Abstract
INTRODUCTION
Volunteering is linked to cognitive benefits in aging, but evidence in diverse populations is limited.
METHODS
We examined volunteering and cognition in Kaiser Healthy Aging and Diverse Life Experiences Study/Study of Healthy Aging in African Americans participants (N = 2789) with unimpaired cognition at baseline. Volunteering and frequency of volunteering in the past year at baseline were self‐reported, and cognition (executive function [EF], verbal episodic memory [VEM]) was assessed with the Spanish and English Neuropsychological Assessment Scale across 4 waves (range of follow‐up: 2–6 years). Linear mixed‐effect models adjusted for demographics.
RESULTS
Participants were 73.8 ± 7.8 years on average; 62% women; 45% Black, 21% White, 18% Asian, and 17% Hispanic/Latin(x); and 47% reported volunteering. Volunteers had higher baseline EF and VEM than non‐volunteers, with the largest gains among those volunteering a few times per week. Volunteering was not associated with rates of cognitive decline.
DISCUSSION
Volunteering was associated with better baseline cognition but not slower decline, suggesting immediate cognitive benefits for racially and ethnically diverse older adults.
Highlights
Kaiser Healthy Aging and Diverse Life Experiences Study and Study of Healthy Aging in African Americans are a racially/ethnically diverse cohort (18% Asian, 47% Black, 17% Latin[x], 21% White) reporting volunteering within 12 months prior to baseline.
Late‐life (55+ years) volunteering is associated with better executive function (β = 0.173, 95% confidence interval [CI]: 0.114–0.232) and verbal episodic memory (β = 0.132, 95% CI: 0.071–0.192) after adjusting for age, gender/sex, education, race/ethnicity, instrumental activities of daily living, and self‐rated health.
Volunteering in late life, a few times per week, is associated with the highest magnitude of executive function (β = 0.216, 95% CI: 0.128–0.305) and once per week with verbal episodic memory (β = 0.189, 95% CI: 0.082–0.297) versus no volunteering, but the magnitude did not increase with more frequent volunteering.
Those who volunteered had similar domain‐specific cognitive decline compared to those who did not.
Keywords: cognition, dementia, race, volunteering
1. BACKGROUND
Volunteering entails offering services or work for a cause or activity without financial compensation. An expanding body of evidence suggests that volunteering may confer benefits to brain and cognitive health. 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 Volunteering in late life has been demonstrated in previous studies to be associated lower risk of dementia, 1 , 6 higher cognition, 2 , 3 , 4 , 5 , 6 , 8 slower cognitive decline, 3 , 4 , 6 , 8 and less risk of cognitive impairment. 7 This is particularly significant for older adults, who face an elevated risk for cognitive decline as well as Alzheimer's disease and related dementias (ADRD). 9 Consequently, volunteering, specifically formal volunteering, which is a type of volunteering that involves providing services or work within a structured organization or group, is increasingly regarded as a potential protective and modifiable lifestyle intervention that could mitigate cognitive decline in late life. Nonetheless, even though minoritized groups are at higher risk of ADRD compared to non‐Hispanic White adults, volunteering in racially and ethnically diverse older adults is often understudied and remains insufficiently understood. 10
The bulk of the research on volunteering and general health focuses on formal volunteering. 3 , 11 , 12 , 13 , 14 In an aging population, volunteering can be an important activity for the health and well‐being of older individuals and their communities. 15 Adults in old age may become less physically active, socially isolated, and less cognitively stimulated, leading to increased risk for cognitive deficiency and ADRD. The Lancet Commission on dementia reported physical inactivity, depression, and social isolation during mid and late life as risk factors attributable to dementia. 16 These risks are especially prominent after retirement when older adults experience drastic changes in their day‐to‐day activities. 12 , 17 , 18 Therefore, it is vital for older adults to engage in activities that can address these risks.
Volunteering has gained recognition as a beneficial activity for adults in late life. Systematic reviews found that volunteering is linked not only to better general health but better cognitive functioning. 2 , 3 , 5 A systematic review analyzing 15 studies on volunteering and cognition among older adults over the age of 55 has identified a positive impact of volunteering on both overall and domain‐specific cognitive functions, with mild to moderate relationships in 11 out of the 15 studies. 5 Another systematic review incorporated an additional 14 studies, including adults outside of the United States. 2 This review demonstrated that 9 out of the 14 studies indicate a positive correlation between volunteering and cognition. Many studies were conducted in homogeneous populations, leaving gaps in understanding about whether volunteering benefits across ethno‐racially diverse groups of older adults.
Theoretical frameworks have been proposed to explain the mechanism of volunteering for cognitive health. Anderson et al. 3 posit that volunteering operates through the pathways of physical, social, and cognitive activity that affect functional well‐being, which supports biological and psychological processes that lower dementia risk. The authors found that an increasing number of volunteering hours improves overall health, psychosocial well‐being, and functional limitations. This finding has been supported in other studies. 2 , 5 , 7 , 19 The revised scaffolding theory of aging and cognition (STAC‐R) 8 framework suggests that volunteering activates the compensatory scaffolding process, consisting of brain activity such as neurogenesis and recruitment, which may counteract the adverse effects of aging. A study drawing upon STAC‐R found that volunteering for more hours was associated with better working memory and processing. 8
In this study, we examined how late‐life volunteering is associated with cognitive function and cognitive decline in a racially and ethnically diverse cohort of older adults. Additionally, we examined how the frequency of volunteering is associated with cognitive function and cognitive decline. We hypothesized that those who volunteered would have better cognitive functioning and less cognitive decline than those who did not volunteer, and that those who volunteered more often would experience better cognitive performances and less cognitive decline than those who volunteered less frequently.
2. METHODS
2.1. Study population
The Kaiser Healthy Aging and Diverse Life Experiences Study (KHANDLE) and Study of Healthy Aging in African Americans (STAR) are harmonized cohort studies of 3050 individuals with the aim of evaluating how life‐course sociocultural and cardiovascular risk factors are associated with cognitive decline and risk for cognitive impairment in racially and ethnically diverse older adults. Participants of KHANDLE and STAR are long‐term members of the Kaiser Permanente Northern California Health System—an integrated health system—who resided in the California Bay Area and the Sacramento area. Participants were aged ≥ 65 years old for KHANDLE and ≥ 50 years old for STAR at the time of enrollment in 2017 and 2018, respectively. To date, STAR consists of three waves of data, and KHANDLE consists of four waves of data that range over 6 years of follow‐up. Exclusion criteria included diagnoses of dementia, other neurodegenerative disease (frontotemporal dementia, Lewy body disease, Pick's disease, Parkinson's disease with dementia, or Huntington's disease) in their electronic medical records, and other health conditions that may prevent study interviews (hospice activity in the past 12 months, history of severe chronic obstructive pulmonary disease in the past 6 months, congestive heart failure hospitalizations in the past 6 months, and history of end stage renal disease or dialysis in the past 12 months). 20 , 21 Both studies were approved by the University of California Davis and Kaiser Permanente Institutional Review Boards (IRB).
RESEARCH IN CONTEXT
Systematic review: We conducted a literature search using PubMed. We identified peer‐reviewed articles that examined the association of volunteering with cognitive performance, cognitive decline, cognitive impairment, and/or dementia.
