spousal-based exercise programs on long-term exercise adherence among older adults.
Therefore, the purpose of this prospective follow-up study was to examine the effects of an exercise program for older married couples on long-term exercise adherence.
Discussion
To our knowledge, this study is the first to prospectively examine the efficacy of an exercise program for older married couples in maintaining exercise adherence over a 24-week follow-up period. The major finding of this study was that older couples—who were recruited, attended the exercise sessions, and exercised together—were more likely to maintain greater adherence to walking exercise over the 24 weeks following termination of the exercise program than were non-couples. We also found that the older couples received significantly greater social support from family members than did the non-couples, which might have led to their greater adherence to walking exercise after support from exercise instructors ceased.
Several recent studies have provided evidence of significant relationships between marital status and health outcomes [
15‐
21]. Most of these studies indicated that being single, divorced, or widowed is a risk of adverse health outcomes, which suggest that the presence of a spouse is key for helping individuals adopt a more active lifestyle. Wankel et al. similarly reported that social support from a spouse can contribute to greater activity, especially among the older population [
22]. Sallis et al. noted that the physical activity level of older women was more strongly associated with spousal support than was that of younger women [
23]. Given the fact that support from a spouse has such a robust influence on individuals’ health behavior, especially among older adults, our study was right to confirm the efficacy of an exercise program for older married couples on exercise adherence.
The results of the current study broaden current evidence by showing that an exercise program targeting older married couples can improve regular walking exercise during both the supervised and unsupervised periods. This expands on Wallance et al.’s study, wherein they retrospectively compared the attendance and dropout rates of an exercise program during a 12-month intervention period between married pairs (16 pairs) and married singles (
n = 32) in middle-aged adults [
8]. Notably, monthly attendance was significantly higher and dropout rate was significantly lower (by 13.9% and 36.7%, respectively) among married pairs than among married singles. Interestingly, the most frequent reason for dropping out was lack of support from the spouse. Our previous study also compared the rate of full attendance for exercise programs between a CG and an NCG during an exercise intervention period [
9]. This previous study showed that the full attendance rate of the CG was significantly higher (by about 9.1%) than was that of the NCG. This was particularly true in the latter half of the intervention (weeks 6–8). Additionally, the adherence rate of walking exercise in the CG was significantly higher than that of the NCG during the intervention period. Therefore, our previous study was consistent with the notion that an exercise intervention targeting older married couples may be useful for maintaining exercise program participation and walking during the intervention period.
Although the mechanism behind our results is somewhat unclear, they might be partially explainable by exercise social support. Specifically, support from spouses appears to encourage individuals to participate in walking. Oka et al. identified a number of psychological, social, and environmental factors related to exercise participation among 1932 Japanese adults using structural equation modeling [
10]. They presented that social support influenced exercise practice via exercise self-efficacy. As such, a possible mechanism for the effectiveness of the intervention is that support from the spouse promotes individuals’ exercise self-efficacy, which in turn increases their motivation to participate in walking. However, we found that, although there was a significant time × group interaction for support from family (Fig.
3), there was no significant interaction for exercise self-efficacy between the two groups (Table
2). Thus, the mechanism behind our results appears to be that support from the spouse directly influences older adults’ exercise adherence, independent of exercise self-efficacy. It must be noted that we did not directly investigate the specific effect of support from a spouse; as such, the idea that the CG may have received greater support from their spouse (e.g., encouragement and motivation for walking) compared with the NCG remains mere speculation.
We found no significant difference in the adherence rate of strength exercise between the two groups. In general, exercise adherence is associated with a variety of factors that fall into a range of categories, such as routine-related, intrinsic, biophysical, psychosocial, environmental, and resource-related factors. McArthur et al. qualitatively identified the enablers and barriers to adherence to regular exercise among middle-aged women [
24]. The most oft-reported enabling factor was “an established daily structure that incorporated exercise” (broad theme: routine-related factors). For example, participants who wanted to integrate regular exercise into their daily routine thought it important to make it a habitual lifestyle activity that was as unconscious as daily tooth-brushing. The frequency of strength exercise was notably higher than was that of walking, suggesting that most participants felt that strength exercise was one of their habitual lifestyle activities. As such, because most participants had integrated strength exercise into their routine lifestyle, support from a spouse might not have had any effect on the practice of this exercise. The strength exercise utilized in the present study may thus be a useful intervention for older adults who live alone, as it appears to be easy to maintain and does not require social support.
Although the adherence rate of walking exercise was significantly higher in the CG than in the NCG, both rates equally declined over time, and there was no significant interaction in walking adherence between the two groups (data not shown). Chogahara et al. reviewed various studies on how social support related to participation in exercise among older adults [
25,
26]. They found that most previous studies had overly emphasized the positive aspects of social support, while its negative aspects were largely neglected. Indeed, Barnett et al. explored and described how spousal support can influence both spouses’ physical activity behavior using a qualitative approach, and suggested that excessive demands from the spouse can negatively influence adherence [
27]. This means that we cannot exclude the possibility that support from spouse negatively affected exercise participation during the study period. For example, if a husband complained to his wife about her walking speed (or vice versa), it could decrease both of their motivations to continue walking. Coexistence of the negative and positive effects of spousal communication may be the reason that there was no significant interaction in walking adherence between the two groups. An educational exercise intervention that improves the effects of spousal support on exercise participation will be needed in the future for older married adults.
Limitations
The strengths of the current study were that it was the first to prospectively identify the effects of an exercise intervention for older married couples on exercise adherence. This novel strategy has the potential to encourage older males and people with little interest in the adherence to walking exercise in health promotion activities through spousal invitation. However, there were several limitations in this study. First, there was a possibility of selection bias because participants who were interested in practicing exercise are more likely to participate in such studies. Thus, the results of this study might not be generalizable to the overall Japanese population, especially among older adults who are not interested in practicing exercise. Second, there is a possibility of arbitrary bias due to the non-randomized group allocation and because blinding was not possible in this study design. A better designed approach to stimulate exercise adherence that would allow for recruitment of control groups should be used in the future [
28,
29]. These methods would provide more detailed insights into the processes underlying spousal support. Third, although the current study prospectively assessed exercise adherence using an exercise diary, which could stimulate adherence to exercise, using an objective evaluation of exercise (e.g., an accelerometer) would have provided more reliable results. Finally, the final sample size in the NCG (
n = 59) was slightly below the required sample size (
n = 62) because of exclusion of many participants with regular exercise habits, withdrawals after consent, and the limited study period given by our research funding. However, this study was sufficiently powered to detect the difference in adherence to walking exercise. A well-designed randomized controlled trial using an objective measurement of exercise adherence and an appropriate sample size will be needed to validate the novel findings of this study. Additionally, future studies should identify whether not-married couple units, such as pairs of friends (friend-pair-based exercise programs), have a similar effect on exercise adherence as married couples do.