The roots of brain health may reach back further than retirement, midlife or even the first job. A study of 13,357 adults in China links childhood socioeconomic circumstances with patterns of thinking and memory decades later. It also points to opportunities for learning and engagement across life.
Published by Hongwei Liu and colleagues in npj Aging on October 15, the findings are observational. They do not mean that childhood disadvantage fixes anyone’s cognitive future, or demonstrate a way to prevent dementia.
How the researchers looked back at childhood
The team used the Study on Ageing and Adult Health in China, with assessments in 2007 to 2010, 2014 to 2015, and 2018 to 2019. Participants were at least 50 at entry, with an average age of 62.9. Although the study spans roughly a decade, average individual follow-up was 5.3 years. People could enter at different waves.
Childhood circumstances were estimated from participants’ reports of their parents’ education and main occupations. The researchers scored occupational skill requirements, combined the measures into a standardized index, and divided it into higher and lower groups. This captures one aspect of advantage, rather than directly measuring childhood income, housing, nutrition or adversity.
Thinking skills were assessed with immediate and delayed word recall, repeating number sequences, and naming animals. The tests cover memory, attention, working memory and aspects of executive function, the skills used to organize and manage mental tasks.
Better starting levels and different patterns of decline
A statistical model identified two broad patterns: higher cognitive scores with more gradual decline, comprising 25.3% of participants, and lower scores with faster decline, comprising 74.7%. These are model-defined groups, not dementia diagnoses or an inevitable course for everyone within them.
People in the higher childhood socioeconomic group were more likely to follow the favorable pattern: the adjusted risk ratio was 1.57, with a 95% confidence interval of 1.48 to 1.66. This is a relative comparison of membership in a modeled trajectory, not a 57% reduction in dementia risk.
Higher childhood socioeconomic status was also associated with higher cognitive scores at age 60 and slower decline across attained age. The latter distinction matters: the main model combines differences between people of different ages with changes observed within individuals. It does not track each person continuously from age 50 to 99.
What cognitive reserve means
Cognitive reserve describes the capacity to cope with aging or brain damage while continuing to function. One way to picture it is having several routes to a destination: when one becomes less efficient, other strategies may help. It is a concept about resilience, not a store of spare brain cells.
The study did not measure reserve directly with scans or pathology. It used a combined score based on participants’ own education, occupational complexity and social engagement. Engagement included activities such as visiting friends, attending meetings and participating in community groups.
A higher reserve-proxy score was associated with favorable cognitive patterns after accounting for childhood socioeconomic status. Education was the component most consistently associated with both cognitive level and decline when the components were considered together. Work complexity and social engagement were more clearly associated with cognitive level.
Adult circumstances were considered, but not all were adjusted away
The main childhood models accounted for sex, urban or rural residence and birth cohort. Cognitive-score models also accounted for familiarity with repeated testing. Participants’ education was incorporated into the reserve score and examined separately in additional analyses.
The researchers recorded current household circumstances, marital status, smoking, activity, sleep, diet and health conditions. They deliberately left these out of the primary adjustment set because these factors could lie along the pathway from early circumstances to later cognition. This analytical choice depends on assumptions about how the factors relate. It does not establish that adult health and lifestyle cannot explain part of the association.
A life-course perspective, with important uncertainties
Participants had to be able to complete the individual questionnaire, limiting applicability to people with substantial cognitive impairment. Childhood information was recalled many years later. Reserve was measured indirectly, and only three assessment waves were available. Poorer cognition could itself reduce social participation, while unmeasured influences could affect both opportunities and cognitive outcomes.
The mediation analyses suggest reserve-related experiences may statistically account for part of the childhood association. They cannot prove a causal chain or quantify how much an educational or social intervention would help. Within-person estimates were also less precise than the main age-based comparisons, particularly for the reserve score.
For dementia-prevention research, the useful question is whether improving opportunities throughout life can change later outcomes. The study supports investigating that question, rather than proving that a particular activity prevents dementia. Childhood circumstances are influences to understand, not a personal prognosis or a reason to give up on later-life opportunities.
Original research
Liu H, Guo Y, Zhong J and colleagues. Childhood socioeconomic status, cognitive reserve, and cognitive aging trajectories: a 10-year population-based cohort study. npj Aging, October 15, 2026. The full open-access accepted manuscript was reviewed. The publisher identifies it as an Article in Press awaiting final editing.