Peifeng Hu
· Clinical ProfessorUniversity of California, Los Angeles · Geriatrics and Gerontology
Active 1991–2026
About
Peifeng Hu is a HS Clinical Professor in the Department of Medicine at UCLA. His research activities focus on aging, neurodegenerative biomarkers, cognitive decline, and the relationships between inflammation, metabolic factors, and aging-related health outcomes. He has contributed to longitudinal studies such as the LASI-Biomarkers and LASI-DAD, investigating neurodegenerative biomarkers, cognitive function, and health disparities in older populations in India and the United States. His work includes examining the associations between blood-based biomarkers and cognitive decline, diabetes prevalence, and socioeconomic factors affecting health in aging populations. Dr. Hu has also been involved in studies on inflammation, hemolysis, and their impact on aging and neurodegenerative diseases, as well as the validation of biomarker assays. His research aims to improve understanding of aging processes and health disparities, with a particular focus on neurodegeneration, inflammation, and metabolic health in diverse populations.
Research topics
- Medicine
- Gerontology
- Computer Science
- Data Mining
- Psychology
- Psychiatry
- Demography
- Environmental health
- Physical therapy
- Geography
- Internal medicine
- Medical education
Selected publications
Journal of the American Geriatrics Society · 2026-02-11
articleOpen accessBACKGROUND: The Longitudinal Aging Study in India-Diagnostic Assessment of Dementia (LASI-DAD) is a nationally representative study of cognitive aging and dementia in older adults. This study aims to investigate the longitudinal relationship between neurodegenerative biomarkers such as amyloid-beta (Aβ) 42/40 ratio, glial fibrillary acidic protein (GFAP), neurofilament light chain (NfL), phosphorylated tau (pTau), and total tau, and cognitive decline among older adults in India. METHODS: Data from 1181 participants (aged ≥ 60 years) who completed both Wave 1 (2017-2019) and Wave 2 (2022-2024) of LASI-DAD with biomarker data were analyzed. General cognitive factor score was used as a comprehensive measure of cognitive performance, and biomarkers including NFL, Aβ42/Aβ40 ratio, GFAP, pTau, and total Tau were measured using Simoa technology. Generalized Estimating Equations (GEE) were used to analyze the association between change in cognition and biomarkers, adjusting for demographic covariates and comorbidities. RESULTS: , p = 0.002) were associated with greater cognitive decline over time. Domain-specific analysis showed GFAP and NfL were linked to language and memory decline; Aβ42/Aβ40 ratio was associated with language decline. CONCLUSIONS: Our findings suggest that biomarkers, including GFAP and NfL, are associated with cognitive decline over time in older adults in India. These biomarkers may serve as important indicators for monitoring cognitive aging and dementia risk in this population.
HbA1c alone might not reliably indicate India's diabetes prevalence – Authors' reply
The Lancet Global Health · 2026-01-07
articleOpen accessLovely Gupta and colleagues question the use of HbA1c in the Longitudinal Aging Study in India (LASI). We wish to make three points. First, we did not “concede … that HbA1c-only diagnosis is flawed”, and we disagree with the letter's suggestion that we did not meaningfully discuss HbA1c-related measurement issues. In our Discussion, we devoted a paragraph of 350 words, including 14 references, to the strengths and limitations of our use of HbA1c. We refer readers to this text. Additionally, allow us to clarify that HbA1c is largely unaffected by common haemoglobinopathies when measured by modern laboratory methods, such as the assay used in LASI.
