Discordance in chromosomal and self-reported sex in the UK Biobank: Implications for transgender- and intersex-inclusive data collection.

There is growing need to distinguish between sex and gender. While sex is assigned at birth, gender is socially constructed and may not correspond to one's assigned sex. However, in most research studies, sex or gender is assessed in isolation or the terms are used interchangeably, which has implications for research accuracy and inclusivity. We used data from the UK Biobank to quantify the prevalence of disagreement between chromosomal and self-reported sex and identify potential reasons for discordance. Among approximately 200 individuals with sex discordance, 71% of discordances were potentially explained by the presence of intersex traits or transgender identity. The findings indicate that when describing sex- and/or gender-specific differences in health, researchers may be limited in their ability to draw conclusions regarding specific sex and/or gender health information.

Comparison of approaches to control for intracranial volume in research on the association of brain volumes with cognitive outcomes.

Most neuroimaging studies linking regional brain volumes with cognition correct for total intracranial volume (ICV), but methods used for this correction differ across studies. It is unknown whether different ICV correction methods yield consistent results. Using a brain-wide association approach in the MRI substudy of UK Biobank (N = 41,964; mean age = 64.5 years), we used regression models to estimate the associations of 58 regional brain volumetric measures with eight cognitive outcomes, comparing no correction and four ICV correction approaches. Approaches evaluated included: no correction; dividing regional volumes by ICV (proportional approach); including ICV as a covariate in the regression (adjustment approach); and regressing the regional volumes against ICV in different normative samples and using calculated residuals to determine associations (residual approach). We used Spearman-rank correlations and two consistency measures to quantify the extent to which associations were inconsistent across ICV correction approaches for each possible brain region and cognitive outcome pair across 2320 regression models. When the association between brain volume and cognitive performance was close to null, all approaches produced similar estimates close to the null. When associations between a regional volume and cognitive test were not null, the adjustment and residual approaches typically produced similar estimates, but these estimates were inconsistent with results from the crude and proportional approaches. For example, when using the crude approach, an increase of 0.114 (95% confidence interval [CI]: 0.103-0.125) in fluid intelligence was associated with each unit increase in hippocampal volume. However, when using the adjustment approach, the increase was 0.055 (95% CI: 0.043-0.068), while the proportional approach showed a decrease of -0.025 (95% CI: -0.035 to -0.014). Different commonly used methods to correct for ICV yielded inconsistent results. The proportional method diverges notably from other methods and results were sometimes biologically implausible. A simple regression adjustment for ICV produced biologically plausible associations.

Representativeness of Participants in the ACCORD Trial Compared to Middle-aged and Older Adults Living with Diabetes in the United States.

BACKGROUND: We evaluated whether participants in the landmark Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial represent US adults aged ≥40 with diabetes. METHODS: Using the nationally representative 2017-2020 prepandemic National Health and Nutrition Examination Survey data, we made operational definitions of ACCORD eligibility criteria. We calculated the percentage of individuals aged ≥40 with diabetes and HbA1c ≥ 6.0% or ≥ 7.5% who met operational ACCORD eligibility criteria. RESULTS: Applying survey sampling weights to 715 National Health and Nutrition Examination Survey participants aged ≥40 with diabetes and HbA1c ≥ 6.0% (representing 29,717,406 individuals), 12% (95% confidence interval [CI] = 8%, 18%) met the operational ACCORD eligibility criteria. Restricting to HbA1c ≥ 7.5%, 39% (95% CI = 28%, 51%) of respondents met the operational ACCORD eligibility criteria. CONCLUSIONS: ACCORD represented a minority of US middle-aged and older adults with diabetes. Given the differential risk profile between ACCORD participants and the general population with diabetes, extrapolating the trial findings may not be appropriate.

Overcoming data gaps in life course epidemiology by matching across cohorts.

