Body mass index and two-year change of in vivo Alzheimer’s disease pathologies in cognitively normal older adults

This study was performed as part of the Korean Brain Aging Study for Early Diagnosis and Prediction of Alzheimer’s Disease (KBASE), an ongoing prospective cohort study conducted from 2014 [21]. As of 2018, 297 CN adults between 55 and 90 years old were recruited and received a baseline evaluation, including a comprehensive clinical assessment and BMI measurement. Among them, 194 participants who had completed both baseline and two-year follow-up neuroimaging scans for brain Aβ deposition were included in the current study. The inclusion criteria were as follows: (a) age 55–90 years, (b) Clinical Dementia Rating score of 0, and (c) no diagnosis of mild cognitive impairment or dementia. The exclusion criteria were as follows: (a) any serious medical, psychiatric, or neurological disorder that could affect mental function; (b) any severe communication problem that would render clinical examination or brain scanning difficult; (c) contraindications to magnetic resonance imaging (MRI), such as a pacemaker or claustrophobia; (d) absence of a reliable informant; (e) illiteracy defined as a lack of the ability to read; and (f) participation in another clinical trial or treatment with an investigational product. Research clinicians determined the presence of any exclusion criteria by referring to the results of laboratory examinations and MRI scans. They also evaluated the clinical data collected by trained nurses during systematic interviews of participants and their reliable informants during the screening period. More detailed information on the recruitment of the KBASE cohort has been presented in a previous report [21]. The study was approved by the Institutional Review Board of the Seoul National University Hospital and SNU-SMG Boramae Medical Center, South Korea. All participants provided written informed consent.

The participants underwent comprehensive baseline clinical assessments based on the KBASE protocol [21] by trained psychiatrists. The assessments incorporated the Korean version of the Consortium to Establish a Registry for Alzheimer`s Disease Assessment (CERAD-K) [22, 23]. The presence of vascular risk factors (VRFs), including diabetes, hypertension, dyslipidemia, coronary heart disease, transient ischemic attack, and stroke, was assessed from data collected during systematic interviews by trained nurses with participants and their informants. Based on the number of VRFs, the vascular risk score (VRS) was calculated [24] and treated as a continuous variable for analyses. The Geriatric Depression Scale (GDS) [25] was used to measure the severity of depressive symptoms. Smoking status (never/former/smoker), alcohol intake status (never/former/drinker), and lifetime physical activity were evaluated through interviews with nurses. The Lifetime Total Physical Activity Questionnaire [26] was used to assess lifetime physical activity. A metabolic equivalent (MET) value was assigned to the intensity of activity based on the compendium of physical activities [27].

BMI was calculated as weight in kilograms divided by the square of the height in meters. It was measured at the baseline visit. Trained research nurses measured the participants’ height and body weight using standard anthropometric methods.

All participants underwent [¹¹C] Pittsburgh compound B (PiB)-positron emission tomography (PET) scans using a 3.0 T Biograph mMR (PET-MR) scanner (Siemens, Washington DC, USA). These scans were conducted according to the manufacturer’s protocols at baseline and two-year follow-up visit. We described the details of PiB-PET image acquisition and preprocessing previously [28]. The automatic anatomic labeling algorithm and the region combination method [29] were used to determine regions of interest (ROIs) and to characterize PiB retention in the frontal, lateral parietal, posterior cingulate-precuneus, and lateral temporal regions. A global cortical ROI (consisting of the four smaller ROIs) was also defined. The global Aβ retention value, the standardized uptake value ratio (SUVR) for the global cortical ROI, was calculated by dividing the mean values for all voxels of the global cortical ROI by a mean reference region. For the analysis of baseline data, the inferior cerebellar gray matter in the spatially unbiased infratentorial template for the cerebellum (SUIT) atlas [30] was used as the reference region. A participant was classified as Aβ positive if the SUVR was > 1.21 [31]. For longitudinal analysis, the reference region included the inferior cerebellar grey matter, cerebellar white matter (thresholded at 50%), pons, and cerebrum white matter (thresholded at 95% and eroded by three voxels) [32, 33].

A subset of subjects (n = 45) underwent two [¹⁸F] AV-1451 PET scans using a Biograph True Point 40 PET/CT platform (Siemens, USA) per the manufacturer’s guidelines at a two-year time interval. While the first PiB-PET imaging was performed during the baseline visit, the first AV-1451 PET imaging was performed at an average of 2.55 (standard deviation = 0.26) years after that visit. The details of AV-1451 PET imaging acquisition and preprocessing have been described previously [28]. We quantified the AV-1451 SUVR of a priori ROI of the “AD-signature region” of tau accumulation to estimate cerebral tau deposition. This was a size-weighted average of the partial volume-corrected uptake by the entorhinal, amygdala, parahippocampal, fusiform, inferior temporal, and middle temporal ROIs [34, 35]. It was done using the cerebral hemispheric white matter ROI from FreeSurfer in the partial volume code [36] as a reference region. The literature recommends using cerebral white matter as the reference region for intensity normalization in longitudinal AV-1451 PET data analysis [37].

We tested linear mixed-effects (LME) models with random intercepts to examine the relationships between late-life BMI and longitudinal change in AD neuropathological biomarkers. All models included Aβ or tau deposition values as dependent variables on the first and second PET scan. Model 1 included baseline BMI, age, sex, APOE4, baseline Aβ or Tau and their interactions with time. In Model 2, we additionally controlled for VRS and its interaction with time to adjust for the confounding effects of vascular risk factors, considering the well perceived role of vascular risk factors in AD development [38, 39]. A random intercept was included for each subject, and time was calculated as the number of years from baseline. For exploratory purposes, the LME model including baseline BMI, age, APOE4, baseline Aβ or tau and their interactions with time was analyzed for each sex. Statistical analyses were performed using R version 4.0, and jamovi version 2.2.1 (The jamovi project, www.​jamovi.​org). In all analyses, p < 0.05 was considered as statistical significance.

The datasets generated and analyzed during the present study are not publicly available, owing to ethics considerations and privacy restrictions. Data might be obtained from the corresponding author after approval by the Institutional Review Board of the Seoul National University Hospital, South Korea.

The demographic and clinical characteristics of all subjects are presented in Table 1.

Baseline BMI was not significantly associated with global Aβ deposition change during the two-year follow-up period for models 1 and 2. In contrast, a lower baseline BMI was significantly associated with a greater increase in tau deposition in the AD-signature region over two years (Table 2). When we conducted the same analyses including three BMI strata (below -1 SD, median BMI, above 1SD) instead of BMI as a continuous variable for the purpose of demonstration, the results were similar (Fig. 1 and Table 3). We also performed sensitivity analyses, including the GDS score, smoking status, alcohol intake status, and lifetime physical activity as additional covariates, which showed similar findings (Table 4).

A lower baseline BMI was associated with increased tau deposition over two years in men, but not in women (Table 5). As for Aβ changes, neither women nor men showed significant association between baseline BMI and cerebral Aβ changes over two years.