After all, do findings from clinical physical activity interventions among breast cancer patients apply to all socioeconomic groups? A systematic review of representativeness

Abstract

Abstract Background Considering the widening social health inequalities, socioeconomic representativeness in clinical trials is essential to analyses of feasibility and validity relating to socioeconomic groups. Physical activity trials on breast cancer patients are particularly relevant because the relationship between socioeconomic status and disease burden reverses after diagnosis and because low socioeconomic groups appear less inclined initially to engage in the type of activity commonly prescribed in such trials. A step-by-step exclusion process in which low-socioeconomic groups are less likely to register if at all included, adhere well, and to complete the intervention, could reinforce health inequality. Hence, our aim was to a) examine representativeness in physical activity trials on breast cancer patients, b) investigate the association between socioeconomic status and physical activity intervention adherence, and c) explore associations between representativeness and the relationship between socioeconomic status and intervention adherence. Methods Systematic, computerized searches for articles reporting physical activity trials in breast cancer patients were performed in PubMed, CINAHL, AMED, EMBASE and PsycINFO. Additional citation-based searches resulted in 37 articles. In articles where education, marital status or ethnicity were used as indicators of socioeconomic status, we compared the distribution of socioeconomic groups in the study sample and the country where the study was conducted. Through such proxy calculations, we estimated the studies' socioeconomic representativeness. Results Of the 30 studies that applied education as an indicator of socioeconomic status, 23 favoured high-socioeconomic groups; Of the 18 American studies where ethnicity was the socioeconomic indicator, 15 favoured white participants. The difference between study samples and country population was less pronounced in studies where marital status was the socioeconomic indicator. Only six studies reported associations between socioeconomic status and trial adherence, hampering solid conclusions on this relationship and possible associations between estimated representativeness and a socioeconomic status and trial adherence relationship. Conclusions Non whites and patients with low education are underrepresented in physical activity trials on breast cancer patients. Using leisure-time physical activity in breast cancer treatment may not be suitable for all groups. Hence, the results from most of these studies may be considered inequitable with regard to different socioeconomic status.