Lifecourse epidemiology is hampered by the absence of large studies with exposures and outcomes measured at different life stages on the same individuals. We describe when the effect of an exposure (A) on an outcome (Y) in a target population is identifiable in a combined ("synthetic") cohort created by pooling an early-life cohort including measures of A with a late-life cohort including measures of Y. We enumerate causal assumptions needed for unbiased effect estimation in the synthetic cohort and illustrate by simulating target populations under four causal models. From each target population, we randomly sampled early- and late-life cohorts and created a synthetic cohort by matching individuals from the two cohorts based on mediators and confounders. We estimated the effect of A on Y in the synthetic cohort, varying matching variables, the match ratio, and the strength of association between matching variables and A. Finally, we compared bias in the synthetic cohort estimates when matching variables did not d-separate A and Y to the bias expected in the original cohort. When the set of matching variables includes all variables d-connecting exposure and outcome (i.e., variables blocking all back- and front-door pathways), the synthetic cohort yields unbiased effect estimates. Even when matching variables did not fully account for confounders, the synthetic cohort estimate was sometimes less biased than comparable estimates in the original cohort. Methods based on merging cohorts may hasten the evaluation of early- and mid-life determinants of late-life health but rely on available measures of both confounders and mediators.

The Association Between Physical Activity and Cognition in a Racially/Ethnically Diverse Cohort of Older Adults: Results From the Kaiser Healthy Aging and Diverse Life Experiences Study.

OBJECTIVE: Most prior research on physical activity (PA) and cognition is based on predominantly white cohorts and focused on associations of PA with mean (average) cognition versus the distribution of cognition. Quantile regression offers a novel way to quantify how PA affects cognition across the entire distribution. METHODS: The Kaiser Healthy Aging and Diverse Life Experiences study includes 30% white, 19% black, 25% Asian, and 26% Latinx adults age 65+ living in Northern California (n = 1600). The frequency of light or heavy PA was summarized as 2 continuous variables. Outcomes were z-scored executive function, semantic memory, and verbal episodic memory. We tested associations of PA with mean cognition using linear regression and used quantile regression to estimate the association of PA with the 10th-90th percentiles of cognitive scores. RESULTS: Higher levels of PA were associated with higher mean semantic memory (b = 0.10; 95% CI: 0.06, 0.14) and executive function (b = 0.05; 95% CI: 0.01, 0.09). Associations of PA across all 3 cognitive domains were stronger at low quantiles of cognition. CONCLUSION: PA is associated with cognition in this racially/ethnically diverse sample and may have larger benefits for individuals with low cognitive scores, who are most vulnerable to dementia.

Association of cancer history with structural brain aging markers of Alzheimer's disease and related dementias risk.

INTRODUCTION: Cancer survivors are less likely than comparably aged individuals without a cancer history to develop Alzheimer's disease and related dementias (ADRD). METHODS: In the UK Biobank, we investigated associations between cancer history and five structural magnetic resonance imaging (MRI) markers for ADRD risk, using linear mixed-effects models to assess differences in mean values and quantile regression to examine whether associations varied across the distribution of MRI markers. RESULTS: Cancer history was associated with smaller mean hippocampal volume (b = -19 mm3 , 95% CI = -36, -1) and lower mean cortical thickness in the Alzheimer's disease signature region (b = -0.004 mm, 95% CI = -0.007, -0.000). Quantile regressions indicated individuals most vulnerable to ADRD were more affected by cancer history. DISCUSSION: Some brain MRI markers associated with ADRD risk were elevated in adults with a history of cancer. The magnitude of the adverse associations varied across quantiles of neuroimaging markers, and the pattern suggests possible harmful associations for individuals already at high ADRD risk. HIGHLIGHTS: We found no evidence of an inverse association between cancer history and ADRD-related neurodegeneration. Cancer history was associated with smaller mean hippocampal volume and lower mean cortical thickness in the Alzheimer's disease signature region. Quantile regressions indicated individuals most vulnerable to ADRD were more affected by cancer history.

Genetic risk score for Alzheimer's disease predicts brain volume differences in mid and late life in UK biobank participants.

