An integrated intervention model to improve behavioural change in inactive members of fitness centres

Background: The physical inactivity (PIA) prevalence in adults is not improving. Fitness centres (fitness centres, leisure centres, gyms, and any other semi- or non-structured PA providers) are part of the solution as they are a major contributor to the built environment and are used by thousands of adults to start or maintain a physical activity (PA) behaviour. However, it remains unclear how to effectively promote and maintain PA in these populations. To promote PA at fitness centres issues like the low attendance rate (1.1 to 5.6 times/month), the high dropout levels (>60% within the first 3 months), and the difficulties for monitoring PA need to be addressed. Attendance is a more feasible behaviour to target and monitor, and might increase the members' PA levels, but research in this regard is scarce. Behaviour change models like the COM-B Model might help to identify the barriers for attendance, while the Behaviour Change Wheel (BCW) framework might work to inform the developing of the intervention targeting attendance. Therefore, the aims of this thesis are to explore the PA participation at fitness centres, to identify the main barriers and determinants for attendance fitness centres and to develop a behaviour change intervention according to the BCW to target attendance rate at fitness centres. Methods: Three empirical chapters (Chapters 3 to 5) were conducted. Chapter 3 examined the PA levels from a database with 3,627 members of Spanish fitness centres and the PA levels reported in the Eurobarometer 472 for the Spanish population (n=1,002). This comparison was performed considering factors like the age (18-29 years; 30-44 years; 45-59 years; 60-69 years; ≥70 years), gender (men or women), or the origin of the performed weekly PA (i.e., waking-PA, moderate-PA, or vigorous-PA). The PA levels were analysed using the International Activity Questionnaire short version (IPAQ-SF), and respondents were grouped into Low-PA (Inactive), Moderate-PA, and High-PA according to the criteria set for the IPAQ-SF. Chapter 4 designed and validated a questionnaire for measuring factors influencing attendance at fitness centres using the COM-B Model. Three phases were performed: i) content generation of the questionnaire items through literature searches and focus groups, ii) expert review of the relevance and clarity of the items to determine sufficient construct validity, iii) assessment of content validity and internal reliability through the administration of an online survey to 180 participants from British fitness centres and 430 participants from Spanish fitness centres Chapter 5 developed a general intervention to target attendance at fitness centres. The BCW was used as a framework to inform the intervention development. An iterative approach of four iterations were performed to develop the intervention and the Theoretical Domains Framework (TDF) was used to underpinning the mechanisms of change. The questionnaire developed in chapter 4 was used to inform the main determinants and barriers for increase attendance to fitness centres members. Results: Fitness centres' members showed lower prevalence of PIA and a higher prevalence of High PA than the general population regardless the age and gender (p < 0.05). However, women were less active than men in both studied datasets (p < 0.05). Low-PA group and Moderate-PA group from the Eurobarometer dataset reported >70% of their total MET-minutes/week from walking-PA whilst those fitness centre's members reported significant higher engagement in vigorous-PA. Chapter 4 shows a final version of the questionnaire with 35-items approved in the expert review with acceptable construct validity (e.g., Aikens' V for relevance and clarity ≥ 0.97). However, this questionnaire failed the content validity and internal reliability analysis. A shorter version of 17-item questionnaire showed acceptable validity and internal reliability (e.g., confirmatory factorial analysis: CMIN/DF = 204; CMIN = 175.497; DF = 86; CFI = 0.965; NFI = 0.919; RMSEA = 0.50; PClose = 0.457). Chapter 5 identified low-attendance members (< 2 access/week) that reports low-PA levels as the members who might benefit the most from increasing their attendance rate. On the other hand, each item from the questionnaire developed in Chapter 4 was linked to the most promising intervention functions and behaviour change techniques. Furthermore, a mode of delivery for these behaviour change techniques and an example of how to do it was identified. Discussion and conclusion: Fitness centres may play a key role to address PIA as they engage most of their members in regular PA, including women and older adults. Also, members seem to be more active and perform more vigorous-PA than the general population. However, gender and age differences exist in this population, therefore, attention to these factors should be made. Attendance might be an effective way to address PIA at fitness centres. The 17-items questionnaire developed in Chapter 4 has acceptable reliability and validity, so it might be used to develop future interventions targeting members' attendance. Chapter 5 describes the first structured intervention model to enhance attendance rate in low-attendance-and-inactive members of fitness centres. Further research on the testing of the intervention model defined in Chapter 5 in a fitness centre setting could provide a structure and evidenced-based intervention which can be implemented at scale.