Interpretation: Volunteering later in life is associated with better cognition in a racially and ethnically diverse group. Volunteering a few times per month to once per week shows the strongest association with cognition and has a plateau effect, in which more volunteering was not associated with higher cognition.
Future directions: Our study provides evidence that volunteering benefits cognition and may lower the risk of dementia. Future research should include different types of volunteering activities, more racially and ethnically diverse populations, and should examine volunteering maintenance to better understand cognitive decline.
Participants who were ≥ 90 years (n = 18) at baseline were excluded due to their ages being capped at age 89.99 to ensure participants were not identifiable and to reduce potential bias from extreme age‐related cognitive variability and small sample size. Individuals with missing outcome data (n = 127) were excluded because their cognitive trajectory could not be estimated, and those with missing exposure data (n = 3) could not be classified with respect to volunteering status. Finally, participants with missing covariate information (n = 123) were excluded to allow for consistent adjustment across models. The final analytical sample comprises 2789 participants at baseline.
2.2. Cognitive measures
Participants received a comprehensive cognitive assessment using the Spanish and English Neuropsychological Assessment Scale (SENAS) during each wave of the study. The SENAS is a battery of cognitive tests that has undergone extensive development using item response theory methods for valid comparison across racially and ethnically diverse groups. The cognitive domains of executive function and verbal episodic memory were assessed. The executive function domain is constructed from component tasks of category fluency, phonemic (letter) fluency, and working memory (digit‐span backward, list sorting), and the verbal episodic memory domain is obtained from a multi‐trial word‐list‐learning test. More information on the SENAS can be found elsewhere. 22 , 23 Participants who received cognitive assessments had an average follow‐up of 4.8 years for KHANDLE and 2.3 years for STAR. All scores were z standardized to the baseline mean of both cohorts combined.
2.3. Volunteering status
Formal volunteering, henceforth referred to as volunteering, was assessed during each wave of the study, but only baseline volunteering was used for the analyses. Participants were asked, “Have you spent any time in the last 12 months doing volunteer work for religious, educational, health‐related, or other charitable organizations?” Those who answered “yes” to the question were categorized as having volunteered during late life, and those who answered “no” or “don't know” were categorized as having not volunteered during late life. For participants who marked “yes” to volunteering, we considered volunteer frequency for each participant with a question asking, “How frequently do you do volunteer work?” and possible responses were: (1) daily, (2) a few times per week, (3) once a week, (4) two to three times a month, (5) once every month, or (6) one to five times per year.
2.4. Covariates
Baseline sociodemographic characteristics include gender/sex, education attainment, and self‐identified race and ethnicity. Gender/sex, dichotomized as men or women, was self‐reported or derived from the electronic medical record and represents a mixture of both gender and sex. Education attainment was reported as the last education level achieved (high school or less, some college but no degree, associate's degree, bachelor's degree, master's degree, doctoral or higher‐level degree) and was dichotomized as college degree versus no college degree. Race and ethnicity were self‐identified and categorized as Asian, Black or African American, Hispanic or Latin(x) (hereafter Hispanic), non‐Hispanic White (hereafter White), and other or multi‐racial. One participant with missing gender/sex and five participants with missing education data were excluded from the analytical sample.
Baseline measures of functional status and self‐rated health were assessed. Functional status was captured using measures of activities of daily living (ADL) and instrumental activities of daily living (IADL). ADL was captured using a 9‐item questionnaire (i.e., How much difficulty do you have walking for a quarter of a mile?), and IADL was captured using a three‐item questionnaire (i.e., How much difficulty do you have managing your money?). Participants had the option to answer on a scale of: (1) no difficulty, (2) some difficulty, (3) much difficulty, or (4) unable to do. An average score for ADL and IADL was calculated for all responses. For self‐rated health, participants reported their health as: (1) poor, (2) fair, (3) good, (4) very good, or (5) excellent. Fifty‐seven participants were excluded due to missing ADL, 1 participant was excluded due to missing IADL, and 38 participants were excluded due to missing self‐rated health.
Additionally, lifestyle factors were also assessed using physical exercise, alcohol use, and lifetime smoking status. Physical exercise consists of light and vigorous exercise, with light exercise including “walking, dancing, softball, bowling, etc.,” and vigorous exercise including “cycling, jogging, swimming laps, tennis, etc.” Participants reported whether they engaged in these physical activities as follows: (1) never, (2) several times a year, (3) several times a month, (4) several times a week, and (5) every day or almost every day. Alcohol use was reported as having ever or never drunk alcohol beverages such as beer, wine, or liquor. Smoking status was reported as having ever or never smoked ≥ 100 cigarettes in their lifetime. Twenty‐six participants were missing the overall combined lifestyle factors.
2.5. Statistical analysis
Descriptive analyses of participants’ baseline characteristics were examined and stratified by volunteering status. The association of baseline volunteering status and cognitive decline was evaluated using linear mixed models. Years from baseline were used as the time scale for up to four waves of data, and are appropriate for our time‐varying exposures, such as volunteering, measured at one time‐point (baseline). 24 , 25 Random intercepts and slopes were allowed in the mixed models, and cognitive decline was defined in the model by the interaction of the volunteering exposure with the timescale variable. All mixed models were adjusted for practice effects by the inclusion of a first visit indicator, a commonly used method to account for differences in cognitive performances between each wave by conferring a benefit or jump up in cognitive performances at the first visit. 26 An indicator for interview mode (in‐person vs. over the phone) was also adjusted in all models due to changes in how participants were interviewed between study waves during the COVID‐19 pandemic in 2020. Successive models were assessed for volunteering status starting from (1) no covariate adjustment, (2) adjustment for baseline age and gender/sex, (3) further adjusted for education, and (4) further adjusted for race and ethnicity, ADL, IADL, and self‐rated health. Furthermore, model 4 was examined stratified by race and ethnicity. Additional analyses were performed to examine whether lifestyle factors (physical exercise, alcohol use, and lifetime smoking status) confounded the associations. Furthermore, sensitivity analyses were performed by restricting participants to those age ≥ 65 years (n = 2487) to ensure no selection bias due to age differences between the harmonized cohorts.
To account for attrition across study waves, inverse probability weights (IPWs) for participation in each study wave were applied to the fully adjusted models. The IPW for participation is a tool that calculates the probability of participants being included in each wave of the study and applies weights to each participant to address dropout or attrition in participants across each subsequent wave. 27 , 28 Pooled logistic regression was used to calculate probabilities of participants remaining in the study for each follow‐up wave and probabilities conditional on confounding variables. Stabilized IPW was created by dividing the marginal probabilities by conditional probabilities and was applied to each participant per wave in the mixed model. All analyses were conducted in SAS version 9.4.