INTRODUCTION: We estimated the ages when associations between Alzheimer's disease (AD) genes and brain volumes begin among middle-aged and older adults. METHODS: Among 45,616 dementia-free participants aged 45-80, linear regressions tested whether genetic risk score for AD (AD-GRS) had age-dependent associations with 38 regional brain magnetic resonance imaging volumes. Models were adjusted for sex, assessment center, genetic ancestry, and intracranial volume. RESULTS: AD-GRS modified the estimated effect of age (per decade) on the amygdala (-0.41 mm3 [-0.42, -0.40]); hippocampus (-0.45 mm3 [-0.45, -0.44]), nucleus accumbens (-0.55 mm3 [-0.56, -0.54]), thalamus (-0.38 mm3 [-0.39, -0.37]), and medial orbitofrontal cortex (-0.23 mm3 [-0.24, -0.22]). Trends began by age 45 for the nucleus accumbens and thalamus, 48 for the hippocampus, 51 for the amygdala, and 53 for the medial orbitofrontal cortex. An AD-GRS excluding apolipoprotein E (APOE) was additionally associated with entorhinal and middle temporal cortices. DISCUSSION: APOE and other genes that increase AD risk predict lower hippocampal and other brain volumes by middle age.

Comparison of Imputation Strategies for Incomplete Longitudinal Data in Life-Course Epidemiology.

Incomplete longitudinal data are common in life-course epidemiology and may induce bias leading to incorrect inference. Multiple imputation (MI) is increasingly preferred for handling missing data, but few studies explore MI-method performance and feasibility in real-data settings. We compared 3 MI methods using real data under 9 missing-data scenarios, representing combinations of 10%, 20%, and 30% missingness and missing completely at random, at random, and not at random. Using data from Health and Retirement Study (HRS) participants, we introduced record-level missingness to a sample of participants with complete data on depressive symptoms (1998-2008), mortality (2008-2018), and relevant covariates. We then imputed missing data using 3 MI methods (normal linear regression, predictive mean matching, variable-tailored specification), and fitted Cox proportional hazards models to estimate effects of 4 operationalizations of longitudinal depressive symptoms on mortality. We compared bias in hazard ratios, root mean square error, and computation time for each method. Bias was similar across MI methods, and results were consistent across operationalizations of the longitudinal exposure variable. However, our results suggest that predictive mean matching may be an appealing strategy for imputing life-course exposure data, given consistently low root mean square error, competitive computation times, and few implementation challenges.

Propensity Scores in Health Disparities Research: The Example of Cognitive Aging and the Hispanic Paradox.

BACKGROUND: Individuals of Mexican ancestry in the United States experience substantial socioeconomic disadvantages compared with non-Hispanic white individuals; however, some studies show these groups have similar dementia risk. Evaluating whether migration selection factors (e.g., education) associated with risk of Alzheimer disease and related dementia (ADRD) explain this paradoxical finding presents statistical challenges. Intercorrelation of risk factors, common with social determinants, could make certain covariate patterns very likely or unlikely to occur for particular groups, which complicates their comparison. Propensity score (PS) methods could be leveraged here to diagnose nonoverlap and help balance exposure groups. METHODS: We compare conventional and PS-based methods to examine differences in cognitive trajectories between foreign-born Mexican American, US-born Mexican American, and US-born non-Hispanic white individuals in the Health and Retirement Study (1994-2018). We examined cognition using a global measure. We estimated trajectories of cognitive decline from linear mixed models adjusted for migration selection factors also associated with ADRD risk conventionally or with inverse probability weighting. We also employed PS trimming and match weighting. RESULTS: In the full sample, where PS overlap was poor, unadjusted analyses showed both Mexican ancestry groups had worse baseline cognitive scores but similar or slower rates of decline compared with non-Hispanic white adults; adjusted findings were similar, regardless of method. Focusing analyses on populations where PS overlap was improved (PS trimming and match weighting) did not alter conclusions. CONCLUSIONS: Attempting to equalize groups on migration selection and ADRD risk factors did not explain paradoxical findings for Mexican ancestry groups in our study.

Association between cancer and dementia risk in the UK Biobank: evidence of diagnostic bias.

