Sedentary behaviour, musculoskeletal pain conditions and type 2 diabetes

Sedentary behaviour (too much sitting, as distinct from too little physical activity or exercise) is of rising public health concern. It has been associated with increased risks of multiple chronic diseases, including cardiovascular conditions, metabolic disorders such as type 2 diabetes (T2D), and all-cause mortality. Also, there is growing evidence of potential risk associations with musculoskeletal pain (MSP) conditions. Importantly, MSP conditions have emerged as common comorbidities in people living with cardiometabolic conditions, especially so in those living with T2D.

The co-occurrence of excessive sedentary behaviour, T2D and MSP conditions, which is much more common in older adults is concerning. MSP conditions can be a barrier to regular physical activity participation in adults. An adequate level of moderate-to-vigorous intensity physical activity (MVPA) has been a cornerstone in the management of T2D and has also been known to be beneficial for pain management. The coexistence of T2D and MSP conditions may render many adults physically inactive and vulnerable to engaging in prolonged periods of sitting during waking hours, due at least in part to functional impairment and pain. Consequently, being physically inactive and engaging in excessive sedentary behaviour may have further detrimental impacts on both T2D and MSP conditions. Currently, the coexistence of MSP conditions and T2D in adults and the potential relationships with sedentary behaviour have been largely unexplored.

This thesis, therefore, aimed to explore the evidence on sedentary behaviour, MSP conditions, and T2D with the broad aim of understanding the associations of sedentary behaviour with pain related to musculoskeletal systems in adults and whether such potential relationships differ in those living with and without T2D.

To address this aim, an existing prospective dataset from an epidemiological study, the Australian Diabetes, Obesity and Lifestyle Study (AusDiab) and a cluster-randomised control trial, the Stand-Up Victoria Study, as well as a cross-sectional dataset from the Maastricht Study were analysed. First, a systematic review (Study 1), was conducted using the standard Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines with the purpose of identifying the existing evidence on associations of sedentary behaviour in occupational and non-occupational settings with MSP conditions, and to identify knowledge gaps to inform the thesis’ empirical studies. This review (Study 1) found evidence of cross-sectional associations of both occupational and non-occupational sedentary behaviour with MSP conditions, with the associations in the occupational domain being dependent on the nature and the physical demand of the occupation. Evidence on prospective associations was inconclusive; however, there was a probable indication of a protective association of sedentary behaviour (device-measured) with some MSP conditions in tradespeople. Additionally, reducing desk-based (office) workers’ sitting time was observed to be correlated with reduced MSP conditions or discomfort. Also, the review identified a lack of a sufficient number of device-measured sedentary behaviour-based studies as well as prospective studies as key literature gaps.

Secondly, three empirical studies informed by the knowledge gaps from the systematic review were conducted. Study 2 (Maastricht Study dataset – data of 2827 participants were analysed): In this study, logistic regression and restricted cubic spline statistical methods were utilised to examine the linear and non-linear cross-sectional associations of device-measured daily sitting time with MSP outcomes in adults with normal glucose metabolism (NGM), prediabetes, and T2D. Evidence of a cross-sectional association was observed between device-measured daily sitting time and knee pain (in a linear function). The association was only significant in those with T2D but not in those with prediabetes or NGM. No significant associations were observed for neck, shoulder, or low back pain; however, the relationships appeared to be curvilinear but statistically non-significant.

Study 3 (AusDiab dataset – 4099 participants’ data were analysed): This study utilised a multilevel growth curve modelling to examine the prospective relationships of the common leisure-time sedentary behaviour, television-viewing (TV) time with bodily pain in adults with and without T2D over 12 years. The findings showed that bodily pain severity increases with age, and an increased volume of TV time at any given time point was significantly associated with increased bodily pain severity. The observed relationship was more pronounced in those with T2D than those without. In reference to those with NGM, the effect of T2D and prediabetes on bodily pain severity increased with increasing TV time, significantly so only in those with T2D when the TV time threshold increased above 2.5 hours per day.

Study 4 (Stand Up Victoria Study dataset – pooled data of 224 participants were analysed): Using compositional data analysis framework, prospective relationships with changes in multisite MSP of changes in desk-based workers' sitting, standing, and stepping, as well as the short-bouts and long-bouts of these behaviours at three- and 12-months were examined. Further, compositional isotemporal substitution modelling was performed to examine the impact of reallocating time among these behaviours on MSP outcomes. The findings demonstrated that in the short term (at three months) increased standing relative to changes in stepping and sitting composition was significantly associated with increased multisite MSP outcomes, and increased stepping relative to changes in sitting and standing was significantly associated with reduced multisite MSP outcomes. Reduced sitting relative to changes in standing and stepping was not significantly associated with multisite MSP changes at three months. Further, no significant associations were observed for changes in short-bouts relative to long-bouts of these behaviours with the MSP outcomes. In the longer term (at 12 months), there were no significant associations observed for the relationships. Noteworthy, increased standing appeared not to worsen multisite MSP outcomes in the long term. Additionally, reallocating time from sitting at baseline to standing or stepping at follow-ups with the other behaviour held constant at the mean could favourably impact multisite MSP outcomes. Likewise, favourably reallocating time from baseline to follow-ups between the short and long bouts of a given behaviour while volumes of time spent in other behaviours are kept constant may have beneficial impacts on multisite MSP outcomes, especially in the longer term at 12 months.

In summary, the findings indicate that there is evidence of cross-sectional associations of sedentary behaviour with MSP conditions. The cross-sectional evidence appears stronger for knee pain, with evidence observed for both self-reported and device-measured sedentary behaviour, an association which seems to be driven mainly by the presence of T2D. Furthermore, the thesis found evidence of a prospective association of increased sedentary behaviour with increased bodily pain severity, a relationship which was more pronounced in those with T2D than those without. Additionally, reducing desk-based workers’ sitting by increasing standing and stepping, would unlikely have adverse impacts on MSP outcomes, especially in the long term. These findings provide some implications for practice and future research in this context. They could also help to inform future work directed at developing an improved understanding of the potential biological mechanisms of sedentary behaviour’s role in T2D/MSP conditions relationships in adults.