Journal of the American Geriatrics Society · 2026-02-10
articleBACKGROUND: Cognitive aging and dementia are major public health challenges in India's aging population. This study examines associations between cognition and neurodegenerative biomarkers among community-dwelling older adults using data from the Longitudinal Aging Study in India-Diagnostic Assessment of Dementia (LASI-DAD) Wave 2. METHODS: LASI-DAD Wave 2 included 4635 participants aged ≥ 60 years. Cognitive assessments covered memory, executive function, language, and visuospatial abilities. Age-, sex-, and education-adjusted cut-offs identified low cognitive performance. Blood biomarkers, including total tau, phosphorylated tau-181 (pTau-181), neurofilament light chain (NfL), glial fibrillary acidic protein (GFAP), and β-amyloid (Aβ42/Aβ40 ratio) were collected and log-transformed. Generalized linear model (GLM) adjusted for demographic and clinical variables were used to examine the associations between biomarker levels and poor cognitive performance. RESULTS: Median participant age was 70 years (IQR: 66-76); 54.6% were female. Poor cognitive performance was present in 828 (27.6%) participants. Higher NfL (Median 27.2 vs. 21.8 pg/mL, p < 0.001), GFAP (Median 120.4 vs. 102.9 pg/mL, p < 0.001) and p-Tau 181 (Median 37.5 vs. 35.4, p = 0.003) were associated with poor cognition. In GLM, NfL (RR: 1.411, 95% CI 1.162-1.714, p = 0.001), GFAP (RR: 1.425, 95% CI 1.147-1.769, p = 0.003) were associated with poor cognition, while Aβ42/Aβ40 ratio total tau and p-Tau 181 were not (adjusted for age, sex, body mass index, serum creatinine, hypertension, diabetes, heart disease and stroke). Additionally, higher NfL and GFAP were associated with poor performance in most cognitive domains, except recognition memory. Higher Aβ42/Aβ40 ratio with poor recognition memory, higher p-Tau 181 with lower orientation scores, and lower total tau with lower recognition memory and visuospatial score. CONCLUSION: Among older Indian adults, elevated NfL and GFAP levels were strongly associated with poor cognitive performance. These biomarkers could aid future dementia screening and prognostication efforts. Longitudinal studies are needed to validate their role in large-scale screening programs.
The Lancet Global Health · 2025-08-07 · 14 citations
articleOpen accessBACKGROUND: India is a country of 1·4 billion people that contributes to much of the global diabetes burden. Updated evidence on the state of the diabetes epidemic among middle-aged and older adults is imperative given that the risk of diabetes increases with age and that clinical and public health interventions can prevent diabetes complications. We aimed to estimate the prevalence, awareness, treatment, and control of diabetes in a nationally representative and state-representative sample of Indians aged 45 years and older. METHODS: <7·0%), (3) proportion with blood pressure control (measured blood pressure <140/90 mm Hg), and (4) proportion self-reporting use of lipid-lowering medications. Outcomes were assessed in the national sample; by state and union territory; and across individual-level characteristics of age, sex, rural versus urban area of residence, education, economic status, and BMI. FINDINGS: Diabetes prevalence among adults aged 45 years and older in India was 19·8% (95% CI 19·4-20·2), which amounted to 50·4 million people (49·4-51·4). Prevalence among men and women was similar (men, 19·6% [95% CI 19·0-20·2] and women, 20·1% [19·5-20·6]). Urban diabetes prevalence (30·0% [95% CI 29·1-30·8]) was approximately twice as high as rural prevalence (15·0% [14·6-15·5]). States with higher levels of economic development tended to have greater age-standardised prevalence (standardised regression coefficient for gross domestic product per capita 0·65 [95% CI 0·45-0·85]). Overall, 60·1% (59·0-61·2) of individuals were aware of their diabetes. Of individuals with diagnosed diabetes, 45·7% (44·3-47·2) achieved glycaemic control, 58·9% (57·5-60·4) achieved blood pressure control, and 6·4% (5·8-7·2) were taking a lipid-lowering medication. INTERPRETATION: Our findings emphasise the urgent need to scale up policies to better prevent, detect, manage, and control diabetes among middle-aged and older adults in India. FUNDING: US National Institute on Aging; Ministry of Health and Family Welfare, Government of India.
