Borderline personality features moderate trajectories of health and well-being around retirement

Why we asked

Retirement is one of the biggest transitions in later life. For most people it brings a continuation of the health and well-being they already had, but some people decline, and we would like to know who. Personality is a natural place to look. Most work has examined ordinary Big Five traits, and much less is known about extreme, maladaptive traits. Borderline personality features, which include emotional swings, impulsivity, and an unstable sense of self, are a good candidate. Work can give structure and a sense of identity, so losing it might be especially hard for people high in these features. Also, people with more borderline features tend to have fewer resources going into retirement, such as savings, social support, and physical health.

What we did

We used data from the St. Louis Personality and Aging Network, a community sample of older adults. Our analysis included 590 people who were working at the first assessment (aged 55 to 65) and retired at some point while in the study. We treated borderline features as a continuum rather than a diagnosis, combining an interview, self-reports, and reports from people who knew the participant well. We then modeled depressive symptoms, physical functioning, and life satisfaction from about seven years before to ten years after each person’s retirement, using Bayesian multilevel models. The question was whether borderline features shifted the trajectory around retirement. We did not preregister this study.

What we found

Borderline features were associated with worse outcomes at every point we measured. At retirement, for example, they correlated strongly with depressive symptoms (r = .56) and moderately with physical functioning (r = −.26) and life satisfaction (r = −.31). What we did not find was much evidence that retirement changed these gaps. For physical functioning and life satisfaction, trajectories for people high and low in borderline features ran roughly parallel. For depressive symptoms there was a modest difference. People below average in borderline features stayed flat across retirement, while those above average showed rising symptoms that slowed after they retired.

What it means

We had expected retirement to be harder for people high in borderline features, and the evidence for that is limited. One possibility, which is speculation, is that the effects of these traits show up early and leave little room for further change. Our sample was fairly narrow in age, we did not know why people retired, and not everyone completed every wave. Future studies should include people with a wider range of ages and reasons for retirement.

The gap in well-being was clear, and the retirement-specific effect was small.