Interactive effect of socio-eco-demographic characteristics and perceived physical activity barriers on physical activity level among older adults

The hypothesis of the present study was that evaluating the moderation effect of perceived PA barriers on association between SED characteristics and PA of older adults would help expand upon our knowledge on these associations and explain some of the inconsistencies in the relevant literature. Two out of the six direct associations evaluated between SED factors and PA were significant (Table 3). Besides, approximately 28.5% of the tested interactions between SED factors and perceived PA barriers in relation to PA were found to be significant (Table 3). These findings demonstrate that there is moderation by perceived PA barriers, suggesting there would be further complexities than could be explained only through SED factors.

The results of the present research show that about one fifth of older adults was characterized as sedentary (Table 2). Besides, 44.8% of them found to be under-active regular, meaning that less than the half of them regularly do some light PA every week. Only 11.7% of them met the minimum recommended weekly PA, doing ≥30 min moderate physical activities, 5 or more days a week or doing ≥20 min of vigorous physical activities, 3 or more days a week. Therefore, 88.3% of older adults did not meet the PA level recommended by world leading institutions in health and fitness, including American College of Sports Medicine [28], which is consistent with other findings on PA of Iranian older adults [29]. Further, according to the data available from a review study on PA of older adults, the prevalence of low PA among Iranian older adults found in this study was higher than those observed in other 63 low-to-middle income countries studied previously [30].

In the relevant literature, there have been controversial results on whether increasing age among older adults is associated with PA participation. Resnick et al. (2000) observed no significant differences in the ages of those who exercise regularly and those who did not [31]. Conversely, Janke et al. (2006) studied the leisure habits of older adults and found a significant decrease in PA between the ages of 70 and 80 [32]. In the present study, it was observed that PA participation declines with age. Participants aged 60–75 years demonstrated significantly higher PA levels compared to their older counterparts. Also, in the regression analysis, there was a tendency (albeit not significantly) towards significant effect of age for PA outcomes.

The results showed that there are significant differences in PA participation of older adults when they were compared in relation to education attainment and living status. More specifically, those with college education (and higher) and those living in their private houses reported significantly higher PA levels. Similarly, significant main effects for PA outcomes were found by education and living status in the regression analysis. Previous findings on older adults have consistently explained that participation in PA is lower among those who are less educated [32,33,34,35]. Besides, the type of residency and local culture might be associated with PA behavior. Walking has been reported to be less among those living in a rented accommodation than that of home owners [36]. However, it was shown that the risk of inactive lifestyle was 1.47 times higher for women living alone [37]. Therefore, further research seems to be needed to better explain the relationship between type of residency and its culture with PA.

It is well established that of the chief PA barriers across old ages are health problems and associated fear of injury/falling. Mathews et al. (2010) investigated the perceived PA in a multicultural older adults population and found health problems, fear of falling, and inconvenience as common barriers [38]. Additionally, our findings demonstrated that when “fear of injury” is perceived as a critical barrier, younger older adults would participate in PA more than their older peers. In agreement with these results, a review study highlighted that when studying PA of oldest old it is necessary to focus on a number of factors among which fear of injury receives an especial attention [39]. Also, “lack of skill” moderated the effect of age for PA and when it is perceived as a barrier, older adults would demonstrate lower PA participation. This evidence together with previously mentioned findings in comparison of age classifications necessitate the implementation of PA policies and priorities specifically established for older adults, considering their fears, needs, and preferences.

Although no significant effect of gender for PA outcome was found, there was a moderation by “fear of injury” and when it is thought to be a potential barrier, women would participate in PA less than their male counterparts. This finding is consistent with the results of Gobbi et al. (2012) among Brazilian older adults, reporting the “fear of falling and being hurt” as the most important barrier among women [40]. Similarly, through a research on Icelandic population, low self-efficacy and fear of falling have been shown to be significant limiting factors for many older women and suppressed their PA participation [41]. Gobbi et al. (2012) also explained that women reported significantly more barriers than men in all 60–69, 70–79, and ≥ 80 year age groups, even though it was marginally significant among the last two categories [40]. Therefore, the evidence on older adults from different geographical locations and cultures shows that women commonly face further perceived limiting factors when compared to men and fear of injury/falling is amongst the most significant barriers.

