Social inequalities in physical activity Implications of research practices – the case of studying breast cancer survivors

While social inequalities in health persist, a growing field of physical activity research continues documenting the positive effects of physical activity on mental and physical health. Hence, studies have suggested that there might be an association between physical activity and socioeconomic status. However, public physical activity interventions targeting low socioeconomic groups have shown limited effects, and the associations between physical activity and socioeconomic status seem more complex than previously assumed. Besides, recent research has given reason to believe that there may be practices within the knowledge production in physical activity research as such, that may cause incomplete knowledge about how these variables relate. The present thesis illuminates the implications of methodological practices within physical activity research as explanations of the reported differences in physical activity between socioeconomic groups. The main question is whether conventional methodological choices affect the reported physical activity in socioeconomic groups differently. In addition to traditional analyses of socioeconomic differences in aspects of physical activity, Papers I-III study the impact of applying various conceptualisations of physical activity, the impact of variations across instruments for collecting physical activity data, and the generalisability of a controlled physical activity intervention across socioeconomic groups, respectively. The three studies included in the thesis employ different designs. The first paper is a systematic article analysis of 56 previously published papers examining the relationships between physical activity and socioeconomic status, aiming to determine whether findings are linked only to the activity domain investigated, and perhaps not total physical activity. Paper II has a cross-sectional design, applying questionnaires, activity logs, and interviews among 52 breast cancer survivors to uncover possible limitations in standard physical activity questionnaires. In Paper III, we analyse adherence rates in a physical activity treatment intervention among 47 newly diagnosed breast cancer patients to identify socioeconomically (among other) predictors. The findings suggest that there is evidence for a positive relationship between physical activity and socioeconomic status only for leisure-time physical activity. The positive relationship between the mentioned variables is absent, or negative, in other physical activity domains - especially within the occupational physical activity. In the study of long-term breast cancer survivors, alternative methods assessing physical activity, particularly time-geographic activity logs, reveal socioeconomic differences in experiences from-, and routines for-, physical activity, without there necessarily being different amounts of total physical activity. In the study of newly diagnosed breast cancer survivors, we found a higher income among those who completed the physical activity intervention comparing to those who withdrew, although there was no significant association between socioeconomic status and adherence rate. A lack of socioeconomic sample representativeness may, in part, explain the lack of associations between adherence rates and socioeconomic status in our material. The scientific understanding of the relationship between physical activity and socioeconomic status seems to be largely based on exercise and leisure-time physical activity data. The use of traditional questionnaires yields insufficient information about differences in physical activity between socioeconomic groups. It is unclear whether exercise interventions in breast cancer treatment are equally suitable to all socioeconomic groups; however, they may risk excluding lower socioeconomic groups and thus reduce the external validity