Background and aims Smoking is associated with increased risk for brain aging/atrophy and dementia. Few studies have examined early associations with brain aging. This study aimed to measure whether adult men with a history of heavier smoking in early mid‐life would have older than predicted brain age 16–28 years later. Design Prospective cohort observational study, utilizing smoking pack years data from average age 40 (early mid‐life) predicting predicted brain age difference scores (PBAD) at average ages 56, 62 (later mid‐life) and 68 years (early old age). Early mid‐life alcohol use was also evaluated. Setting Population‐based United States sample. Participants/cases Participants were male twins of predominantly European ancestry who served in the United States military between 1965 and 1975. Structural magnetic resonance imaging (MRI) began at average age 56. Subsequent study waves included most baseline participants; attrition replacement subjects were added at later waves. Measurements Self‐reported smoking information was used to calculate pack years smoked at ages 40, 56, 62, and 68. MRIs were processed with the Brain‐Age Regression Analysis and Computation Utility software (BARACUS) program to create PBAD scores (chronological age—predicted brain age) acquired at average ages 56 (n = 493; 2002–08), 62 (n = 408; 2009–14) and 68 (n = 499; 2016–19). Findings In structural equation modeling, age 40 pack years predicted more advanced age 56 PBAD [β = −0.144, P = 0.012, 95% confidence interval (CI) = –0.257, −0.032]. Age 40 pack years did not additionally predict PBAD at later ages. Age 40 alcohol consumption, but not a smoking × alcohol interaction, predicted more advanced PBAD at age 56 (β = −0.166, P = 0.001, 95% CI = –0.261, –0.070) with additional influences at age 62 (β = −0.115, P = 0.005, 95% CI = –0.195, –0.036). Age 40 alcohol did not predict age 68 PBAD. Within‐twin‐pair analyses suggested some genetic mechanism partially underlying effects of alcohol, but not smoking, on PBAD. Conclusions Heavier smoking and alcohol consumption by age 40 appears to predict advanced brain aging by age 56 in men.

People age at different rates and in different biological systems that may differentially contribute to accelerated decline. Better understanding of biological aging may contribute to identification of better targets for intervention. In 1005 VETSA participants we created 3 indicators of biological age: physiological age (PA), frailty, and brain age. PA included hemoglobin, glucose, lipids, height, weight, waist, systolic and diastolic blood pressure, and age. PA was calculated using the Klemera and Doubal (2006) method. The frailty index summed 37 health deficits (Jiang et al. 2017). A machine learning algorithm was used to estimate brain age across cortical and subcortical regions (Liem et al, 2017); predicted brain age subtracted from chronological age comprised the predicted brain age difference score (PBAD). Frailty and PBAD were calculated at waves 1, 2 and 3 when participants were average age 56, 62, and 68, respectively. PA markers were only available at waves 2 and 3. Outcome measures included mortality by wave 3 and scores on AD-related plasma biomarkers—Neurofilament light (NFL), Tau, and AB40 and AB42 at wave 3. Frailty at wave 1 and 2 predicted mortality. Frailty at wave 1 was significantly associated with wave 3 NFL, AB42 and AB40. Wave 2 & 3 frailty was associated with all biomarkers. Neither PA nor PBAD predicted biomarkers or mortality. The results are striking given the relatively young age of the sample. Even as early as one’s 50s, frailty in a community-dwelling sample predicted accelerated decline and mortality when the outcome age was only 66-73.

BackgroundAbout 7 million older American adults are affected by frailty (Bandeen‐Roche et al., 2015). Frailty places older adults at an increased risk of adverse physical and mental health outcomes, including Alzheimer’s Disease (AD). However, prior to old age, frailty or its relationship to AD biomarkers such as beta‐amyloid 40 and 42 (Aβ40, Aβ42), tau, and neurofilament light (NfL) are understudied.MethodParticipants in this prospective longitudinal observational study were community‐based men of predominantly European ancestry. Data were collected across 3 waves at average ages 56 (2002‐08), 62 (2009‐14), and 68 (2016‐19). We calculated frailty index (FI) scores using 37 out of 49 items from an index previously validated in the UK Biobank (Williams et al. 2019; Jiang et al. 2017) which assesses health deficits across multiple physiological systems. FI is calculated as a proportion of deficits out of the total, with scores ranging from 0 (no deficits) to 1 (all deficits present). AD‐related plasma biomarkers for Aβ40,and Aβ42, total tau, and NfL were assayed at average ages 56 and 68. Biomarker values were adjusted for study site and storage time.ResultAverage FI score at age 56 was .2, equivalent to having about 7 health deficits. Frailty was significantly correlated (rs=.67‐.74) over time. Higher frailty at both age 56 and 62 doubled the risk for mortality across the 12 years of the study. Age 56 frailty was significantly associated with age 56 tau. Age 68 frailty was significantly associated with age 68 NfL, tau, Aβ40 and Aβ42. In models adjusted for age, education, and ethnicity, age 56 frailty predicted age 68 NfL (B=0.64, t=2.72, p=0.007); associations with age 68 Aβ40 (B=0.39, t=1.96, p=0.0513) and Aβ42 were attenuated (B=0.27, t=1.79, p=0.0743). Age 62 frailty mediated associations between age 56 tau and age 68 tau and NfL, but not other biomarkers.ConclusionThe results are striking given the relatively young age of the sample. Even as early as one’s 50s, links between frailty and increased risk for AD emerged. Public health groups have recommended routine frailty screening in adults over 65, but it may be important to start screenings earlier.