3. RESULTS
3.1. Demographics
There were 2789 participants with an average age of 73.8; standard deviation (SD) ± 7.8 in this study (Table 1). Fewer than half (n = 1344, 47.0%) of participants reported having volunteered and were on average slightly younger (72.8; SD ± 7.4) than those who did not volunteer (74.5; SD ± 8.0). Approximately 62.1% (n = 1731) identified as women, with 66.7% of women having volunteered compared to men. Those who volunteered were more educated (54.1% with a college degree or higher) than those who did not volunteer (39.0% with a college degree or higher). The racial composition of this cohort consisted of 499 Asian (17.9%), 1246 Black (46.7%), 463 Hispanic (16.6%), 578 White (20.7%), and 3 other participants (0.1%) who did not identify in any of the previous racial and ethnic categories. Across the racial and ethnic groups, a higher proportion of Black and White participants volunteered (47.0% and 22.4%, respectively) than did not volunteer (42.5% and 19.3%, respectively) relative to Asian and Hispanic participants who volunteered (17.1% and 13.4%, respectively) and did not volunteer (18.7% and 19.4%, respectively). Among those who volunteered, most reported volunteering a few times per week (28.2%), and the fewest reported volunteering daily (6.6%).
TABLE 1.
Baseline characteristics of the Kaiser Healthy Aging and Diverse Life Experience (KHANDLE) and Study of Health Aging in African Americans (STAR) cohorts stratified by volunteering and no volunteering status.
| No volunteering N (%), mean (SD) | Volunteering N (%), mean (SD) | Total N (%), mean (SD) | |
|---|---|---|---|
| Number of participants | 1478 (52.99) | 1311 (47.01) | 2789 (100) |
| Baseline age | 74.65 (8.00) | 72.83 (7.44) | 73.79 (7.79) |
| Gender/sex: women | 856 (57.92) | 875 (66.74) | 1731 (62.07) |
| Race/ethnicity | |||
| Asian | 276 (18.67) | 223 (17.01) | 499 (17.89) |
| Black/African American | 628 (42.49) | 618 (47.14) | 1246 (44.68) |
| Hispanic/Latin(x) | 287 (19.42) | 176 (13.42) | 463 (16.6) |
| Non‐Hispanic White | 285 (19.28) | 293 (22.35) | 578 (20.72) |
| Others | 2 (0.14) | 1 (0.08) | 3 (0.11) |
| Education: college graduate | 577 (39.04) | 709 (54.08) | 1286 (46.11) |
| Self‐rated health | |||
| Poor | 47 (3.18) | 22 (1.68) | 69 (2.47) |
| Fair | 318 (21.52) | 182 (13.88) | 500 (17.93) |
| Good | 604 (40.87) | 521 (39.74) | 1125 (40.34) |
| Very good | 415 (28.08) | 462 (35.24) | 877 (31.44) |
| Excellent | 94 (6.36) | 124 (9.46) | 218 (7.82) |
| Activities of daily living (ADL) | 0.63 (0.75) | 0.49 (0.61) | 0.56 (0.69) |
| Instrumental activities of daily living (IADL) | 0.28 (0.57) | 0.19 (0.38) | 0.24 (0.49) |
| Light exercise a | |||
| Every day or almost every day | 510 (34.81) | 481 (37.06) | 991 (35.87) |
| Several times per week | 466 (31.81) | 464 (35.75) | 930 (33.66) |
| Several times per month | 186 (12.7) | 188 (14.48) | 374 (13.54) |
| Several times per year | 100 (6.83) | 62 (4.78) | 162 (5.86) |
| Never | 203 (13.86) | 103 (7.94) | 306 (11.07) |
| Vigorous exercise a | |||
| Every day or almost every day | 87 (5.94) | 99 (7.63) | 186 (6.73) |
| Several times per week | 225 (15.36) | 287 (22.11) | 512 (18.53) |
| Several times per month | 134 (9.15) | 151 (11.63) | 285 (10.31) |
| Several times per year | 131 (8.94) | 166 (12.79) | 297 (10.75) |
| Never | 888 (60.61) | 595 (45.84) | 1483 (53.67) |
| Alcohol uses lifetime: yes a | 971 (66.28) | 916 (70.57) | 1887 (68.30) |
| Smoking uses lifetime (100 cigarettes or more): yes a | 669 (45.67) | 533 (41.06) | 1202 (43.50) |
| Volunteer frequency | |||
| Never | 1478 (100) | ||
| 1–5 times/year | 182 (13.88) | ||
| 1–2 times/month | 166 (12.66) | ||
| 2–3 times/month | 262 (19.98) | ||
| Once/week | 246 (18.76) | ||
| Few times/week | 369 (28.15) | ||
| Daily | 86 (6.56) | ||
| Number of participants per wave | |||
| Wave 1 | 1478 (52.99) | 1311 (47.01) | 2789 |
| Wave 2 | 1010 (50.58) | 987 (49.42) | 1997 |
| Wave 3 | 890 (49.42) | 911 (50.58) | 1801 |
| Wave 4 | 398 (46.39) | 460 (53.61) | 858 |
Missing participants: light exercise = 11, vigorous exercise = 20, alcohol use lifetime = 4, smoking use lifetime = 2.
Abbreviation: SD, standard deviation.
3.2. Association of volunteering with cognition and cognitive decline
3.2.1. Executive function
In the unadjusted mixed model, participants who volunteered had, on average, higher baseline executive function (β = 0.425, 95% confidence interval [CI]: 0.355–0.494) compared to participants who did not volunteer (Table 2). The estimated association slightly attenuated when adjusted for age and gender/sex (β = 0.329, 95% CI: 0.264–0.395), education (β = 0.232, 95% CI: 0.169–0.295), and for race and ethnicity, ADL, IADL, and self‐rated health (β = 0.173, 95% CI:0.114–0.232). There was no association between volunteering and cognitive decline in executive function (Figure 1).
TABLE 2.
The association between late‐life volunteering and cognitive decline adjusted for covariates (n = 2789).
| Executive function (β, 95% confidence interval) | ||||
|---|---|---|---|---|
| Volunteering status | Model 1 | Model 2 | Model 3 | Model 4 |
| Years from baseline | −0.026 (−0.037, −0.014) | −0.027 (−0.038, −0.015) | −0.027 (−0.039, −0.015) | −0.025 (−0.037, −0.013) |
| Never | Reference | Reference | Reference | Reference |
| Volunteering | 0.425 (0.355, 0.494) | 0.329 (0.264, 0.395) | 0.232 (0.169, 0.295) | 0.173 (0.114, 0.232) |
| Years from baseline*Volunteering | −0.002 (−0.015, 0.012) | −0.001 (−0.015, 0.012) | −0.001 (−0.014, 0.013) | −0.001 (−0.015, 0.012) |
| Verbal episodic memory (β, 95% confidence interval) | ||||
| Years from baseline | −0.063 (−0.079, −0.046) | −0.062 (−0.078, −0.046) | −0.064 (−0.081, −0.048) | −0.076 (−0.093, −0.059) |
| Never | Reference | Reference | Reference | Reference |
| Volunteering | 0.347 (0.278, 0.417) | 0.228 (0.165, 0.292) | 0.167 (0.104, 0.230) | 0.132 (0.071, 0.192) |
| Years from baseline x volunteering | −0.008 (−0.027, 0.01) | −0.006 (−0.024, 0.012) | −0.005 (−0.023, 0.014) | 0.000 (−0.019, 0.019) |
Model 1: Adjusted for practice effect and interview mode.
Model 2: Model 1 + adjusted for age and gender/sex.
Model 3: Model 2 + education.
