Abstract

Background: Cardiovascular disease is a major global health burden, emphasizing the importance of blood pressure medications and lipid-modifying agents in prevention and treatment. Accurate self-reported data on the use of these drugs are vital for epidemiological research, yet the influence of age, sex, and cognitive function on reporting validity is not well understood. This study assesses self-reported use of these two drug classes within a large population-based Norwegian cohort and examines how demographic and cognitive factors affect reporting accuracy.

Methods: Cross-sectional study using data on self-reported medication use in the fourth wave of the population-based Trøndelag Health Study (HUNT4, 2017-19). Sensitivity, specificity and Cohen´s Kappa Statistics were calculated to assess accuracy using the Norwegian Prescribed Drug Registry as the gold standard. Analyses were stratified by age and sex, and by cognitive function for participants aged 70 years or older in the substudy HUNT4 70+.

Results: A total of 55,086 participants (54.3% women) were included in this study, and the mean age (SD) was 54.0 (17.4) years. Out of these, 8,762 also participated in the substudy HUNT4 70+ (2017-2019), where 3,132 participants had mild cognitive impairment and 779 had dementia. In the total sample, sensitivity and specificity were >90% for both blood pressure medication and lipid-modifying agents. Kappa statistics indicated substantial agreement for blood pressure medication (0.74-0.79), and substantial to almost perfect agreement for lipid-modifying agents (0.79-0.81). Agreement was highest among middle-aged participants and lower in the oldest. Chi-square tests revealed significant sex differences for both medication groups (p<0.001), demonstrating higher agreement among women. Furthermore, cognitive function was strongly associated with reporting accuracy, with the highest agreement among cognitively healthy participants.

Conclusion: Self-reported data on the use of blood pressure medication and lipid-modifying agents were highly reliable. However, agreement was influenced by age, sex, and cognitive function.

Tilgang til artikkelen