The Long and Short of It: Investigating Relationships between Perceived Stress and Cognitive Functioning Across Multiple Timescales in Middle to Late Adulthood

Abstract

As the global population of older adults continues to grow, an increasing number of people are expected to experience normative age-related cognitive decline and potential cognitive impairment, raising challenges for public health and support systems. At the same time, trajectories of cognitive functioning, which often begin to diverge in midlife, can vary widely across individuals. Thus, clarifying how potential risk factors relate to different facets of cognitive functioning is an important goal for research aimed at improving cognitive outcomes across middle to later adulthood. One factor that is of particular relevance to the present dissertation is perceived stress (i.e., the degree to which a person feels stressed), which has been regularly found to predict short- and long-term cognitive functioning. Yet, the conditions under which these relationships emerge, and if and when they may differ across middle to later adulthood remain unclear. This dissertation addressed these topics through two complementary studies by examining the associations between perceived stress and cognitive functioning in middle-aged and older adults across multiple timescales. Study 1 used data from the Recording Everyday Activity and Cognition using Tablets (REACT) study (N = 78, Mage = 60.55 years), which included 2 weeks of daily-life assessments of perceived stress and performance in three domains of fluid cognitive functioning (Gf): processing speed, visuospatial memory, and working memory. Participants included middle-aged adults (n = 46, Mage = 52.53) and older adults (n = 32, Mage = 72.06). A series of multilevel models tested whether greater daily perceived stress was associated with lower same-day Gf within individuals and whether higher average daily perceived stress across the study period was associated with slower improvement over time between individuals. Age group moderation of these effects was also examined, with the expectation that negative associations between daily perceived stress and Gf would be stronger among older adults. Associations between daily perceived stress and same-day Gf did not reach statistical significance in any cognitive domain. However, higher average perceived stress across the study period was associated with slower rates of improvement in processing speed and working memory in the overall sample, and with slower improvement in visuospatial memory among older adults relative to middle-aged adults. Study 2 utilized data from the Manchester Longitudinal Study of Cognition (N = 1,067, Mage = 63.71 years), in which middle-aged (n = 585, Mage = 58.82) and older adults (n = 482, Mage = 69.67) completed up to four waves of Gf and crystallized cognitive functioning (Gc) assessments at approximately 4-year intervals. At the second wave of testing, participants also rated the perceived stressfulness of 45 life events experienced during the prior 2-year period across five domains: work, family, finances, health, and social relationships. A series of multilevel models were estimated separately for total stressfulness of life events (SLE) and domain-specific SLE in relation to Gf and Gc outcomes. In the overall sample, total, health, and social-related SLE were associated with greater post-event declines in Gf. In addition, higher total and family-related SLE predicted greater post-event Gf decline among middle-aged adults relative to older adults. Similar age group moderation involving health and social-related SLE and later Gf decline was also observed in some model iterations, although these findings were not consistent across all models. By contrast, no significant associations between total or domain-specific SLE and Gc were observed in the overall sample. However, higher total and health-related SLE were associated with greater post-event Gc decline among older adults relative to middle-aged adults. Across the two studies, perceived stress was most consistently linked to between-person differences in how cognition changed over time, whether that change took the form of attenuated gains under repeated daily testing or steeper decline following periods characterized by more stressful life events. These findings support a view of stress and cognitive aging as dynamic, context-dependent, and developmentally patterned in distinct ways. Overall, associations between perceived stress and cognitive functioning appear to depend not only on whether stress is experienced, but also on how it is appraised, when it occurs, which cognitive processes are examined, and where individuals find themselves across the adult developmental continuum. Taken together, these findings suggest that prevention and practice efforts may benefit from incorporating a developmentally sensitive approach that includes perceived stress as a potentially modifiable pathway to support cognitive functioning across middle to later adulthood.