Autobiographical memory specificity and the persistence of depressive symptoms in HIV-positive patients: Rumination and social problem-solving skills as mediators

Abstract

Individuals infected with human immunodeficiency virus (HIV) are at elevated risk for depressive conditions, which in turn can negatively impact health-related behaviours and the course of illness. The present study tested the role of autobiographical memory specificity and its interaction with perceived stress in the persistence of depressive symptoms among dysphoric HIV-positive individuals. Additionally, we examined whether rumination and social problem solving mediated these effects. Results indicated that memory specificity moderated the impact of perceived stress, such that perceived stress was more strongly associated with follow-up depressive symptoms among those with greater memory specificity. Rumination, but not social problem solving, mediated this effect. Implications of these findings are discussed.

Keywords:

• Overgeneral autobiographical memory
• Depression
• Stress
• HIV/AIDS
• Rumination
• Problem solving

Notes

¹Analyses were also conducted using general memories (defined as total number of categoric and extended memories), with the same pattern of results found. Likewise, the same pattern of results was found when examining each valence of cue words (negative, neutral, positive) separately.

²The form of this interaction was the same as that with depressive symptoms as the dependent variable: perceived stress prospectively predicted increases in rumination at high, b=1.36, t=2.11, p<.05, but not low, b= − 0.45, t=0.76, p=.45, memory specificity.

³Because rumination and depressive symptoms were both assessed at Time 2, it is possible that depressive symptoms acted as the mediator for the relationship between the interaction of perceived stress and memory specificity and rumination (instead of rumination acting as the mediator as presented above). Using a bootstrap analysis with 5000 resamples and a 95% confidence interval, we examined the indirect effect of this “reverse” mediation model. In this analysis, depressive symptoms failed to significantly mediate the moderated effect in predicting change in rumination; the confidence interval contained 0, indirect effect=.13, CI=(−0.03; 0.32).

⁴We should note that it is also possible that the specific memories experienced by participants with the highest level of perceived stress were intrusive memories that were emotionally distressing. We thank Filip Raes for suggesting this possibility to us.