Epidemiological studies have identified an inverse association between cancer and dementia. Underlying methodological biases have been postulated, yet no studies have systematically investigated the potential for each source of bias within a single dataset. We used the UK Biobank to compare estimates for the cancer-dementia association using different analytical specifications designed to sequentially address multiple sources of bias, including competing risk of death, selective survival, confounding bias, and diagnostic bias. We included 140,959 UK Biobank participants aged ≥ 55 without dementia before enrollment and with linked primary care data. We used cancer registry data to identify cancer cases prevalent before UK Biobank enrollment and incident cancer diagnosed after enrollment. We used Cox models to evaluate associations of prevalent and incident cancer with all-cause dementia, Alzheimer's disease (AD), and vascular dementia. We used time-varying models to evaluate diagnostic bias. Over a median follow-up of 12.3 years, 3,310 dementia cases were diagnosed. All-site incident cancer was positively associated with all-cause dementia incidence (hazard ratio [HR] = 1.14, 95% CI: 1.02-1.29), but prevalent cancer was not (HR = 1.04, 95% CI: 0.92-1.17). Results were similar for vascular dementia. AD was not associated with prevalent or incident cancer. Dementia diagnosis was substantially elevated in the first year after cancer diagnosis (HR = 1.83, 95% CI: 1.42-2.36), after which the association attenuated to null, suggesting diagnostic bias. Following a cancer diagnosis, health care utilization or cognitive consequences of diagnosis or treatment may increase chance of receiving a dementia diagnosis, creating potential diagnostic bias in electronic health records-based studies.

Association of Social Integration with Cognitive Status in a Multi-Ethnic Cohort: Results From the Kaiser Healthy Aging and Diverse Life Experiences Study.

We evaluated overall and race-specific relationships between social integration and cognition in older adults. Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) cohort participants included 1343 Asian, Black, Latino, or non-Latino White Kaiser Permanente Northern California members. We estimated the effect of social integration on verbal episodic memory, semantic memory, and executive function derived from the Spanish and English Neuropsychological Assessment (SENAS) Scales. Social integration scores included marital status; volunteer activity; and contact with children, relatives, friends, and confidants. We estimated covariate-adjusted linear mixed-effects models for baseline and 17-month follow-up cognition. Social integration was associated with higher baseline cognitive scores (average  ß = 0.066 (95% confidence interval: 0.040, 0.092)) overall and in each racial/ethnic group. The association did not vary by race/ethnicity. Social integration was not associated with the estimated rate of cognitive change. In this cohort, more social integration was similarly associated with better late-life cognition across racial/ethnic groups.

California's Mental Health Services Act and Mortality Due to Suicide, Homicide, and Acute Effects of Alcohol: A Synthetic Control Application.

California's Mental Health Services Act (MHSA) substantially expanded funding of county mental health services through a state tax, and led to broad prevention efforts and intensive services for individuals experiencing serious mental disorders. We estimated the associations between MHSA and mortality due to suicide, homicide, and acute effects of alcohol. Using annual cause-specific mortality data for each US state and the District of Columbia from 1976-2015, we used a generalization of the quasi-experimental synthetic control method to predict California's mortality rate for each outcome in the absence of MHSA using a weighted combination of comparison states. We calculated the association between MHSA and each outcome as the absolute difference and percentage difference between California's observed and predicted average annual rates over the postintervention years (2007-2015). MHSA was associated with modest decreases in average annual rates of homicide (-0.81/100,000 persons, corresponding to a 13% reduction) and mortality from acute alcohol effects (-0.35/100,000 persons, corresponding to a 12% reduction). Placebo test inference suggested that the associations were unlikely to be due to chance. MHSA was not associated with suicide. Protective associations with mortality due to homicide and acute alcohol effects provide evidence for modest health benefits of MHSA at the population level.

Extension of Mendelian Randomization to Identify Earliest Manifestations of Alzheimer Disease: Association of Genetic Risk Score for Alzheimer Disease With Lower Body Mass Index by Age 50 Years.

