Cohort profile: Zhejiang Healthy Aging Cohort Study (ZHACS)

The Zhejiang Healthy Aging Cohort Study (ZHACS) is an open representative cohort of elderly living in Zhejiang, China aged ≥60 years, which is designed focusing on mental and physical health and is multidisciplinary in orientation, involving the collection of economic, social, psychological, cognitive, daily behavioral, dietary, occupational, biological, and genetic data. The study commenced in 2014, and four waves of follow-up have been completed, with response rates of 88.44%, 77.01%, 91.95%, and 86.74%, respectively. Follow-up frequencies were once a year before 2017 and every 3 years since 2018. A total of 25 111 participants has been engaged in the study, with 17 317 individuals enrolled in three phases (2014–2015: n = 10 901; 2018: n = 4974; 2021: n = 1442), and 7794 new additions in each wave. As of the latest follow-up (2021–2023), 13 340 participants remain under active follow-up, of whom 3692 have been followed for 9 years without omission. Cohort data are stored at the Zhejiang CDC. Researchers interested in the ZHACS data can contact Xue Gu (E-mail: email protected). The world’s aging population is facing rising rates of chronic diseases, functional impairments, and multimorbidity, exacerbated by aging-related physiological decline and modifiable risk factors such as poor diet and inactivity 1. Socioeconomic challenges, including limited healthcare resources and insufficient social support, have also contributed to the worsening health outcomes of older adults 2. Numerous research projects have provided references for monitoring the health of older adults, such as the UK Biobank (UKB), the English Longitudinal Study of Ageing (ELSA), and a study on the health and living conditions of older people in Japan (AGES) 3–5. Chinese longitudinal cohort studies on this topic include the China Health and Aged Care Tracking Survey (CHARLS), West China Elderly Population Health Cohort Study, Hainan Centenarian Cohort Study, and Beijing Brain Health for the Elderly Initiative (BABRI) 6–8. Nonetheless, there is a paucity of representative cohorts focusing on older adults that aim to measure health issues prevalent among community-dwelling populations 9, 10. The Zhejiang Health Aging Cohort Study (ZHACS) was established in 2014 to represent a longitudinal examination of the older adults’ population in Zhejiang and was intended to supplement the extensive data for a large-scale cohort of seniors. This study aimed to ascertain the progression of age-related diseases to adverse outcomes such as dementia, frailty, depression, and all-cause mortality by estimating the rate of progression and determining any associated factors. Consequently, this study also aimed to provide policymakers with evidence to facilitate the development of more effective policies to promote healthy aging in China. The ZHACS was an open cohort that recruited adults aged 60 years and above from the general population in 12 study sites out of 90 cities, districts, and counties across Zhejiang Province, eastern China. Beyond the age requirement, individuals were eligible if they were registered as residents or had lived for ≥6 months in the area; they also had to be free of major disability, severe cognitive impairment, and profound hearing loss. Moreover, each year, eligible older adults were invited to join the cohort, with additions at least matching the number of deaths in the previous year, thereby maintaining cohort size. Each study site encompassed one to three subdistricts or towns, from which one to five communities or administrative villages were selected as clusters. Within these clusters, 17 Community Health Service Centers (CHSCs) formed the primary care network with registered residents and served as data-collection units (Supplementary Table S1, in the supplementary material). The cohort was purposely assembled from diverse regions across Zhejiang that exhibit contrasting disease patterns and exposure profiles, rather than being intended as a microcosm of the wider population 9. Study sites were selected to cover the central–northern plains (Tongxiang City, Yuecheng District, Yiwu City, Binjiang District, and Nanxun District), eastern–southern coast (Haishu District, Yuhuan City, Haiyan County, and Fenghua District), western–southern mountainous regions (Changshan County, Jingning She Autonomous County), and islands (Putuo District), which had robust infrastructure, including reliable death-reporting systems, broadband coverage, air-courier capacity for blood samples, and sustained administrative support. At each study site, the municipal or prefectural Centers for Disease Control and Prevention extracted a complete roster of eligible older adults from the Resident Health Records database for each selected cluster, totaling 24 346 individuals across 12 sites. A sample of 1500–1700 people per site (18 700 in total, 76.81% of the eligible population) was invited by letter and telephone, except for those in Putuo District, the easternmost tourist island, where only 800 eligible individuals were invited. In total, 17 317 older adults attended the baseline survey, with a response rate of 92.60%. Local community general practitioners from the CHSCs conducted computer-assisted face-to-face interviews that incorporated mental function assessments and physical measurements. Those who were bedridden, disabled, with hearing impairment or severe dementia, or who declined to respond to the survey were considered non-participants. The number of participants represented by ZHACS over time is shown