Résumé
Objective: Subjective age is consistently related to memory performance and global cognitive function among older adults. The present study examines whether subjective age is prospectively related to specific domains of cognitive function. Method: Participants were drawn from the Health and Retirement Study (HRS, N = 2549, Mean Age = 69.66, SD = 7.36) and the Midlife in the United States Survey (MIDUS, N = 2499, Mean Age = 46.24, SD = 11.25). In both samples, subjective age, depressive symptoms, chronic conditions, and demographic factors were assessed at baseline. Four domains of cognition were assessed 8 years later in the HRS and almost 20 years later in the MIDUS: episodic memory, speed-attention-executive, verbal fluency, and numeric reasoning. HRS also assessed visuospatial ability. Results: Regression analysis revealed that an older subjective age was related to worse performance in the domains of episodic memory and speed-attention-executive in both samples. The effect size for the difference between a younger and an older subjective age was d = 0.14 (MIDUS) and d = 0.24 (HRS) for episodic memory and d = 0.25 (MIDUS) and d = 0.33 (HRS) for speed-attention-executive. Feeling older was related to lower verbal fluency in HRS (d = 0.30) but not in MIDUS, whereas no association was found with numeric reasoning in either sample. An older subjective age was related to lower visuospatial ability in HRS (d = 0.25). Conclusion: Subjective age is prospectively related to performance in different cognitive domains. The associations between subjective age and both episodic memory and speed-attention-executive functions were replicable and robust over up to 20 years of follow-up.