Model 4: Model 3 + race/ethnicity (ref: Black group) + activities of daily living + instrumental activities of daily living + self‐rated health.
FIGURE 1.
Executive function and verbal episodic memory predicted by volunteer status adjusted for practice effects, interview mode, baseline age, gender/sex, education, race/ethnicity, activities of daily living, instrumental activities of daily living, and self‐rated health. VE, verbal episodic.
3.2.2. Verbal episodic memory
Participants who volunteered also had, on average, higher baseline verbal episodic memory in unadjusted models compared to no volunteering (β = 0.347, 95% CI: 0.278–0.417). The estimated association also attenuated when adjusted for age and gender/sex (β = 0.228, 95% CI: 0.165–0.292), education (β = 0.167, 95% CI: 0.104–0.230), and for race and ethnicity, ADL, IADL, and self‐rated health (β = 0.132, 95% CI: 0.071–0.192). There was no association between volunteering and cognitive decline in verbal episodic memory, and verbal episodic memory decline was steeper than executive function (Figure 1).
3.3. Association of volunteering frequency with cognition and cognitive decline
3.3.1. Executive function
The frequency of volunteering, regardless of how often it occurs, is associated with higher baseline executive function in both unadjusted models and models adjusted for all covariates, compared to no volunteering (Table 3). When models were adjusted for age, gender/sex, education, race/ethnicity, ADL, IADL, and self‐rated health, there was a significant association observed in those who volunteered two to three times per month (β = 0.214, 95% CI: 0.113–0.316), once per week (β = 0. 161, 95% CI: 0.058–0.265), and a few times per week (β = 0.216, 95% CI: 0.128–0.305) compared to no volunteering. There was no association between any volunteering frequency and executive function decline.
TABLE 3.
The association between late‐life volunteering frequency categories and cognitive decline adjusted for covariates (n = 2789).
| Executive function (β, 95% confidence interval) | ||||
|---|---|---|---|---|
| Volunteer frequency | Model 1 | Model 2 | Model 3 | Model 4 |
| Years from baseline | −0.025 (−0.037, −0.014) | −0.026 (−0.038, −0.015) | −0.027 (−0.039, −0.015) | −0.025 (−0.037, −0.013) |
| Never | Reference | Reference | Reference | Reference |
| 1–5 times/year | 0.339 (0.196, 0.481) | 0.196 (0.062, 0.330) | 0.154 (0.028, 0.281) | 0.118 (−0.001, 0.236) |
| 1 times/(month or two) | 0.337 (0.186, 0.487) | 0.247 (0.106, 0.388) | 0.155 (0.022, 0.289) | 0.102 (−0.022, 0.226) |
| 2–3 times/month | 0.549 (0.427, 0.672) | 0.419 (0.304, 0.535) | 0.297 (0.187, 0.407) | 0.214 (0.113, 0.316) |
| Once/week | 0.365 (0.238, 0.491) | 0.304 (0.186, 0.422) | 0.221 (0.109, 0.333) | 0.161 (0.058, 0.265) |
| Few times/week | 0.455 (0.348, 0.561) | 0.380 (0.280, 0.480) | 0.273 (0.178, 0.368) | 0.216 (0.128, 0.305) |
| Daily | 0.436 (0.235, 0.638) | 0.351 (0.162, 0.539) | 0.210 (0.031, 0.388) | 0.151 (−0.017, 0.318) |
| Years from baseline x 1–5 times/year | 0.015 (−0.013, 0.044) | 0.015 (−0.014, 0.043) | 0.014 (−0.014, 0.043) | 0.015 (−0.014, 0.043) |
| Years from baseline x 1 times/(month or two) | −0.006 (−0.035, 0.023) | −0.005 (−0.034, 0.024) | −0.004 (−0.033, 0.025) | −0.003 (−0.032, 0.026) |
| Years from baseline x 2–3 times/month | −0.001 (−0.024, 0.022) | −0.001 (−0.024, 0.022) | −0.001 (−0.024, 0.022) | −0.002 (−0.025, 0.021) |
| Years from baseline x once/week | −0.003 (−0.025, 0.020) | −0.002 (−0.025, 0.021) | −0.002 (−0.024, 0.021) | −0.001 (−0.024, 0.021) |
| Years from baseline x few times/week | −0.002 (−0.021, 0.018) | 0.000 (−0.020, 0.019) | 0.000 (−0.020, 0.020) | −0.002 (−0.022, 0.018) |
| Years from baseline x daily | −0.023 (−0.059, 0.012) | −0.023 (−0.058, 0.013) | −0.022 (−0.058, 0.013) | −0.022 (−0.058, 0.015) |
| Verbal episodic memory (β, 95% confidence interval) | ||||
| Years from baseline | −0.063 (−0.079, −0.046) | −0.062 (−0.078, −0.045) | −0.064 (−0.081, −0.048) | −0.076 (−0.093, −0.059) |
| Never | Reference | Reference | Reference | Reference |
| 1–5 times/year | 0.348 (0.205, 0.492) | 0.186 (0.055, 0.317) | 0.159 (0.031, 0.287) | 0.127 (0.003, 0.251) |
| 1 times/(month or two) | 0.268 (0.117, 0.420) | 0.157 (0.019, 0.295) | 0.100 (−0.036, 0.235) | 0.061 (−0.068, 0.190) |
| 2–3 times/month | 0.451 (0.328, 0.574) | 0.289 (0.177, 0.402) | 0.212 (0.101, 0.322) | 0.175 (0.070, 0.280) |
| Once/week | 0.381 (0.255, 0.507) | 0.291 (0.176, 0.406) | 0.239 (0.126, 0.351) | 0.189 (0.082, 0.297) |
| Few times/week | 0.325 (0.218, 0.432) | 0.228 (0.131, 0.325) | 0.159 (0.063, 0.255) | 0.129 (0.037, 0.220) |
| Daily | 0.179 (−0.024, 0.381) | 0.095 (−0.089, 0.280) | 0.005 (−0.176, 0.186) | −0.013 (−0.187, 0.162) |
| Years from baseline x 1–5 times/year | −0.015 (−0.054, 0.024) | −0.013 (−0.052, 0.026) | −0.013 (−0.052, 0.025) | −0.004 (−0.044, 0.035) |
| Years from baseline x 1 times/(month or two) | −0.015 (−0.055, 0.025) | −0.012 (−0.051, 0.028) | −0.009 (−0.049, 0.030) | −0.001 (−0.041, 0.039) |
| Years from baseline x 2–3 times/month | −0.006 (−0.037, 0.025) | −0.004 (−0.035, 0.027) | −0.003 (−0.033, 0.028) | 0.002 (−0.029, 0.034) |
| Years from baseline x once/week | −0.011 (−0.041, 0.020) | −0.008 (−0.039, 0.023) | −0.007 (−0.037, 0.024) | −0.003 (−0.034, 0.028) |
| Years from baseline x few times/week | −0.007 (−0.034, 0.020) | −0.004 (−0.031, 0.023) | −0.003 (−0.029, 0.024) | −0.002 (−0.029, 0.025) |
| Years from baseline x daily | 0.005 (−0.043, 0.053) | 0.007 (−0.041, 0.055) | 0.008 (−0.040, 0.056) | 0.017 (−0.034, 0.067) |
Model 1: Adjusted for practice effect and interview mode.
