Scottish Health Survey - topic report: physical activity

6. Discussion and Conclusion

There is extensive evidence that physical activity contributes to health, as well as extensive evidence of persistent inequalities in health. Differentials in physical activity matter then, because they have the potential to contribute to alleviation of these persistent inequalities in health.

Logistic regression is helpful in indicating what factors are associated with and may predict meeting physical activity guidelines, i.e. the patterning of physical activity. However, as noted earlier, a limitation of this study is only being able to compare those factors for which there are variables present in the Scottish Health Survey. The factors discussed here can only explain part of the variation in physical activity. It is also important to remind ourselves that many of the factors examined in the models for this study are likely to have bidirectional relationships with physical activity. For example, having a BMI of >30+ (obese) was found to be associated with a lower likelihood of meeting physical activity guidelines. Equally, it could be said that not meeting physical activity guidelines is associated with being obese. No clear direction of causality can be claimed. The study of correlates, however, is regarded as important despite these limitations in helping to develop and improve interventions and identify priority target population groups, although this type if research is just one part of a wider evidence base that should be taken into consideration.

The analysis on the 2012 SHeS found that, for all adults, demographic and health and lifestyle factors were of primary importance with meeting the 2012 physical activity guidelines, whereas socioeconomic factors were overall less influential. The exception to this was economic activity status, which showed a relatively strong influence on likelihood of meeting the guidelines. Those reporting being unable to work were much less likely to achieve recommended levels of physical activity compared to those in paid work, even after having controlled for disability or health status.

While it is clear the physical activity guideline change in 2011 (from 30 minutes of moderate physical activity on five or more days per week to accumulation of 150 minutes of moderate physical activity per week, with no stipulation on frequency) had a substantial effect on the proportion of the population who meet the guidelines, the analyses in this study suggest there was not such a dramatic effect on the distribution of characteristics of those most likely to meet the physical activity guidelines. Overall, very little difference was observed in the demographic, socioeconomic and health and lifestyle patterning of those in 2012 who met the new guidelines compared to respondents to the 2011 survey meeting the old guidelines. A very similar range of the most influential factors were observed for 2011: economic activity status, health, disability, BMI, age and sex. However, it is important to note that this study only examined the patterning in relation to this one aspect of the guidelines. Further research is required that investigates other areas of the guidelines, such as the patterning of sedentary behaviour; of those who are very low active; vigorous activity, muscle strengthening activity and looking at the different age groups.

It is recognised that there are multiple domains that contribute to accumulation of total physical activity, however, research has tended to focus primarily on leisure time physical activity. This is likely to be most representative of a combination of sport and exercise and walking for leisure. Focus on leisure time physical activity and/or total physical activity may be obscuring differences in associated characteristics for different types of physical activity. The factors that contribute to each different domain and how they compare to each other is under-researched.

The results from this study do suggest that focussing on the patterning of total physical activity can indeed disguise very different patterning in the different domains of activity that make up total physical activity, For example, a strong association of socioeconomic factors and lack of gender association was observed in relation to occupational physical activity, whereas gender is a key factor for total physical activity and socioeconomic factors were found to be less influential. Also, a reversal of the gender association was observed with walking and housework physical activity, where women were found to be more likely than men to participate. This contrasts with total physical activity or sport and exercise where men are found to be more likely to participate.

Walking was also found to increase with age in women, especially age 45 to 74 years. Age was not associated at all with walking in men and, for women, no socioeconomic factors were associated with walking. These findings add to the building evidence on the importance of walking as a leveller of inequalities in participation in physical activity.

It is important to note, that domain specific physical activity was taken to be any participation in activity in that domain. This differed from total physical activity which was calculated as a certain quantity of participation. As such, the patterning of total physical activity is not directly comparable to that of domain specific physical activity. This is an area which could benefit from future research.

Age, gender and poor health remain key factors associated with physical activity, in common with other research. Bauman et al. (2012) conclude that poor health is a determinant of physical activity, not just a correlate. This will inevitably be challenging to address, although evidence is promising for a number of primary care based approaches to promoting physical activity in those with health problems.

Despite the finding that more women participate in walking and housework, when total physical activity is calculated, the participation in these activities is not enough to counteract the overall gender effect, suggesting more action needs to be taken to encourage greater activity levels in women.

