Patterns of response by sociodemographic characteristics and recruitment methods for women in UK population surveys and cohort studies

Abstract

Women are an important public health focus, because they are more likely to experience some social determinants of disease, and they influence family health. Little research has explored the sociodemographic representativeness of women in research studies. We examined the representativeness of female respondents across four sociodemographic factors in UK population surveys and cohort studies. Six UK population-based health surveys (from 2009-2013) and eight Medical Research Council cohort studies (from 1991 to 2014) were included. Percentages of women respondents by age, income/occupation, education status, and ethnicity were compared against contemporary population estimates. Women aged <35 years were under-represented. The oldest women were under-represented in four of nine studies. Within income/occupation, at the highest deprivation level, the range was 4 percent under-representation to 43 percent over-representation; at the lowest level, it was 6 percent under-representation to 21 percent over representation. Of nine studies reporting educational level, four under-represented women without school qualifications, and three under-represented women with degrees. One of five studies over-represented non-white groups and under-represented white women (by 9 percent). Response patterns varied by topic and recruitment and data collection methods. Future research should focus upon the methods used to identify, reach, and engage women to improve representativeness in studies addressing health behaviors.

Non-response in a national health survey in Germany: An intersectionality-informed multilevel analysis of individual heterogeneity and discriminatory accuracy

“…However, it is feasible to draw a summary of dimensions of social location that have been associated with study participation across different contexts and study designs. In population-based surveys, ethnic minority status, low income, low level of education and young and old age were associated with lower response proportions [14][15][16]. Furthermore, male sex, low socioeconomic status, being an unskilled worker, having no children and being unmarried were associated with low study participation in cross-sectional and cohort designs [13,[17][18][19].…”

Investigating people’s attitudes towards participating in longitudinal health research: an intersectionality-informed perspective

“…Participants in longitudinal research are more likely to be educated to at least undergraduate level [ 11 25 ]. In cohort studies, women aged 75 and older and women aged 35 and younger are often underrepresented groups amongst women [ 35 ], and the influence of a steady partnership on response proportions differs by sex/gender [ 66 ]. However, most analyses of social heterogeneity have either exclusively focused on specific subgroups [ 61 67 4 37 42 ] or have been quantitative and descriptive in nature, missing to engage the complex and situated nature of study participation (also see [ 29 ].…”

Lifestage differences in young UK women’s reasons for research participation

“…Young adulthood is a valuable stage to focus on lifestyle behaviour change since this is when young women may be leaving parental homes and making independent lifestyle decisions (Barker et al, 2018). Yet, young women are less likely than older women to take advantage of health-related activities, such as cervical screening, or to participate in health research to identify needs, target interventions and inform services (Howcutt et al, 2017a;NHS Digital Screening and Immunisations Team, 2017).…”