Model 2: Model 1 + adjusted for age and gender/sex.
Model 3: Model 2 + education.
Model 4: Model 3 + race/ethnicity (ref: Black group) + activities of daily living + instrumental activities of daily living + self−rated health.
3.3.2. Verbal episodic memory
The unadjusted model indicates that volunteering more frequently is associated with better verbal episodic memory (Table 3). After adjusting for age, gender/sex, education, race/ethnicity, ADL, IADL, and self‐rated health, the association only remains significant for volunteering one to five times per year (β = 0.127, 95% CI: 0.003–0.251), two to three times per month (β = 0.175, 95% CI: 0.070–0.280), once per week (β = 0.189, 95% CI: 0.082–0.297), and a few times per week (β = 0.129, 95% CI: 0.037–0.220). There was no association between any volunteering frequency and verbal episodic memory decline.
3.4. Association of volunteering and volunteering frequency with cognition and cognitive decline stratified by race and ethnicity
Volunteering was associated with executive function adjusted for age, sex/gender, education, ADL, IADL, and self‐rated health in all race and ethnic groups (Asian, Black, Hispanic, and White participants; Table 4). Volunteering was only associated with verbal episodic memory in Black (β = 0.146, 95% CI: 0.055–0.236) and White (β = 0.139, 95% CI: 0.007–0.272) participants adjusted for age, sex/gender, education, ADL, IADL, and self‐rated health. Volunteering two to three times per month was associated with executive function in Asian, Black, and Hispanic participants, and volunteering a few times per week was associated with executive function in Black, Hispanic, and White participants. Volunteering once per week was associated with verbal episodic memory in only Asian and Black participants. There was no association of volunteering or volunteering frequency with executive function and verbal episodic memory decline in any racial or ethnic group.
TABLE 4.
The association between late‐life volunteering and volunteering frequency with cognition and cognitive decline stratified by race and ethnicity (n = 2789).
| Executive function (β, 95% confidence interval) | ||||
|---|---|---|---|---|
| Race/ethnicity | Asian | Black or African American | Hispanic/Latin(x) | Non‐Hispanic White |
| Years from baseline | −0.030 (−0.054, −0.006) | −0.010 (−0.030, 0.011) | −0.030 (−0.059, −0.001) | −0.033 (−0.057, −0.009) |
| Never | Reference | Reference | Reference | Reference |
| Volunteering | 0.230 (0.092, 0.367) | 0.157 (0.072, 0.242) | 0.166 (0.023, 0.309) | 0.180 (0.039, 0.320) |
| Years from baseline x volunteering | −0.022 (−0.05, 0.005) | −0.008 (−0.029, 0.014) | 0.026 (−0.01, 0.063) | 0.009 (−0.019, 0.037) |
| Verbal episodic memory (β, 95% confidence interval) | ||||
|---|---|---|---|---|
| Volunteering status | Asian | Black or African American | Hispanic/Latin(x) | Non‐Hispanic White |
| Years from baseline | −0.088 (−0.124, −0.052) | −0.054 (−0.084, −0.024) | −0.069 (−0.109, −0.030) | −0.100 (−0.132, −0.068) |
| Never | Reference | Reference | Reference | Reference |
| Volunteering | 0.096 (−0.055, 0.247) | 0.146 (0.055, 0.236) | 0.115 (−0.038, 0.267) | 0.139 (0.007, 0.272) |
| Years from baseline x volunteering | 0.008 (−0.033, 0.049) | 0.001 (−0.030, 0.033) | −0.014 (−0.06, 0.032) | 0.008 (−0.027, 0.043) |
| Executive function (β, 95% confidence interval) | ||||
|---|---|---|---|---|
| Volunteering frequency | Asian | Black or African American | Hispanic/Latin(x) | Non‐Hispanic White |
| Years from baseline | −0.030 (−0.054, −0.006) | −0.010 (−0.03, 0.011) | −0.030 (−0.059, −0.001) | −0.033 (−0.057, −0.009) |
| Never | Reference | Reference | Reference | Reference |
| 1–5 times/year | 0.122 (−0.172, 0.415) | 0.092 (−0.071, 0.254) | 0.155 (−0.154, 0.463) | 0.201 (−0.099, 0.502) |
| 1 times/(month or two) | 0.108 (−0.159, 0.375) | 0.032 (−0.146, 0.211) | 0.167 (−0.170, 0.504) | 0.226 (−0.077, 0.528) |
| 2–3 times/month | 0.300 (0.039, 0.561) | 0.205 (0.064, 0.346) | 0.306 (0.019, 0.592) | 0.138 (−0.094, 0.371) |
| Once/week | 0.242 (0.018, 0.467) | 0.168 (0.015, 0.322) | 0.203 (−0.072, 0.478) | 0.073 (−0.169, 0.316) |
| Few times/week | 0.339 (0.124, 0.553) | 0.193 (0.065, 0.321) | 0.113 (−0.108, 0.334) | 0.246 (0.041, 0.450) |
| Daily | −0.021 (−0.519, 0.477) | 0.208 (−0.029, 0.445) | 0.034 (−0.327, 0.395) | 0.222 (−0.182, 0.626) |
| Years from baseline x 1–5 times/year | −0.018 (−0.086, 0.049) | 0.007 (−0.034, 0.047) | 0.061 (−0.030, 0.153) | 0.026 (−0.033, 0.084) |
| Years from baseline x 1 times/(month or two) | 0.000 (−0.059, 0.060) | −0.020 (−0.064, 0.024) | 0.064 (−0.022, 0.151) | −0.016 (−0.077, 0.046) |
| Years from baseline x 2–3 times/month | −0.024 (−0.074, 0.025) | −0.008 (−0.042, 0.026) | 0.009 (−0.067, 0.086) | 0.017 (−0.028, 0.062) |
| Years from baseline x once/week | −0.014 (−0.058, 0.029) | 0.014 (−0.023, 0.051) | −0.001 (−0.064, 0.063) | −0.006 (−0.053, 0.040) |
| Years from baseline x few times/week | −0.048 (−0.091, −0.004) | −0.017 (−0.049, 0.015) | 0.043 (−0.011, 0.097) | 0.019 (−0.020, 0.057) |
| Years from baseline x daily | 0.003 (−0.074, 0.079) | −0.043 (−0.102, 0.017) | −0.035 (−0.135, 0.065) | −0.01 (−0.083, 0.062) |
| Verbal episodic memory (β, 95% confidence interval) | ||||
|---|---|---|---|---|
| Volunteer frequency | Asian | Black or African American | Hispanic/Latin(x) | Non‐Hispanic White |
| Years from baseline | −0.088 (−0.125, −0.052) | −0.054 (−0.084, −0.024) | −0.069 (−0.109, −0.030) | −0.099 (−0.131, −0.068) |
| Never | Reference | Reference | Reference | Reference |
| 1–5 times/year | 0.116 (−0.209, 0.441) | 0.066 (−0.108, 0.239) | 0.244 (−0.088, 0.576) | 0.209 (−0.075, 0.493) |
| 1 times/(month or two) | −0.091 (−0.385, 0.204) | 0.131 (−0.06, 0.323) | −0.048 (−0.406, 0.310) | 0.080 (−0.207, 0.368) |
| 2–3 times/month | 0.098 (−0.187, 0.384) | 0.213 (0.063, 0.364) | 0.263 (−0.044, 0.569) | 0.142 (−0.077, 0.361) |
| Once/week | 0.254 (0.008, 0.500) | 0.185 (0.021, 0.349) | 0.119 (−0.170, 0.409) | 0.167 (−0.061, 0.395) |
| Few times/week | 0.123 (−0.112, 0.357) | 0.138 (0.002, 0.275) | 0.105 (−0.128, 0.338) | 0.129 (−0.063, 0.321) |
| Daily | −0.361 (−0.903, 0.180) | 0.066 (−0.188, 0.321) | −0.076 (−0.463, 0.312) | 0.083 (−0.298, 0.465) |
| Years from baseline x 1–5 times/year | −0.036 (−0.136, 0.065) | 0.038 (−0.022, 0.099) | −0.065 (−0.182, 0.052) | −0.025 (−0.100, 0.049) |
| Years from baseline x 1 times/(month or two) | 0.007 (−0.081, 0.095) | 0.015 (−0.050, 0.081) | −0.013 (−0.117, 0.092) | −0.008 (−0.085, 0.069) |
| Years from baseline x 2–3 times/month | 0.034 (−0.039, 0.107) | −0.022 (−0.073, 0.028) | −0.01 (−0.104, 0.083) | 0.031 (−0.026, 0.088) |
| Years from baseline x once/week | 0.005 (−0.059, 0.069) | 0.018 (−0.036, 0.073) | 0.008 (−0.072, 0.088) | −0.029 (−0.088, 0.031) |