The bivariate analysis shows a decrease in meeting the physical activity recommendations with age. Recognition of the decrease in physical activity with age has tended to manifest itself as initiatives aimed at those age 60+, yet once other factors such as health and income have been controlled for, the relationship with age was found to be more complex. The likelihood of participating in walking tended to increase with age in women till age 75+. There was found to be an increase in likelihood of meeting the new physical activity guidelines in both men and women up to about age 45 years (after a sharp decrease for men between ages 16-24 and 25-34) and then it declines. There is undoubtedly a marked difference between the youngest and oldest age categories, however, the lack of clear patterning in the other age groups in total physical activity and in the physical activity domains suggests more research to clarify the influence of age is warranted.

The socioeconomic patterning of physical activity is often cited as an important consideration, however, it is not always clear how much of an impact it has. Gidlow et al. (2006), in a systematic review of the relationship between socioeconomic factors and physical activity, concluded there was consistent (if weak) evidence for a positive association between higher levels of leisure time or moderate to vigorous intensity physical activity (MVPA) in those with higher socioeconomic status. The results from this study support this association and add to this evidence. Gidlow et al. however, found that the evidence was less convincing for a relationship between total physical activity (also referred to as general or habitual physical activity) and socioeconomic measures. The authors highlighted how the vast majority of the studies they encountered used self-report of physical activity and how this has been criticised for inaccurate capture of habitual/total physical activity with a bias towards recall of structured or more intense physical activity typically associated with sport and exercise and leisure time physical activity (LTPA). Thus any relationships that have been shown may actually be reflecting the relationship to LTPA or MVPA. Also, there has been limited account taken of the differences that may exist in patterning by domains of physical activity. This study has shown that there are differences in socioeconomic patterning across the different domains which may explain some of the inconsistencies encountered in the evidence base. There is an argument that higher participation in some domains of physical activity associated with lower socioeconomic groups, notably occupational physical activity, counteracts domains where higher participation rates are associated with higher socioeconomic status, such as the sport and exercise domain. This serves to weaken the relationship between socioeconomic status and meeting physical activity guidelines when looking at overall physical activity and may explain the weaker association of socioeconomic factors with meeting physical activity guidelines found in this study. This is an area that requires further research to establish more clearly the influence of socioeconomic factors and would benefit from greater use of objective measures that are better able to capture habitual/total physical activity.

Despite the uncertainty regarding the level of influence of socioeconomic factors in overall physical activity, it is clear there are inequalities in participation in certain domains of physical activity that have important implications. Although those with lower socioeconomic status are shown here to be more likely to participate in occupation physical activity, physical activity associated with occupations has been decreasing with the increase in technology and de-industrialisation over the years^,. This means that occupational physical activity will inevitably make up less and less of total physical activity and action will need to be taken to ensure higher participation of those from lower socioeconomic groups in other domains of physical activity.

A further consideration is that many of the health and lifestyle factors found here to have a negative relationship with achieving physical activity recommendations have been shown to be more likely to be present in those in lower socioeconomic groups, for example obesity, smoking, poor mental and physical health^,. This reinforces the message that socioeconomic factors are important within a broader perspective.

Conclusion

Our analysis has demonstrated that in 2012 multiple demographic and health factors had the strongest association with meeting the new physical activity guidelines, alongside economic activity status. The change to the physical activity guidelines in 2011 has had little substantial effect on the overall patterning of who is most likely to meet the guidelines, though the range of significant factors associated with meeting the new guidelines increased. Differential patterning by physical activity domain is important to the development of more targeted approaches to promote physical activity and the analysis highlighted how recreational physical activity is more socially patterned, compared to total physical activity. This may be a growing issue in future if longer term trends continue of decline in the manual labour sector and decreasing housework activity due to labour saving devices. The long term implication is that leisure physical activity, both sport and exercise as well as walking and other non-sport physical activity, becomes more important to total physical activity. Finally, this study adds further evidence to the importance of walking in addressing inequalities in physical activity participation and highlights the importance on the recent Let's Get Scotland Walking - The National Walking Strategy and the work that continues to implement this at national and local level.