Previous research supports the benefits of marriage on health and the correlation between being married and health-related behaviors, and shows that spousal concordance in health behaviors extends into later life [14]. However, the evidence on PA remains to be inconsistent. The majority of studies indicated that marriage is inversely correlated with PA [7]. On the other hand, one study opposes these findings by showing that getting married is positively associated with PA [42]. So, this lack of agreement is likely to be moderated by other factors than age and gender. In the present research, “lack of time” was the single moderator for PA outcomes and among those perceived that “lack of time” is not a PA barrier for them, higher PA levels were shown by singles. Therefore, we are still unable to draw conclusions on this moderation and further research needs to be elucidated.

The effect of employment for the PA outcomes was moderated by four subscales of BBAQ. Among those reported that “lack of willpower”, “fear of injury”, and “lack of skill” are not PA barriers, employed older adults reported greater PA participation. Similar results were seen when “lack of resources” was a barrier. The results of the present study on the moderation of perceived PA barriers on the effect of employment for PA outcomes appears to be controversial and we are not able to draw any solid conclusion on. One possible explanation, especially for “lack of resources”, would be the few numbers of employed participants at the age of 60 and over, as seen in the present study (by only ~ 15%). In this regard, Sawchuk et al. (2011) examined the reliability of BBAQ among older adults. They reported Cronbach’s alpha as 0.87 for total scale. Nevertheless, it was shown to be 0.45 for “lack of resources” [43]. So, this would have affected the results. Altogether, there seem to be moderation effects by perceived PA barriers on PA.

Regarding education attainment and living status, there was moderation by each of the BBAQ subscales, generally being consistent with the main effects of the education and living status. The effect of education for PA involvement would be explained through several assumptions. First, more educated participants would have further knowledge on the advantages of PA. Second, this people may represent better socioeconomic status, and then, better access to resources for being physically active. Lastly, more educated people may be exposed to more PA programs through educational system [3]. That could be the case for those living in their own private houses and these individuals may represent a higher socioeconomic stratum. Besides, they would generally possess a better physical health compared to those living in other places including nursing homes, which in turn, could significantly affect older adults’ participation in PA programs.

The present research had several strengths and limitations. First, this is the first study to investigate the moderation of perceived barriers of PA on the correlation between SED characteristics and PA in older population. Besides, we took barriers from various domains using a validated questionnaire on PA of older adults. Therefore, this multifaceted approach helps us better understand the perceived PA barriers among older adults and the moderation they might have on the effect of SED characteristics for PA participation. As the limitations, due to funding restrictions, it was not possible to include larger population and we collected data only from the minimum sample size suggested by previous studies. While, a bigger sample size commonly results in more confident findings. Another limitation was the lack of calculation of the inter-reliability of the examiners.

Additionally, as a result of low response fraction, we were unable to report income which is of the most important factors in studies on SED characteristics and this could have unraveled further on the effects and relationships investigated in this study. Lastly, we did not account for the possible consequences of running multiple comparisons. To avoid rejection of null hypothesis too readily, adjustments for multiple comparisons are recommended. These adjustments, such as alpha error corrections, reduce Type I error for null associations but they elevate Type II error, as an overconservative approach is considered and in these cases true relationships might be rejected. As we cannot argue that the first explanation for the significant relationships found in the present study is “chance” and we followed a theoretical framework for our hypotheses, correction of alpha is not desirable [44]. However, due to possibility of Type I error inflation in multiple comparisons, especially regarding moderation effects, further studies need to be conducted to draw solid conclusions.

留言 (0)

沒有登入
gif