| Years from baseline x few times/week | −0.014 (−0.077, 0.050) | −0.019 (−0.066, 0.028) | −0.012 (−0.080, 0.056) | 0.029 (−0.020, 0.078) |
| Years from baseline x daily | 0.086 (−0.026, 0.198) | 0.006 (−0.083, 0.095) | −0.032 (−0.155, 0.091) | 0.013 (−0.080, 0.105) |
Note: Models adjusted for practice effects, interview mode, age, gender/sex, education, activities of daily living, instrumental activities of daily living, and self‐rated health.
3.5. Supplemental analyses
3.5.1. Association of volunteering and volunteering frequency adjusted for lifestyle factors
Analyses that adjusted for physical exercise, alcohol use, and smoking use did not change the estimates for executive function (β = 0.173, 95% CI: 0.115–0.232) but failed to converge for verbal episodic memory (Table S1 in supporting information). Similar trends were observed for the same covariate adjustments in models examining volunteering frequency. No associations were observed for volunteering or volunteering frequency with cognitive decline in both executive function and verbal episodic memory after adjusting for self‐rated health, physical exercise, alcohol use, or smoking use.
3.5.2. Sensitivity analyses: association of volunteering and volunteering frequency with IPW and restriction to age ≥ 65
Application of IPW for selection did not significantly change the associations observed, which remained similar to unweighted estimates (Table S2 in supporting information). In analyses restricted to participants age ≥ 65, the association of volunteering and volunteering frequency remains similar to the associations observed in the overall sample (Table S3 in supporting information).
4. DISCUSSION
In a racially and ethnically diverse cohort of older adults without ADRD, volunteering in late life was associated with higher baseline cognition, even after adjusting for potential sociodemographic and health‐related confounders. However, volunteering was not associated with rates of cognitive decline relative to no volunteering. Furthermore, volunteering a few times per month to once per week was associated with the highest level of cognitive function in both domains (memory and executive function) compared to those who did no volunteering. The result from this study supports the growing literature indicating that volunteering is associated with better cognition. 1 , 2 , 4 , 5 , 7 , 8 , 19 , 29 , 30
Multiple studies have consistently found that volunteering in late life is associated with less risk for cognitive impairment, 7 cognitive complaints, 1 and better cognitive performance scores. 4 Volunteering confers benefits that may target age‐related risk factors by allowing opportunities for older adults to experience increased physical activity, mental stimulation, socialization, and cognitive engagement. One framework explains that volunteering impacts late‐life cognitive function and cognitive decline by activating physical and psychosocial pathways that contribute to better brain and cognitive function, although these effects may vary depending on the amount and type of volunteering older adults. 3 The findings of our study are consistent with these past studies and contribute to the growing body of evidence suggesting that volunteering in later life may serve as a feasible option for improved cognition. Furthermore, our study emphasizes that this association is also pertinent among a racially and ethnically diverse population of older adults. However, not all individuals may be able to volunteer due to structural or functional barriers for older adults. For example, older adults may live in an area with little or no access to volunteering opportunities or have limited time or resources to volunteer. Furthermore, adults with physical impairment or disability may also be unable to volunteer. To address these potential barriers, we adjusted for confounding factors such as education, functional status (self‐rated health, IADL), and lifestyle factors (alcohol and smoking use) in our models.
Our hypothesis that volunteering more frequently is associated with higher cognition was supported. We found that volunteering once per week and up to two to three times per month was associated with the highest magnitude of cognitive scores, while daily volunteering had a comparatively weaker association with cognition, indicating a “plateau phenomenon” in which increasing frequency of volunteering incrementally improves cognitive performances to a certain point. These findings are supported in other studies that observed a diminishing return on cognition for the amount of time volunteering. 3 , 8
In this study, we found that volunteering itself was associated with higher domain‐specific cognition, but cognitive decline did not differ compared to those who did not volunteer across a period of ≈ 6 years. Our results contrast with findings from a longitudinal study on 11,000 participants from the Health and Retirement Survey (HRS) assessing data across nine waves, in which volunteering was associated with higher cognitive levels and higher cognition over time. 8 Similarly, other studies also found volunteering to be associated with slower cognitive decline over time. 4 , 7 , 19 In one study over a period of 2 years, volunteering was associated with lower decline in measures of verbal fluency and recall. 4 Another study using HRS panel data from 1998 to 2012 found that volunteering at one time‐point is associated with a lower likelihood of being cognitively impaired 14 years later. 7 Last, a study using HRS panel data from 2008 to 2016 found a positive association of volunteering with cognition using path models that incorporate cognitive activity, social activity, and physical activity as mediators. 19 The combined results of these studies suggest that volunteering at one time‐point is associated with less or minimal cognitive decline over time, which is contrary to our current findings.
The results from this study provided evidence that volunteering may have a direct impact on brain structure. The benefits of volunteering on brain and cognitive health through compensatory scaffolding or other pathways will need to be tested more thoroughly in future studies. In a randomized controlled study from the Baltimore Experience Corps Trial, a volunteer program for older adults to promote healthy behaviors and activities, researchers found that participants in the volunteering intervention arm had increased neuroimaging cortical and hippocampal brain volumes over time compared to those in the control arm. 31 These types of programs lay the foundation for future policies to implement or fund volunteer‐based interventions that directly support older adults.
