Moderating Effects of Socioeconomic Status in a Randomised Controlled Trial of a Brief Planning Intervention to Promote Physical Activity

Abstract

Background: To address socioeconomic health gaps, there has been a call for tailored behaviour change interventions that meet the specific needs of their targeted audience. However, socioeconomic status is typically operationalised as a single measure such as education, which disregards other indicators and aspects of socioeconomic status, and little is known about the most potent drivers of intervention-generated inequalities. This study aimed to explore the differential effects of socioeconomic status measures of income, education, occupation and IMD area deprivation upon the operation of an established brief behaviour change intervention that has been shown to increase physical activity among manual workers. To achieve this aim, the magnitude of correlations among these socioeconomic status measures was explored and their potential moderating effects upon the effectiveness of the intervention were investigated. Method: A two-armed randomised controlled design was used, with participants (n = 98) assigned to either the intervention (n = 47) or control condition (n = 51). Participants in each condition were given identical physical activity volitional help sheets, the only difference being that individuals in the intervention were encouraged to make “if-then” plans by linking challenges and solutions, whereas those in the control condition were not asked to make “if-then” plans. Individual socioeconomic status measures of income, education and occupation, and IMD area-level deprivation were taken for each participant. The primary outcome measure was leisure centre attendance, which was measured objectively over 12-months using electronic entry swipes. Results: To enable moderator analyses, categorical variables were created distinguishing between high versus low-socioeconomic status participants on socioeconomic status measures of IMD area-deprivation, income, occupation and education. While mean 12-month leisure centre attendance was marginally higher among participants in the intervention condition (M = 20.00, SD = 24.08) compared to those in the control condition (M = 12.55, SD = 16.36), an independent-samples t-test showed that there was no significant statistical difference between the conditions (t (96) = -1.78, p = .079, d = -.36). Two-way, between-group ANOVA’s assessed the main and interaction effects of high versus low socioeconomic status measure of income, education, NS-SEC occupation and IMD-area deprivation upon leisure centre attendance between both conditions. No significant main effect of income, IMD deprivation and occupation status on leisure centre was found, but borderline significant effects of condition on leisure centre was found for income and IMD deprivation (F = (1, 73) = .72, p = .400, d = .2; F = (1, 83) = 4.03, p = .048, d = .14 and (F = (1, 94) = 4.27, p = .041, d = .42). In addition, a main effect of occupation status on leisure centre attendance (F = 1, 94) = 5.09, p = .026, d = .46). Conclusion: The findings highlight that individual- and area-level measures of socioeconomic status are not interchangeable and that leisure centre attendance related research outcomes are predicted by occupation and not income, education or IMD area-deprivation. This means that future research in this area should avoid using a single measure of socioeconomic status to operationalise socioeconomic status in health research.

Date of Award	31 Dec 2020
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Christopher Arden Armitage (Supervisor) & Laura Brown (Supervisor)