Response rates and representativeness from birth to young adulthood in a Swedish birth cohort

In birth-cohort studies, response rates generally decline over time. We aimed to evaluate representativeness and identify predictors of retention up to young adulthood in a population-based birth cohort originally focused on asthma and allergies.

The Barn/Child, Allergy, Milieu, Stockholm, Epidemiology (BAMSE) birth cohort was established in 1994-6 in Stockholm, Sweden. We utilized data from questionnaires and Swedish registers to compare the characteristics of participants at age 24 years (n = 3064) with those of the original cohort (n = 4089) and the general population born in 1994-6 in Stockholm county (n = 68 378) and Sweden (n = 310 976). Logistic regression was used to identify predictors of retention at age 24 years.

The response rate declined over time and was higher in females than in males in adolescence and young adulthood (80.0% versus 70.0% at age 24 years). Compared with the original cohort, participants at age 24 years more often had dispensed asthma medications (6.6% versus 5.8%) and university education (52.6% versus 47.8%), but had less often a mother who smoked during pregnancy (9.6% versus 10.8%). Similar patterns were observed in comparisons with the general population. Female sex [odds ratio (OR) 1.75, 95% confidence interval (CI) 1.49-2.05] and high parental education (OR 1.33, 95% CI 1.12-1.59) were the predominant predictors of retention in the cohort.

Female sex, higher socio-economic status, and asthma were predictive factors of retention in the cohort after 24 years. This may have influenced the observed prevalence rates of asthma and allergies, and associations between potential risk factors and outcomes associated with retention.