Practical Laboratory Medicine · 2025-01-20 · 2 citations
articleOpen accessIntroduction: Hemolysis is a known interference factor that has been found to show erroneous effect. Present study analyzes the impact of hemolysis on the concentrations of protein biomarkers of Alzheimer's disease (Aβ42, t-Tau, p-Tau181) along with novel proteins which are currently under investigation (SIRT1,SIRT2,SIRT6,FOXO3A, NFL, Aβ40, GFAP). Methods: Plasma samples were grouped into two categories: hemolyzed and non-hemolyzed groups. Degree of hemolysis (in percentage) was separately analyzed using Single molecule array (SIMOA) technology. Quantitative analysis for hemolyzed and non-hemolyzed samples were done using surface plasmon resonance (SPR) technology. Results: The SIMOA analysis indicated that at high levels of hemolysis (1000 mg/dL) there was an increase in NFL protein level up to approximately 30 % whereas p-Tau181 did not show much interference even at higher hemolysate concentration. Aβ40, Aβ42 and GFAP showed modest effect up to hemolysis of 250mg/dL-500 mg/dL. SPR analysis of total Tau (t-Tau), p-Tau181, SIRT1, SIRT6 showed the consistency in the result and there was no significant difference in hemolyzed plasma compared to non-hemolyzed samples. Aβ42 and FOXO3A showed decline in hemolyzed plasma compared to non-hemolyzed samples (4.34 ± 0.18ng/ul; 4.95 ± 0.19ng/ul) and (3.83 ± 0.34ng/ul; 5.12 ± 0.46ng/ul), respectively whereas, a significant increase in the concentration was observed for SIRT2; 2.4 ± 0.10ng/ul in hemolyzed compared to 1.30 ± 0.22ng/ul in non-hemolyzed group. Conclusions: High grade hemolysis leads to altered protein concentration associated with neurodegeneration. Present study emphasizes the need to have pre-analytical inspection for hemolysis detection especially in a multicentric biomarker study.
American Journal of Epidemiology · 2025-08-16 · 4 citations
articleOpen accessThis study examines associations between blood-based neurodegenerative biomarkers and cognitive functioning in nationally representative samples of older adults in India and the United States. Using data from the Longitudinal Aging Study in India-Diagnostic Assessment of Dementia (LASI-DAD) and the Health and Retirement Study (HRS), we analyzed 4 biomarkers-the ratio of Amyloid beta 42-40 (Aβ42/Aβ40), Glial Fibrillary Acidic Protein (GFAP), Neurofilament Light Chain (NfL), and Phosphorylated Tau at Threonine 181 (pTau181)-in relation to cognitive outcomes. Higher levels of GFAP and NfL were associated with worse cognitive function and a greater likelihood of dementia in both populations. GFAP and NfL were also associated with cognitive decline in the HRS, but not in LASI-DAD. Higher Aβ42/Aβ40 ratio was associated with worse cognitive functioning and more dementia in HRS; on the other hand, higher levels of the Aβ42/Aβ40 ratio were significantly linked with better cognitive functioning in LASI-DAD. pTau181 was generally not associated with cognitive functioning in either country. The associations of the whole set of markers with subsequent mortality suggest they may serve as markers for general aging. This article is part of a Special Collection on Cross-National Gerontology.
Frontiers in Epidemiology · 2025-07-25
articleOpen accessIntroduction: In low- and middle-income countries, self-reported data on chronic cardiometabolic conditions such as high blood pressure and diabetes are commonly used in large-scale epidemiologic studies because implementing objective measures is challenging in these contexts. However, existing evidence suggests that the sensitivity of such measures may be low, and performance may differ by factors such as age, education, or income. We sought to confirm these prior findings and assess bias due to the use of self-reported data in hypothetical epidemiologic studies considering high blood pressure and diabetes as exposures, outcomes, and confounders. Methods: = 55,392) to assess the performance of self-reported data on high blood pressure and diabetes compared with objective measures, overall and stratified by basic demographic factors. We then compared regression coefficients from models considering self-reported and objective high blood pressure and diabetes as exposures, outcomes, and confounders. In all models, we examined whether the mode of data collection (self-report or objective) for other key variables in the model affected results. Results: The overall sensitivity of self-reported high blood pressure and diabetes was 0.514 and 0.570, respectively; specificity for the two conditions was 0.922 and 0.984. Sensitivity of both conditions increased with age, and was higher among women, those in urban settings, and those with higher educational attainment. Across almost all models considering high blood pressure and diabetes as either exposures or outcomes anti-conservative bias was observed when using self-reported vs. objective measures, regardless of the mode of data collection for other key variables. When high blood pressure and diabetes were considered as confounders, differences between using self-report and objective measures were minimal. Discussion: Anti-conservative bias due to the use of self-reported measures of chronic cardiometabolic conditions in surveys conducted in low- and middle-income contexts may be common. Future studies may seek to quantify the magnitude of anticipated bias in existing data resources and use quantitative bias analysis to formally estimate the potential implications of misclassification.