Weight loss or lower body mass index (BMI) could be an early symptom of Alzheimer disease (AD), but when this begins to emerge is difficult to estimate with traditional observational data. In an extension of Mendelian randomization, we leveraged variation in genetic risk for late-onset AD risk to estimate the causal effect of AD on BMI and the earliest ages at which AD-related weight loss (or lower BMI as a proxy) occurs. We studied UK Biobank participants enrolled in 2006-2010, who were without dementia, aged 39-73, with European genetic ancestry. BMI was calculated with measured height/weight (weight (kg)/height (m)2). An AD genetic risk score (AD-GRS) was calculated based on 23 genetic variants. Using linear regressions, we tested the association of AD-GRS with BMI, stratified by decade, and calculated the age of divergence in BMI trends between low and high AD-GRS. AD-GRS was not associated with BMI in 39- to 49-year-olds (ß = 0.00, 95% confidence interval (CI): -0.03, 0.03). AD-GRS was associated with lower BMI in 50- to 59-year-olds (ß = -0.03, 95% CI: -0.06, -0.01) and 60- to 73-year-olds (ß = -0.09, 95% CI:-0.12, -0.07). Model-based BMI age curves for high versus low AD-GRS began to diverge after age 47 years. Sensitivity analyses found no evidence for pleiotropy or survival bias. Longitudinal replication is needed; however, our findings suggest that AD genes might begin to reduce BMI decades prior to dementia diagnosis.

Proof of Concept Example for Use of Simulation to Allow Data Pooling Despite Privacy Restrictions.

BACKGROUND: Integrating results from multiple samples is often desirable, but privacy restrictions may preclude full data pooling, and most datasets do not include fully harmonized variable sets. We propose a simulation-based method leveraging partial information across datasets to guide creation of synthetic data based on explicit assumptions about the underlying causal structure that permits pooled analyses that adjust for all desired confounders in the context of privacy restrictions. METHODS: This proof-of-concept project uses data from the Health and Retirement Study (HRS) and Atherosclerosis Risk in Communities (ARIC) study. We specified an estimand of interest and a directed acyclic graph (DAG) summarizing the presumed causal structure for the effect of glycated hemoglobin (HbA1c) on cognitive change. We derived publicly reportable statistics to describe the joint distribution of each variable in our DAG. These summary estimates were used as data-generating rules to create synthetic datasets. After pooling, we imputed missing covariates in the synthetic datasets and used the synthetic data to estimate the pooled effect of HbA1c on cognitive change, adjusting for all desired covariates. RESULTS: Distributions of covariates and model coefficients and associated standard errors for our model estimating the effect of HbA1c on cognitive change were similar across cohort-specific original and preimputation synthetic data. The estimate from the pooled synthetic incorporates control for confounders measured in either original dataset. DISCUSSION: Our approach has advantages over meta-analysis or individual-level pooling/data harmonization when privacy concerns preclude data sharing and key confounders are not uniformly measured across datasets.

The competing risk of death and selective survival cannot fully explain the inverse cancer-dementia association.

INTRODUCTION: We evaluated whether competing risk of death or selective survival could explain the reported inverse association between cancer history and dementia incidence (incidence rate ratio [IRR] ≈ 0.62-0.85). METHODS: A multistate simulation model of a cancer- and dementia-free cohort of 65-year-olds was parameterized with real-world data (cancer and dementia incidence, mortality), assuming no effect of cancer on dementia (true IRR = 1.00). To introduce competing risk of death, cancer history increased mortality. To introduce selective survival, we included a factor (prevalence ranging from 10% to 50%) that reduced cancer mortality and dementia incidence (IRRs ranged from 0.30 to 0.90). We calculated IRRs for cancer history on dementia incidence in the simulated cohorts. RESULTS: Competing risk of death yielded unbiased cancer-dementia IRRs. With selective survival, bias was small (IRRs = 0.89 to 0.99), even under extreme scenarios. DISCUSSION: The bias induced by selective survival in simulations was too small to explain the observed inverse cancer-dementia link, suggesting other mechanisms drive this association.

Exposure to Community Homicide During Pregnancy and Adverse Birth Outcomes: A Within-Community Matched Design.

