Improving physical activity for health among inactive adults aged 50 years and older: development, design and delivery of a community-based intervention

As adults get older, they acquire more chronic health conditions, but engage in less  physical activity. Most older adults do not meet the minimum physical activity guidelines  for health. Community-based interventions can address this, but evidence is needed to  understand how they work, and for associations between accelerometer-measured  physical activity and chronic health conditions and healthcare utilisation. The objectives  of this thesis are to: inform the recruitment to, and development of an intervention to  increase physical activity in adults aged 50 years and older – Move for Life; to design a  protocol to test the feasibility of the intervention; to investigate feasibility measures; and  to test the hypothesis that physical activity is related to chronic health conditions and  healthcare use. 

Methods Paper 1 was a qualitative study, involving semi-structured interviews (n=18) and four  focus groups (n=29) with older adults, advocates and professionals; data were analysed  thematically. The purpose of paper 1 was to inform the intervention development and,  specifically, to optimise intervention recruitment, sustainability and scalability. Paper 2  outlined a protocol for a three-arm cluster pilot randomised controlled trial. The setting  was eight Local Sports Partnership (LSP) hubs; each hub was a unit of randomisation (cluster), and individuals were the units of analysis (participants). The hubs were  randomised: true control, usual programme or Move for Life intervention. The true control group was given information about physical activity but was not included in a  programme; the intervention arm involved augmentation of existing physical activity classes with the Move for Life intervention; and usual care groups had physical activity  classes delivered as normal. The recruitment target was 576 participants. Data were  collected at baseline (T0), upon completion of programmes (T1) and at three-month follow-up (T2). Data collected included: demographic data, information on chronic health  conditions, healthcare utilisation data, objective physical parameters, e.g., body mass  index and grip strength. All participants were asked to wear a device on the thigh for  measuring physical activity and sedentary behaviour (activPAL). The third paper used baseline data generated from a sample of adults recruited to the trial, that provided valid  activPAL data (n=485). Hierarchical cluster analysis was conducted using the  accelerometer-measured physical activity variables. Descriptive statistics were used to  investigate associations with chronic conditions and healthcare utilisation. Paper 4  provides a description of the intervention development, using the intervention mapping protocol. The process of intervention development involved: input from key stakeholders,  evidence from the published literature, behavioural change theory, and the experience of  an interdisciplinary team, who understood the context and environment of the setting.  

Results Paper 1 reported on factors for successful recruitment and sustainability: Data analysis  produced three overarching themes. “Age appropriate” explains how communication and  the environment should be adapted to the needs of adults aged 50 years and older.  “Culture and connection” refer to the interplay of individual and social factors that  influence participation, including individual fears and insecurities, group cohesion and  added value beyond the physical gains in these programmes. “Roles and partnerships” outlines how key collaborations may be identified and managed and how local ownership  is key to success and scalability.   Paper 2 outlined a protocol for a three-arm feasibility RCT. Of the 733 participants who  were recruited, 531 were given an accelerometer at T0, 485 of which provided valid data. Of these, 383 were eligible as per the study protocol, i.e., they were aged 50 years and  over and were insufficiently active. Eighty-nine of them (23%) were lost to attrition and  a further 57 (15%) did not provide valid data at T2. Of the 383 eligible participants  providing reliable data at T0, 237 provided reliable data again at T1; the retention rate at  T1 was 61.9%. Of the 485 participants who provided valid activPAL data, 381 (78.6%)  were female, and 382 (80.6%) had private health insurance. In paper 3, four distinct physical activity behaviour profiles were identified: inactive?sedentary (n = 50, 10.3%), low activity (n = 295, 60.8%), active (n = 111, 22.9%) and  very active (n = 29, 6%). The inactive-sedentary cluster had the highest prevalence of  chronic illnesses, in particular, mental illness (p = 0.006) and chronic lung disease  (p = 0.032), as well as multi-morbidity, complex multi-morbidity and healthcare  utilisation. The prevalence of any practice nurse visit (p = 0.033), outpatient attendances  (p = 0.04) and hospital admission (p = 0.034) were higher in less active clusters. The  results have provided an insight into how physical activity behaviour is associated with  chronic illness and healthcare utilisation. A group within the group has been identified  that is more likely to be unwell. Provisions need to be made to reduce barriers for  participation in physical activity for adults with complex multi-morbidity and very low  physical activity. 

Discussion The intervention was designed to fit within existing group-based structured physical  activity programmes run by LSPs, thus maximising the likelihood of translation into  policy and practice. Feasibility outcomes, including recruitment strategies, programme  attendance, attrition and acceptability, are reported. Limitations of the study include the  high proportion of females and participants with private health insurance, as well as the  high baseline levels of physical activity recorded. The thesis outlines recommendations  for a full trial including: an ongoing process of engagement with men and harder to reach  groups to encourage recruitment; engagement with ‘harder to keep’ participants identified  in this study, including those with multi-morbidity and mental health diagnoses; data  collection on income and ethnicity; and approaching target communities, with a view to  active participation, earlier in the planning stage. Trial registration number: ISRCTN11235176