Neurobiology of Disease · 2025-09-17 · 2 citations
articleOpen accessBACKGROUND: In India, studies of anaemia in older populations are lacking despite the adverse effect on cognitive function and dementia. The Longitudinal Ageing Study in India-Harmonized Diagnostic Assessment of Dementia (LASI-DAD) dataset contains detailed measures to allow better understanding of anaemia as a potential risk factor for dementia. METHOD: Linear regression was applied in the LASI-DAD cohort (n = 2758) between blood measures (including anaemia and haemoglobin concentration (g/dL)) with 11 cognitive tests. All models were adjusted for age and gender (full model includes rural location, education, smoking, region, BMI and population weights). The USA-based Health and Retirement Study (HRS) cohort (n = 5720) was used to replicate associations between blood and global cognition. RESULTS: In LASI-DAD, we showed an association between anaemia and poor memory (p-value = 0.0054). We found a positive association between haemoglobin concentration and ten cognitive tests (β = 0.041-0.071,p-value<0.05). The strongest association with haemoglobin was identified for memory-based tests (β = 0.061-0.071,p-value<0.005). Positive associations were shown between the general cognitive score and red blood count tests including mean corpuscular haemoglobin concentration (MCHC,β = 0.06,p-value = 0.0001) and red cell distribution width (RDW,β = -0.11,p-value<0.0001). In the HRS, associations were replicated between general cognitive score and other blood count tests (Red Blood Cell, MCHC, RDW, p-value<0.05). CONCLUSIONS: We have established in a South Asian population that low haemoglobin and anaemia are associated with low cognitive function, therefore indicating that anaemia could be an important modifiable risk factor for dementia. We have validated this result demonstrating both the variability of this risk factor cross-nationally and its generalizable association with cognitive outcomes.
Alzheimer s & Dementia · 2025-12-01
articleOpen accessBACKGROUND: Alzheimer's disease (AD) is characterized by abnormal blood levels of specific proteins associated with amyloid plaques, tau pathology, and neurodegeneration, which are considered biomarkers of AD. To better understand the genetic architecture of AD in South Asians, we assessed the association between single nucleotide polymorphisms (SNPs) previously associated with AD and AD biomarkers with blood-level measures of seven AD biomarkers in 2,547 South Asians from the Harmonized Diagnostic Assessment of Dementia for the Longitudinal Aging Study in India (LASI-DAD). METHOD: We performed SNP-based association analysis between 152 AD- and 9 AD biomarker-associated SNPs with 7 AD biomarkers, including amyloid beta 40 (Ab40), amyloid beta 42 (Ab42), Ab42/Ab40 ratio, total tau (tTau), tau phosphorylated at threonine 181 (pTau181), glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL). The model adjusted for age, sex, the first 10 genetic principal components (PCs), and incorporated the genetic relationship matrix and plate as random effects. Significance was declared at Benjamini-Hochberg FDR q<0.1. Identified loci were followed up with haplotype analysis to understand the potential driving factors. RESULT: Among 152 AD-associated SNPs, rs199515_C (WNT3/MAPT) was associated with tTau in the expected direction (beta=0.21, p = 0.00052), whereas rs2718058_A (NME8) was associated with pTau181 in the unexpected direction (beta=0.13, p = 0.00057). Among the 9 biomarker-associated SNPs, rs242557_A (MAPT) was associated with tTau (beta=-0.139, p = 2.3e-08) and pTau181 (beta=0.099, p = 1.4e-04), and rs429358_C (APOE-e4) was associated with Ab42/Ab40 ratio (beta=-0.141, p = 0.0012). Follow-up analysis in MAPT, the gene that encodes the tau protein, identified a significant association between H1/H2 haplotype and tTau (global-p=0.00205). Specifically, H1j (p = 0.0015), H2 (p = 0.027), H1r (p = 0.031), and H1b (p = 0.041) were associated with lower levels of tTau, whereas H1c (p = 0.04) and H1o (p = 0.012) were associated with higher levels. Moreover, APOE-e4 stratified analysis detected significant H1j-tTau (score=-3.61, p = 0.00031) and H1c-pTau181 (score=2.55, p = 0.01) associations in e4 carriers only. CONCLUSION: The analysis highlights a critical role of MAPT in tau pathology in South Asians. The observed association does not seem to be fully driven by known risk haplotypes. It also provides evidence to suggest modification by APOE-e4. Further investigation of MAPT haplotype structure in this population is warranted.