It is unclear what the effects of volunteering on cognition are within different racial and ethnic groups. The impacts of volunteering in Black populations may not be the same in White populations due to differing patterns of volunteering behaviors between each group. 32 Past studies found that Black individuals were less likely to volunteer compared to their White counterparts, but were more likely to commit more time when engaged in volunteering activities. 29 , 33 The disparities between Black and White Americans may be explained by structural and socioeconomic differences affecting the time, resources, and health needed to commit to volunteering activities. 33 Some studies have found volunteering to have a lower magnitude of association with cognitive health in Black adults compared to White adults, 7 , 29 while another study found no differences but noted having too few Black participants to produce precise results. 8 Many of the empirical studies on volunteering and cognitive functioning do not adequately address racial and ethnic differences in volunteering on cognition, and these studies usually incorporate data from a homogenous population or cohort. 2 , 5 Studies that use HRS data, although nationally representative, typically consist of a majority of non‐Hispanic White participants from higher socioeconomic status. 7 , 8 , 17 , 18 , 19 , 34 In international studies, the target population typically consists of the majority ethnic group within their respective countries. 1 , 2 , 4 , 35 Unlike past studies, KHANDLE contains an almost equal proportion of Asian, Hispanic, White, and Black participants. In our cohort, Asian and Hispanic participants had a lower proportion of individuals who volunteered compared to Black and White participants. When we adjusted for race and ethnicity in our statistical model, the estimated association attenuated slightly but remained significant for volunteering on cognition. When stratified by race and ethnic groups, we found that volunteering was associated with higher baseline executive function in all groups, but only with higher verbal episodic memory in Black and White participants. These results may suggest that race and ethnicity could be important moderators in the relationship between volunteering and cognition.
There were several limitations in our study. First, we did not explore different types of volunteering activities, and we only used one question to capture volunteering. Different formal volunteering activities may be more likely to engage different pathways for the reduction of risk in cognitive decline and dementia. 3 , 5 Second, we only explored volunteering at one time‐point, which was at the start of the study. Volunteering for maintenance for older adults may be critical to maintaining better general and cognitive health. 5 , 7 Last, our analyses were limited to relatively healthy older adults with no diagnosis of cognitive impairment, physical impairment, and who were able to enroll in KHANDLE or STAR. There may be concerns with selection bias in two different ways. One, selection bias for participants who are healthy enough to volunteer 30 and two, selection bias for participants who are healthy enough and were part of the health system to be eligible for the KHANDLE and STAR study enrollment. Moreover, we were unable to determine reverse causality. Regardless, volunteering may still be beneficial to older adults with cognitive impairments. 18
Our study has several strengths. First, we leveraged data from a racially and ethnically diverse cohort, which fills in gaps from previous research with less heterogeneous cohorts. Additionally, our cohort was followed up longitudinally for up to 6 years, providing insight into the association between volunteering and long‐term cognitive decline. We also assessed volunteering frequency to examine whether the amount of volunteering contributes differently to cognitive function and decline. Last, we used a comprehensive neuropsychological battery test, the SENAS, to examine volunteering in two specific cognitive domains, executive function and verbal episodic memory performance.
5. CONCLUSIONS
As adults age, they may engage in few daily activities that help establish social connections, compared to when they were younger or working. The decrease in physical activity, social connectedness, and cognitive engagement puts older adults at greater risk for cognitive impairment and dementia diagnosis. Volunteering may serve as a simple and potentially accessible activity for older adults to counteract age‐related cognitive impairment. Furthermore, as an inherently altruistic behavior, volunteering may also contribute to the benefit of the larger community. Although our study provided promising evidence that volunteering is associated with higher cognitive performance, more work is needed to determine whether increasing volunteering in at‐risk older adults can protect brain and cognitive health and promote volunteering as a public health intervention. Specifically, future studies should incorporate longer follow‐up time to observe changes in cognitive decline and examine different types of formal volunteering activities that are most conducive to brain and cognitive health. Moreover, the use of neuroimaging brain biomarkers would establish a stronger link between volunteering and risk of ADRD. Finally, future studies should consider incorporating socioeconomically diverse participants to reduce potential selection bias and better understand how volunteering affects cognition in the communities most vulnerable to cognitive decline in older age. 10
CONFLICT OF INTEREST STATEMENT
The authors have no potential conflicts of interest to disclose.
CONSENT STATEMENT
The KHANDLE and STAR studies were approved by Kaiser Permanente Northern California and the University of California, Davis Institutional Review Boards.
Supporting information
Supporting Information
Supporting Information
ACKNOWLEDGMENTS
The authors have nothing to report. This work was supported by the National Institute on Aging (NIA) grants (KHANDLE: R01AG052132 and STAR: R01AG050782). The funding sources had no role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Lor Y, Colbeth H, Chanti‐Ketterl M, et al. The impact of volunteering on cognition and cognitive decline in older diverse cohorts: KHANDLE and STAR. Alzheimer's Dement. 2026;22:e71169. 10.1002/alz.71169
Present address: Yi Lor, 4480 2nd Avenue, Sacramento, CA 95817, USA.
REFERENCES
- 1. Griep Y, Hanson LM, Vantilborgh T, Janssens L, Jones SK, Hyde M. Can volunteering in later life reduce the risk of dementia? A 5‐year longitudinal study among volunteering and non‐volunteering retired seniors. PLoS ONE. 2017;12(3):e0173885. doi:10.1371/journal.pone.0173885 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Keefer A, Steichele K, Graessel E, Prokosch HU, PL Kolominsky‐Rabas. Does voluntary work contribute to cognitive performance? – an International Systematic Review. J Multidiscip Healthc. 2023;16:1097‐1109. doi:10.2147/JMDH.S404880 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Anderson ND, Damianakis T, Kröger E, et al. The benefits associated with volunteering among seniors: a critical review and recommendations for future research. Psychol Bull. 2014;140(6):1505‐1533. doi:10.1037/a0037610 [DOI] [PubMed] [Google Scholar]
- 4. Corrêa JC, Ávila MPW, Lucchetti ALG, Altruism LucchettiG. Volunteering and cognitive performance among older adults: a 2‐Year longitudinal study. J Geriatr Psychiatry Neurol. 2022;35(1):66‐77. doi:10.1177/0891988720964260 [DOI] [PubMed] [Google Scholar]
- 5. Guiney H, Machado L. Volunteering in the community: potential benefits for cognitive aging. J Gerontol Ser B. 2018;73(3):399‐408. doi:10.1093/geronb/gbx134 [DOI] [PubMed] [Google Scholar]
