The association of change in peak oxygen uptake with use of psychotropics in community-dwelling older adults - The Generation 100 study.

BACKGROUND

The use of psychotropics is high among the older population and may have detrimental effects on their physical and mental health. Cardiorespiratory fitness (CRF) is a strong and modifiable measure of health and declines with age. We aimed to study the association of change in CRF with use of psychotropics in community-dwelling older adults.

METHODS

We analyzed longitudinal data from 1531 older adults from the Generation 100 study, aged 70-77 years at inclusion, and with a permanent address in Trondheim, Norway. Data on objectively measured peak oxygen uptake (VO) were linked with register data from the Norwegian Prescription Database on prescribed psychotropics. The included psychotropics were antidepressants (N06A), antipsychotics (N05A), anxiolytics (N05B), hypnotics and sedatives (N05C), and N03AE (benzodiazepine derivatives). Analyses were done on any psychotropics as one group, and on the following separate groups: antidepressants (N06A), benzodiazepines (N05BA, N05CD, and N03AE), and z-hypnotics (N05CF). Peak oxygen uptake was measured four times over a five-year period and corresponding medication use was measured as defined daily doses (DDD). A random effects estimator was applied to investigate the association of change in VO with the use of psychotropics.

RESULTS

We found a statistically significant curvilinear association of change in VO with use of any psychotropics and antidepressants. For VO up to ~ 40 ml/kg/min, each 1 ml/kg/min increase was associated by a 3.3 DDD and 2.5 DDD decrease in use of any psychotropics and antidepressants, respectively. A bottoming-out effect was found and increases in VO above ~ 40 ml/kg/min showed increased use of any psychotropics and antidepressants. However, the association of change in VO with use was stronger for changes in the lower continuum of VO levels and decreased with increasing VO. No statistically significant association of change in VO with use of benzodiazepines and z-hypnotics were found. However, because of a non-randomized design, we cannot rule out the possibility of confounding by indication.

CONCLUSIONS

The results of this study show a curvilinear association of change in VO with use of any psychotropics and antidepressants in older adults. This relationship adds a new viewpoint on the adverse effects of psychotropic use and should be considered in interventions and policies aimed at reducing psychotropic medication use among the older population.