Biopsychosocial predictors of perceived life expectancy in a national sample of older men and women.

Perceived life expectancy (PLE) is predictive of mortality risk in older adults, but the factors that may contribute to mental conceptions of PLE are unknown. We aimed to describe the sociodemographic, biomedical, behavioral, and psychological predictors of self-reported PLE estimates among older English adults. Data were from 6662 adults aged 50-79 years in the population-based English Longitudinal Study of Ageing (cross-sectional sample from 2012/13). PLE was assessed in the face-to-face study interview ("What are the chances you will live to be age x or more?" where x = current age plus 10-15 years). Responses were categorized as 'low' (0-49%), 'medium' (50-74%), and 'high' (75-100%). Adjusted prevalence ratios (PRs) and 95% confidence intervals (CIs) for low vs. high PLE were estimated using population-weighted modified Poisson regression with robust error variance. Overall, 1208/6662 (18%) participants reported a low PLE, 2806/6662 (42%) reported a medium PLE, and 2648/6662 (40%) reported a high PLE. The predictors of reporting a low PLE included older age (PR = 1.64; 95% CI: 1.50-1.76 per 10 years), male sex (PR = 1.14; 95% CI: 1.02-1.26), being a smoker (PR = 1.39; 95% CI: 1.22-1.59 vs. never/former smoker), and having a diagnosis of cancer or diabetes. A low sense of control over life was associated with low PLE, as was low satisfaction with life and worse self-rated health. Those with a higher perceived social standing were less likely to report a low PLE (PR = 0.90; 95% CI: 0.87-0.93 per 10-point increase, out of 100). This study provides novel insight into potential influences on older adults' expectations of their longevity, including aspects of psychological well-being. These results should be corroborated to better determine their implications for health-related decision-making, planning, and behavior among older adults.