BACKGROUND: Community violence is an understudied aspect of social context that may affect risk of preterm birth and small-for-gestational age (SGA). METHODS: We matched California mothers with live singleton births who were exposed to a homicide in their Census tract of residence in 2007-2011 to unexposed mothers within the same tract. We estimated risk differences with a weighted linear probability model, with weights corresponding to the matched data structure. We estimated the average treatment effect on the treated of homicide exposure on the risk of preterm birth and SGA during the preconception period and first and second trimester. RESULTS: We found a small increase in risk of SGA associated with homicide exposure in the first trimester (0.14% [95% confidence interval (CI) = -0.01%, 0.30%]), but not for exposure during the preconception period (-0.01% [95% CI = -0.17%, 0.15%]) or the second trimester (-0.06% [95% CI = -0.23%, 0.11%]). Risk of preterm birth was not affected by homicide exposure. When women were exposed to homicides during all three exposure windows, there was a larger increase in risk of SGA (1.09% [95% CI = 0.15%, 2.03%]) but not preterm birth (0.14% [95% CI = -0.74%, 1.01%]). Exposure to three or more homicides was also associated with greater risk of SGA (0.78% [95% CI = 0.15%, 1.40%]). Negative controls indicated that residual confounding by temporal patterning was unlikely. CONCLUSIONS: Homicide exposure during early pregnancy is associated with a small increased risk of SGA.

Using an Online Panel to Crosswalk Alternative Measures of Alcohol Use As Fielded in Two National Samples.

Introduction: Accurate estimation of the health effects of drinking is hampered by inconsistent phrasing of questions about alcohol use in commonly-used health surveys (e.g., HRS, NYLS79), and measurement error in brief self-reports of drinking. We fielded an online survey to a diverse pool of respondents, assessing two versions of alcohol use questions. We used the measurement survey responses to evaluate correspondence across question versions and create a crosswalk between versions of alcohol questions from two different nationally representative studies of middle-aged adults. The measurement model can also be used to incorporate measurement error correction. Methods: Respondents to two measurement survey platforms (Centiment and Qualtrics) were asked drinking frequency and quantity questions as phrased in the Health and Retirement Study (HRS: average days per week drank in the last 3 months; quantity consumed on days drank in the last 3 months) and differently phrased questions from the National Longitudinal Survey of Youth 1979 (NLSY79: days drank in last 30 days, average quantity consumed on days drank). The order in which respondents encountered different versions of the questions was randomized. From these questions, we derived measures of average weekly alcohol consumption. In the online panel data, we regressed responses to the HRS question on responses to the NLSY question and vice versa to create imputation models. HRS (n=14,639) and NLSY79 (n=7,069) participants aged 50-59 self-rated their overall health (range 0-4, 0=excellent and 4=poor). NLSY79 or HRS participants' responses to the alcohol question from the other survey were multiply imputed (k=30) using the measurement model from the measurement survey participant data (k=30). We regressed self-rated health on each alcohol measure and estimated covariate-adjusted coefficients from observed and imputed versions of the questions. Results: The measurement survey (n=2,070) included respondents aged 50+; 64.8% female; 21.4% Hispanic, 23.95% Black, 27.1% White, and 27.6% another ("Other") self-reported racial/ethnic identity. Associations of observed alcohol question responses with self-reported health were slightly smaller than associations of imputed responses for frequency of alcohol use and consumption on days when alcohol was used. For example, using the HRS version of the frequency of alcohol use (days per week), the estimate for the observed question in HRS respondents was ꞵ =-0.045 [-0.055,-0.036]; and the estimate for the imputed version of the HRS question in NLSY79 respondents was ꞵ=-0.051 [-0.065,-0.037]. The estimated effect of average drinks per week was substantially larger for the imputed version of the measure (ꞵ for the observed question in HRS=-0.002 [-0.004,0.001], ꞵ for the imputed version of the HRS measure in NLSY79 respondents=-0.02 [-0.027,-0.012]). Patterns were similar when using the NLSY79 versions of questions as reported in NLSY79 and imputed for HRS respondents. For example, the estimated effect of average drinks per week was substantially larger for the imputed version of the NLSY79 question (ꞵ for the observed question in NLSY79=-0.006 [-0.01,-0.002], ꞵ for the imputed version of the HRS question in NLSY79 respondents=-0.019 [-0.027,-0.01]). Conclusions: Measurement inconsistencies and imperfect reliability are major challenges in estimating effects of alcohol use on health. Collecting additional data using online panels is a feasible and flexible approach to quantifying measurement differences. This approach may enable measurement error corrections, improve meta-analyses, and promote evidence triangulation.

