Personality phenotype has been associated with subsequent dementia in studies of older adults. This study used Project Talent data to examine whether personality during adolescence—a time when pre-clinical dementia pathology is unlikely to be present—confers risk for dementia in later life.

The full text of this article appears in the October 16, 2019 edition of JAMA Psychiatry,

Abstract

Personality phenotype has been associated with subsequent dementia in studies of older adults. However, neuropathology often precedes cognitive symptoms by many years, and may affect personality itself. Thus, it is unclear whether supposed dementia-prone personality profiles (high Neuroticism and low Conscientiousness) are true risk factors, or merely reflections of pre-existing disease.

Objectives: To examine whether personality during adolescence—a time when pre-clinical dementia pathology is unlikely to be present—confers risk for dementia in later life. We also tested whether associations could be accounted for by health factors in adolescence, or differed across socioeconomic status (SES).

Participants: Members of Project Talent (PT), a national sample of high school students in 1960.

Assessments: Ten personality traits were measured by the 150-item PT Personality Inventory. SES was measured by a composite based on parental education, occupation, and property ownership. Participants were also surveyed on demographic factors and height and weight.

Main Outcome and Measures: Medicare records were collected, with dementia diagnoses in the period of 2011-2013 classified according to the US Centers for Medicare and Medicaid Services (CMS) ICD-9-based algorithm. Cox proportional hazard models estimated the relative risk for dementia based on the ten personality traits, testing interactions with SES and adjusting for demographic confounders.

Results: The sample of 82,232 participants was 50% female, with M (SD) ages at baseline of 15.75 (1.67) and at follow-up of 69.51 (1.23). Lower risk of dementia was associated with higher levels of vigor (HR for 1 standard deviation = .93, 95% confidence interval (CI) .90, .97, p < .001). Calm and maturity showed protective associations with later dementia that increased with SES (interaction p’s < .001). At +1 standard deviation of SES, calm showed an HR of .89 (95% CI = .84, .95), maturity showed an HR of .90 (95% CI = .85, .96).

Conclusions and Relevance: The adolescent personality traits associated with later life dementia are similar to those observed in studies of older persons. Moreover, the reduction in dementia risk associated with a calm and mature adolescent phenotype was greater at higher levels of SES. Personality phenotype may be a true risk factor for dementia, preceding it by many decades and interacting with adolescent socioeconomic conditions.