- 6. Han SH, Roberts JS, Mutchler JE, Burr JA. Volunteering, polygenic risk for Alzheimer's disease, and cognitive functioning among older adults. Soc Sci Med 1982. 2020;253:112970. doi:10.1016/j.socscimed.2020.112970 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Infurna FJ, Okun MA, Grimm KJ. Volunteering is associated with lower risk of cognitive impairment. J Am Geriatr Soc. 2016;64(11):2263‐2269. doi:10.1111/jgs.14398 [DOI] [PubMed] [Google Scholar]
- 8. Proulx CM, Curl AL, Ermer AE. Longitudinal associations between formal volunteering and cognitive functioning. J Gerontol B Psychol Sci Soc Sci. 2018;73(3):522‐531. doi:10.1093/geronb/gbx110 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. 2023 Alzheimer's disease facts and figures. Alzheimers Dement. Published online March 14, 2023. doi:10.1002/alz.13016
- 10. Manly JJ, Jones RN, Langa KM, et al. Estimating the prevalence of dementia and mild cognitive impairment in the US: the 2016 health and retirement study harmonized cognitive assessment protocol project. JAMA Neurol. 2022;79(12):1242‐1249. doi:10.1001/jamaneurol.2022.3543 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Borgonovi F. Doing well by doing good. the relationship between formal volunteering and self‐reported health and happiness. Soc Sci Med. 2008;66(11):2321‐2334. doi:10.1016/j.socscimed.2008.01.011 [DOI] [PubMed] [Google Scholar]
- 12. Jongenelis MI, Jackson B, Newton RU, Pettigrew S. Longitudinal associations between formal volunteering and well‐being among retired older people: follow‐up results from a randomized controlled trial. Aging Ment Health. 2022;26(2):368‐375. doi:10.1080/13607863.2021.1884845 [DOI] [PubMed] [Google Scholar]
- 13. Morrow‐Howell N. Volunteering in later life: research frontiers. J Gerontol Ser B. 2010;65B(4):461‐469. doi:10.1093/geronb/gbq024 [DOI] [PubMed] [Google Scholar]
- 14. Yeung JWK, Zhang Z, Kim TY. Volunteering and health benefits in general adults: cumulative effects and forms. BMC Public Health. 2017;18(1):8. doi:10.1186/s12889‐017‐4561‐8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Carr DC, Fried LP, Rowe JW. Productivity & engagement in an aging America: the role of volunteerism. Daedalus. 2015;144(2):55‐67. doi:10.1162/DAED_a_00330 [Google Scholar]
- 16. Livingston G, Huntley J, Liu KY, et al. Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet. 2024;404(10452):572‐628. doi:10.1016/S0140‐6736(24)01296‐0 [DOI] [PubMed] [Google Scholar]
- 17. Han SH, Shih YC, Burr JA, Peng C. Age and cohort trends in formal volunteering and informal helping in later life: evidence from the Health and Retirement Study. Demography. 2023;60(1):99‐122. doi:10.1215/00703370‐10395916 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Lee K, Dabelko‐Schoeny H, Richardson VE. Volunteering served as a transitional role that enhances the well‐being and cognitive health among older adults with cognitive impairments. J Appl Gerontol Off J South Gerontol Soc. 2021;40(11):1568‐1578. doi:10.1177/0733464820982731 [DOI] [PubMed] [Google Scholar]
- 19. Villalonga‐Olives E, Majercak KR, Almansa J, Khambaty T. Longitudinal impact of volunteering on the cognitive functioning of older adults: a secondary analysis from the US Health and Retirement Study. Int J Nurs Sci. 2023;10(3):373‐382. doi:10.1016/j.ijnss.2023.06.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Peterson RL, George KM, Gilsanz P, et al. Racial/ethnic disparities in young adulthood and midlife cardiovascular risk factors and late‐life cognitive domains: the Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) Study. Alzheimer Dis Assoc Disord. 2021;35(2):99‐105. doi:10.1097/WAD.0000000000000436 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. George KM, Gilsanz P, Peterson RL, et al. Impact of cardiovascular risk factors in adolescence, young adulthood, and midlife on late‐life cognition: study of healthy aging in African Americans. J Gerontol Ser A. 2021;76(9):1692‐1698. doi:10.1093/gerona/glab143 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Mungas D, Reed BR, Crane PK, Haan MN, González H, Spanish and English Neuropsychological Assessment Scales (SENAS): further development and psychometric characteristics. Psychol Assess. 2004;16(4):347‐359. doi:10.1037/1040‐3590.16.4.347 [DOI] [PubMed] [Google Scholar]
- 23. Mungas DM, Reed BR, Haan MN, Gonzalez H. Spanish and English Neuropsychological Assessment Scales: relationship to demographics, language, cognition, and independent function. Neuropsychology. 2005;19(4):466‐475. doi:10.1037/0894‐4105.19.4.466 [DOI] [PubMed] [Google Scholar]
- 24. Hayes‐Larson E, Andrews RM, Kezios KL, et al. Approaches to timescale choice in cognitive aging research and potential implications for estimated exposure effects: coordinated analyses in 10 cohorts of older adults. Epidemiol Camb Mass. 2025;36(4):560‐571. doi:10.1097/EDE.0000000000001859 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Weuve J, Proust‐Lima C, Power MC, et al. Guidelines for reporting methodological challenges and evaluating potential bias in dementia research. Alzheimers Dement J Alzheimers Assoc. 2015;11(9):1098‐1109. doi:10.1016/j.jalz.2015.06.1885 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Vivot A, Power MC, Glymour MM, et al. Jump, hop, or skip: modeling practice effects in studies of determinants of cognitive change in older adults. Am J Epidemiol. 2016;183(4):302‐314. doi:10.1093/aje/kwv212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Howe CJ, Cole SR, Lau B, Napravnik S, Eron JJ. Selection bias due to loss to follow up in cohort studies. Epidemiol Camb Mass. 2016;27(1):91‐97. doi:10.1097/EDE.0000000000000409 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Weuve J, Tchetgen Tchetgen EJ, Glymour MM, et al. Accounting for bias due to selective attrition: the example of smoking and cognitive decline. Epidemiol Camb Mass. 2012;23(1):119‐128. doi:10.1097/EDE.0b013e318230e861 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Wang Y, Wong R, Amano T, Shen HW. Associations between volunteering and cognitive impairment: the moderating role of race/ethnicity. Health Soc Care Community. 2022;30(6):e4433‐e4441. doi:10.1111/hsc.13847 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Kail BL, Carr DC. More than selection effects: volunteering is associated with benefits in cognitive functioning. J Gerontol B Psychol Sci Soc Sci. 2020;75(8):1741‐1746. doi:10.1093/geronb/gbaa101 [DOI] [PubMed] [Google Scholar]
- 31. Carlson MC, Kuo JH, Chuang YF, et al. Impact of the Baltimore Experience Corps Trial on cortical and hippocampal volumes. Alzheimers Dement J Alzheimers Assoc. 2015;11(11):1340‐1348. doi:10.1016/j.jalz.2014.12.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Tang F, Copeland VC, Wexler S. Racial differences in volunteer engagement by older adults: an empowerment perspective. Soc Work Res. 2012;36(2):89‐100. 10.1093/swr/svs009 [DOI] [Google Scholar]
- 33. Gonzales E, Shen HW, Wang Y, Martinez LS, Norstrand J. Race and place: exploring the intersection of inequity and volunteerism among older Black and White adults. J Gerontol Soc Work. 2016;59(5):381‐400. doi:10.1080/01634372.2016.1224787 [DOI] [PubMed] [Google Scholar]
- 34. Kim ES, Whillans AV, Lee MT, Chen Y, VanderWeele TJ. Volunteering and subsequent health and well‐being in older adults: an outcome‐wide longitudinal approach. Am J Prev Med. 2020;59(2):176‐186. doi:10.1016/j.amepre.2020.03.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Jiang N. Formal volunteering and depressive symptoms among community‐dwelling older adults in China: a longitudinal cross‐level analysis. Health Soc Care Community. 2022;30(6):e5673‐e5684. doi:10.1111/hsc.13995 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information
Supporting Information