Abstract
Introduction: Depression in adults has been linked with an elevated dementia risk but that is less clear if antidepressants themselves are associated with increased dementia risk. Moreover, research indicates that cardiorespiratory fitness is associated with both depression and dementia. The current study seeks to determine the association between antidepressant (Ad) use and dementia risk, in addition to assessing the potential modifying effect of estimated Cardiorespiratory fitness (eCRF) as an indicator of cardiovascular health.
Methodology: The prospective cohort research was conducted using the data from the third cohort of the Trøndelag Health Study enacted in 2006-08 (HUNT3), and the Norwegian Prescription Database (NorPD). Data from NorPD on Defined Daily Doses (DDD) of purchased antidepressants (ATC code N06) two years before the date of participation in HUNT3 (2006-08) was the exposure variable. Exposure variables were categorized into four groups (not purchased Ad, lowest purchased Ad, second lowest purchased Ad, and highest purchased Ad). All-cause dementia was assessed in the HUNT470+ (2017-19) as an outcome. Adjusted binary logistic regression with odds ratio (OR) and 95% Confidence Interval (CI) was used to estimate the association between the purchase of Ad and the odds of all-cause dementia.
Results: The eligible sample included 8850 individuals who participated both in HUNT3 (2006-08) and HUNT470+ (2017-19). After excluding participants with missing data values on the outcome and adjustment variable we had a total sample of 5223 participants, (55.04% women, 44.96% men) with a mean age (SD) of 66.59 (5.50). Compared to those who did not purchase antidepressants (prior to participation in HUNT3) dementia risk (OR) for three tertile groups of antidepressant users was observed as 0.88 (95% CI=0.47-1.63), 1.30 (95% CI=0.75-2.22), and 2.77 (95% CI=1.74-4.40) respectively in a fully adjusted binary regressed model. We did not support any interaction from eCRF as a modifying factor when eCRF as the P-value for interaction was 0.225.
Conclusion: Our findings exhibit that there is a strong association between the highest purchase of AD and dementia risk. However, the study rejects the hypothesis of eCRF as a potential modifying factor of this association.