Metformin Cessation and Dementia Incidence.

Importance: Prior studies suggested that metformin may be associated with reduced dementia incidence, but associations may be confounded by disease severity and prescribing trends. Cessation of metformin therapy in people with diabetes typically occurs due to signs of kidney dysfunction but sometimes is due to less serious adverse effects associated with metformin. Objective: To investigate the association of terminating metformin treatment for reasons unrelated to kidney dysfunction with dementia incidence. Design, Setting, and Participants: This cohort study was conducted at Kaiser Permanente Northern California, a large integrated health care delivery system, among a cohort of metformin users born prior to 1955 without history of diagnosed kidney disease at metformin initiation. Dementia follow-up began with the implementation of electronic health records in 1996 and continued to 2020. Data were analyzed from November 2021 through September 2023. Exposures: A total of 12 220 early terminators, individuals who stopped metformin with normal estimated glomerular filtration rate (eGFR), were compared with routine metformin users, who had not yet terminated metformin treatment or had terminated (with or without restarting) after their first abnormal eGFR measurement. Early terminators were matched with routine users of the same age and gender who had diabetes for the same duration. Main outcomes and measures: The outcome of interest was all-cause incident dementia. Follow-up for early terminators and their matched routine users was started at age of termination for the early terminator. Survival models adjusted for sociodemographic characteristics and comorbidities at the time of metformin termination (or matched age). Mediation models with HbA1c level and insulin usage 1 and 5 years after termination tested whether changes in blood glucose or insulin usage explained associations between early termination of metformin and dementia incidence. Results: The final analytic sample consisted of 12 220 early terminators (5640 women [46.2%]; mean [SD] age at start of first metformin prescription, 59.4 [9.0] years) and 29 126 routine users (13 582 women [46.6%]; mean [SD] age at start of first metformin prescription, 61.1 [8.9] years). Early terminators had 1.21 times the hazard of dementia diagnosis compared with routine users (hazard ratio, 1.21; 95% CI, 1.12 to 1.30). In mediation analysis, contributions to this association by changes in HbA1c level or insulin use ranged from no contribution (0.00 years; 95% CI, -0.02 to 0.02 years) for insulin use at 5 years after termination to 0.07 years (95% CI, 0.02 to 0.13 years) for HbA1c level at 1 year after termination, suggesting that the association was largely independent of changes in HbA1c level and insulin usage. Conclusions and Relevance: In this study, terminating metformin treatment was associated with increased dementia incidence. This finding may have important implications for clinical treatment of adults with diabetes and provides additional evidence that metformin is associated with reduced dementia risk.

Low- and High-Density Lipoprotein Cholesterol and Dementia Risk Over 17 Years of Follow-up Among Members of a Large Health Care Plan.

