ImportanceStudies investigating the association of threat-related and deprivation-related adverse childhood experiences (ACEs) with later-life cognitive decline are lacking.ObjectivesTo evaluate the independent association of threat-related and deprivation-related ACEs with cognitive decline over time among middle-aged and older Chinese adults and to examine the modifying role of social isolation in such associations.Design, Setting, and ParticipantsThis prospective cohort study used cognitive data from the China Health and Retirement Longitudinal Study (CHARLS) baseline survey that was administered between June 1, 2011, and March 31, 2012, and the CHARLS follow-up survey administered between July 1 and September 30, 2015. The life history survey with information of ACEs was additionally administered between June 1 and December 31, 2014. Statistical analysis was performed from March 1 to July 31, 2022. The study population consisted of middle-aged and older adults (age range, 45-97 years) with complete data on ACEs and 2 cognitive assessments and without cognitive impairment at baseline.ExposuresFive threat-related ACEs (ie, physical abuse, household substance abuse, domestic violence, unsafe neighborhood, and bullying) and 5 deprivation-related ACEs (ie, emotional neglect, household mental illness, incarcerated household member, parental separation or divorce, and parental death) before 17 years of age were queried by questionnaires. The cumulative scores of the 2 ACE dimensions were calculated and grouped into 3 categories as 0, 1, and 2 or more in main analyses.Main Outcomes and MeasuresCognitive function was measured by episodic memory and executive function. Global cognition was further calculated as the total score of these 2 dimensions. The raw scores of each cognitive test were standardized to z scores using baseline means and SDs. Linear mixed-effects models were constructed to examine the association between 2 dimensions of ACEs and the rate of annual cognitive decline. The modifying role of baseline social isolation in such associations was assessed with 3-way interaction tests.ResultsOf the 6466 participants included in main analyses, 3301 (51.1%) were men and the mean (SD) age was 57.2 (8.3) years. Compared with no exposures, experience of 1 deprivation-related ACE was associated with faster cognitive decline in global cognition (β = −0.012 [95% CI, −0.022 to −0.002] SD/y) and executive function (β = −0.010 [95% CI, −0.020 to −0.00002] SD/y), whereas individuals with at least 2 childhood deprivations had faster cognitive declines in all cognitive tests (β = −0.035 [95% CI, −0.050 to −0.019] SD/y for global cognition; β = −0.047 [95% CI, −0.068 to −0.025] SD/y for episodic memory; β = −0.019 [95% CI, −0.034 to −0.004] SD/y for executive function). However, such an association was not observed for threat-related ACEs. In addition, baseline social isolation was a significant modifier in the associations between deprivation-related ACEs and cognitive declines in global cognition (β = −0.033 [95% CI, −0.061 to −0.005] SD/y; P = .02 for 3-way interaction) and executive function (β = −0.032 [95% CI, −0.059 to −0.005] SD/y; P = .02 for 3-way interaction).Conclusions and RelevanceDeprivation-related ACEs, but not threat-related ACEs, were associated with faster decline in later-life cognitive function, whereas social isolation could modify such detrimental impact. These findings highlight the potential benefits of promoting social integration in maintaining later-life cognitive function among individuals who have experienced childhood deprivation.