(C-108) Antidepressant use in a nationally representative cohort of older adults: longitudinal use and polypharmacy

Background: Antidepressants (AD) are widely prescribed. There are several drug classes, and their effectiveness and side effects vary for individuals. Benzodiazepines are deemed inappropriate for older adults.

Objectives: Using a nationally representative longitudinal study of older adults born between 1927 and 1965, the National Social Life, Health and Aging Project (NSHAP), we describe AD use among 4,484 older adults, measured in 2015/16 (W3) and again in 2022/23 (W4), and sociodemographic and pharmacological associations of longitudinal use.

Methods: W3 data were collected by interviewers examining medications in the home, but W4 was multimodal due to the pandemic. Focusing on those who participated in both waves, we use logistic regression models to examine correlates of quitting among W3 AD users (N = 826) and correlates of initiating among W3 AD nonusers (N = 3,601). Pharmacological variables include AD type and polypharmacy. Sociodemographics include age, gender, race/ethnicity, and education.

Results: 18% in W3 and 23% in W4 were AD users. In W3, 54% of AD users were on selective serotonin reuptake inhibitors (SSRI), 16% on serotonin-norepinephrine reuptake inhibitors (SNRI), and 16% on benzodiazepines. Distribution in W4 was similar, with slightly more SNRI use (21%) and less benzodiazepine use (11%). In both waves, 23% of AD users were taking >1 AD. Among those using >1 AD, the most common combinations were SSRIs and benzodiazepines (22% in W3, 14% in W4), SSRIs and atypical ADs (19% in W3, 23% in W4), and SSRIs and phenylpiperazines (16% in W3, 22% in W4).

Between waves, 24% of users quit and 10% of non-users initiated use. Models indicate the odds of quitting decrease by 62% for every extra AD a participant used (OR: 0.38, 95% CI 0.20, 0.74). Black and Hispanic participants had much higher odds for quitting than white participants. Drug-specific trends indicated that participants taking benzodiazepines tended to be more likely to quit (OR: 1.77) and that participants taking SSRIs tended to be less likely to quit (OR: 0.64), although both effects failed to reach statistical significance with small sample sizes. Among W3 nonusers, the odds of initiating were significantly higher for female participants (OR 1.87, 95% CI: 1.32, 2.65) and significantly lower for Black (OR 0.62, 95% CI: 0.38, 1.00) and Hispanic participants (OR 0.49, 95% CI: 0.28, 0.88). Age and education were not associated with any use pattern.

Conclusions: AD use was high among older adults in 2015/16 and even higher in 2022/23. While use of benzodiazepines decreased, they were still being prescribed for many older adults. SSRIs continued to be the most used AD, and were often combined with other ADs. AD type, AD polypharmacy, race/ethnicity, and gender are associated with patterns in AD use among older adults.