Alzheimer s & Dementia · 2025-12-01
articleOpen accessBACKGROUND: Alzheimer's disease (AD) and related dementias are a growing health and economic burden in India, the most populous country in the world. Blood levels of amyloid beta (Aß), tau, and other proteins marking neuronal injury are biomarkers of AD risk. Understanding genetic associations with blood biomarkers of AD in South Asians may help lead to improved prevention and treatment in this understudied population. METHODS: We performed gene-based analyses on 1) missense/loss-of-function single-nucleotide variants (SNVs) and 2) promoter/brain-specific enhancer SNVs across 84 genes previously associated with AD and 22 genes previously associated with AD biomarkers in the Harmonized Diagnostic Assessment of Dementia for the Longitudinal Aging Study in India (LASI-DAD; N = 2,545) with the variant-Set Test for Association using Annotation infoRmation (STAAR) across 7 AD biomarkers measured in blood: Aß40, Aß42, Aß42/Aß40, neurofilament light chain (NfL), glial fibrillary acidic protein (GFAP), tau phosphorylated at threonine 181 (pTau), and total tau (tTau). We adjusted for age, sex, and 10 genetic ancestry principal components, with random intercepts for genetic relatedness and biomarker plate. Analyses incorporated weighted annotation scores (e.g., deleteriousness). Significant results (FDR q<0.1) were followed up with single variant analysis to determine the SNVs driving the associations. RESULTS: We found that 5 genes previously associated with AD (ICA1L, SCIMP, APOE, TSPOAP1, CLU) and 1 gene previously associated with AD biomarkers (APOE) were associated with pTau, Aß42/Aß40, and tTau in missense/LoF analysis in LASI-DAD (FDR q<0.1). The top SNVs of ICA1L, SCIMP, and TSPOAP1 (rs201355940, 17:5223434:C:T, rs144482724) were rare in LASI-DAD (MAFs<0.005) and had high CADD Phred scores (11.2-36.0) indicative of deleteriousness. In promoter/enhancer analysis, 1 AD-associated gene (CCDC6) and 3 AD biomarker-associated genes (CCK, APOC1, SLIT3) were associated with Aß40, Aß42/Aß40, and/or tTau. The top SNV in CCDC6 (rs1171833) had a CADD Phred score of 19.6 and was common in LASI-DAD (MAF=0.27). Several associated variants appeared to have an ancestral origin unique to India. CONCLUSIONS: Rare and common variants in AD- and AD biomarker-associated genes may impact blood levels of AD biomarkers in South Asians. Gene-based analysis may aid in identifying rare deleterious variants not seen in other populations.
Recent grants
NIH · $392k · 2014
NIH · $517k · 2010
Frequent coauthors
- 23 shared
Teresa E. Seeman
University of California, Los Angeles
- 23 shared
Jinkook Lee
University of Southern California
- 14 shared
Gail A. Greendale
University of California, Los Angeles
- 12 shared
Eileen M. Crimmins
University of Southern California
- 10 shared
David B. Reuben
University of California, Los Angeles
- 10 shared
Shelley E. Taylor
General Electric (United States)
- 10 shared
Laura Cousino Klein
Brigham and Women's Hospital
- 9 shared
Joseph A. Loo
University of California, Los Angeles
Education
- 2006
Ph.D., Molecular, Cellular, and Integrative Physiology
University of California, Los Angeles
- 2002
M.S., Molecular, Cellular, and Integrative Physiology
University of California, Los Angeles
- 1999
B.S., Biological Sciences
University of Science and Technology of China
- Resume-aware match score
- Save to shortlist
- AI-drafted outreach
See your match with Peifeng Hu
PhdFit ranks faculty by your research interests, methods, and publications — grounded in their actual work, not templates.
- Free to start
- No credit card
- 30-second signup