BACKGROUND AND OBJECTIVES: The associations of high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C) with dementia risk in later life may be complex, and few studies have sufficient data to model nonlinearities or adequately adjust for statin use. We evaluated the observational associations of HDL-C and LDL-C with incident dementia in a large and well-characterized cohort with linked survey and electronic health record (EHR) data. METHODS: Kaiser Permanente Northern California health plan members aged 55 years and older who completed a health behavior survey between 2002 and 2007, had no history of dementia before the survey, and had laboratory measurements of cholesterol within 2 years after survey completion were followed up through December 2020 for incident dementia (Alzheimer disease-related dementia [ADRD]; Alzheimer disease, vascular dementia, and/or nonspecific dementia) based on ICD-9 or ICD-10 codes in EHRs. We used Cox models for incident dementia with follow-up time beginning 2 years postsurvey (after cholesterol measurement) and censoring at end of membership, death, or end of study period. We evaluated nonlinearities using B-splines, adjusted for demographic, clinical, and survey confounders, and tested for effect modification by baseline age or prior statin use. RESULTS: A total of 184,367 participants [mean age at survey = 69.5 years, mean HDL-C = 53.7 mg/dL (SD = 15.0), mean LDL-C = 108 mg/dL (SD = 30.6)] were included. Higher and lower HDL-C values were associated with elevated ADRD risk compared with the middle quantile: HDL-C in the lowest quintile was associated with an HR of 1.07 (95% CI 1.03-1.11), and HDL-C in the highest quintile was associated with an HR of 1.15 (95% CI 1.11-1.20). LDL-C was not associated with dementia risk overall, but statin use qualitatively modified the association. Higher LDL-C was associated with a slightly greater risk of ADRD for statin users (53% of the sample, HR per 10 mg/dL increase = 1.01, 95% CI 1.01-1.02) and a lower risk for nonusers (HR per 10 mg/dL increase = 0.98; 95% CI 0.97-0.99). There was evidence for effect modification by age with linear HDL-C (p = 0.003) but not LDL-C (p = 0.59). DISCUSSION: Both low and high levels of HDL-C were associated with elevated dementia risk. The association between LDL-C and dementia risk was modest.

Associations Between Handgrip Strength and Dementia Risk, Cognition, and Neuroimaging Outcomes in the UK Biobank Cohort Study.

Importance: The associations between muscle strength and cognitive outcomes have sparked interest in interventions that increase muscle strength for prevention of dementia, but the associations between muscle strength and cognitive aging are unclear, particularly among middle-aged adults. Objective: To evaluate the association between handgrip strength (HGS) and dementia, reduced cognition, and poorer neuroimaging outcomes in a UK population of middle-aged adults. Design, Setting, and Participants: This cohort study evaluated UK Biobank participants aged 39 to 73 years enrolled from 2006 to 2010 with measured HGS and prospectively followed up for dementia diagnosis. Data were analyzed from October 2021 to April 2022. Exposures: HGS assessed in both hands via dynamometer. Main Outcomes and Measures: Outcomes included cognitive test scores (fluid intelligence and prospective memory), brain magnetic resonance imaging measures (total brain volume, white matter hyperintensity, and hippocampal volume), and incident dementia (all-cause, vascular, and Alzheimer disease [AD] from primary care, hospital, or death records) over a median (IQR) of 11.7 (11.0-12.4) years of follow-up. Mixed-effects linear and logistic regressions and Cox proportional-hazard models were used to estimate associations, stratified by gender and adjusted for covariates. Estimates are presented per 5-kg decrement in HGS. To evaluate reverse causation, we assessed whether a polygenic risk score for AD is associated with HGS. Results: A subsample of 190 406 adult participants in the UK Biobank (mean [SD] age, 56.5 [8.1] years; 102 735 women [54%]) were evaluated. A 5-kg decrement in HGS was associated with lower fluid intelligence scores in men (ß, -0.007; 95% CI, -0.010 to -0.003) and women (ß, -0.04; 95% CI, -0.05 to -0.04. A 5-kg decrement in HGS was associated with worse odds of correctly responding to a prospective memory task for men (odds ratio, 0.91; 95% CI, 0.90 to 0.92) and women (odds ratio, 0.88; 95% CI, 0.87 to 0.90). A 5-kg decrement in HGS was associated with greater white matter hyperintensity volume in men (ß, 92.22; 95% CI, 31.09 to 153.35) and women (ß, 83.56; 95% CI, 13.54 to 153.58). A 5-kg decrement in HGS was associated with incident dementia for men (hazard ratio, 1.20; 95% CI, 1.12 to 1.28) and women (hazard ratio, 1.12; 95% CI, 1.00 to 1.26). The AD genetic risk score was not significantly associated with HGS. Conclusions and Relevance: These findings suggest that HGS is associated with measures of neurocognitive brain health among men and women and they add to a growing body of research indicating that interventions designed to increase muscle strength, particularly among middle-aged adults, may hold promise for the maintenance of neurocognitive brain health.