in Fig. 1. Between 2014 and 2015, the cohort completed its baseline survey at seven study sites, and personal interviews were conducted with the participants once a year. Follow-up visits were conducted annually, and by 2017, two waves of follow-ups were completed. The study protocol was revised in 2018 with the addition of four study sites. Since annual follow-ups revealed little change in exposure factors, the follow-up interval was extended every 3 years and implemented at four sites each year. Two additional follow-up waves were conducted between 2018 and 2023 (Supplementary Table S2, in the supplementary material). The fifth wave of follow-up is currently ongoing. Between 2014 and 2023, the ZHACS completed the baseline survey and four follow-up waves, with a cohort of 25 111 participants. The response rate for each wave was defined as the proportion of eligible survivors at the start of the wave who either completed follow-up or had a documented outcome. For this calculation, eligibility was defined as membership not being known to have died or moved outside the study site. The response rates of four wave’s follow-up were 88.44%, 77.01%, 91.95%, and 86.74%, respectively. In the first follow-up wave, 85.82% (9355/10 901) participants were successfully interviewed face-to-face. In the second wave, 8069 participants were followed up, including 6639 who were followed up consecutively. As of the end of the third wave, a total of 10 681 participants had been followed up, with 5567 participants having completed three times’ follow-up assessments. By the conclusion of the fourth wave, a total of 13 340 participants remain under active follow-up, of whom 3692 have been followed for 9 years without omission. These were drawn from the original study sites enrolled at cohort inception in 2014–2015: Changshan (n = 734), Jingning (n = 655), Tongxiang (n = 989), Yuhuan (n = 1154), and Yuecheng (n = 160). Notably, during urbanization, some study-site villages were demolished, causing participants to either relocate permanently or be placed in temporary accommodations. This dispersion of the target population significantly complicated follow-up visits. Consequently, we relocated the project sites to neighboring villages with similar demographics. For participants who were not followed up face-to-face, we continued to collect electronic information on their medical appointments and deaths annually. To enhance data precision and validity, we are negotiating data-use agreements to enable linkage to Hospital Information Systems and medical insurance claims databases for the future phases of the cohort. In 2021, the Haishu site was officially dissolved following administrative redivision, resulting in 2485 participants no longer being followed up through face-to-face surveys, although mortality and healthcare utilization data continued to be synchronously captured. An adjacent administrative district within Ningbo City, the Fenghua District, was incorporated into the cohort, and baseline surveys were initiated with 1442 of the 1500 invited older adults. Loss to follow-up was defined as the absence of a follow-up questionnaire or physical examination attendance or of having a death certificate or out-migration record from an unspecified year onward. Table 1 compares baseline characteristics between the 2656 lost participants and the 18 924 who remained under follow-ups. Individuals who contributed only to the most recent baseline wave were excluded from the “follow-up” category. Age, sex, education, smoking, current alcohol use, health self-assessment, cognitive impairment, and depression symptoms were associated with loss to follow-up (all Ps ≤ .05). It was worth noting that Haishu district exhibited a notably high attrition rate (797/2485) due to the lack of Wave 4 follow-up data. Baseline characteristics of lost to follow-up participants versus those retained in the cohort. The total number (n = 21 580) did not include new participants recruited in Wave 4. The measurements obtained using ZHACS are presented in Table 2. A wide range of socio-demographics, menstruation and reproductive history, behavioral health, disease and medicine, and general health related data were included. Supplementary Table S3, in the supplementary material, provides a synopsis of the measures regarded as the core indicators of functioning, which were integral to one or more measurement cycles. Not all documented measures were recorded in each survey to accommodate the potential for new inquiries. For instance, we monitored a subset of respondents for up to 10 years, since a 3-year monitoring interval was insufficient to reflect the natural progression of cognitive decline, even if there was a bidirectional regression of cognitive functioning. Summary of questionnaire data in ZHACS study. Details of the measurements repeated in subsequent waves of data collection are listed in Table 3. In Waves 1 and 2, the participants completed the entire questionnaire and underwent a comprehensive physical examination. In Waves 3 and 4, the time-invariant items (e.g. education, menstrual, and reproductive history) were not repeatedly asked. New scales and physical function tests were introduced, and the original sleep items was replaced with a validated sleep quality questionnaire. Additionally, starting in Wave 3, participants who were recognized as cognitive impaired were asked to conduct the Alzheimer’s Disease Assessment Scale–Cognitive Subscale (ADAS-Cog) instead of the questionnaire at the current follow-up wave 11. Moreover, the Vital Registration System for the date and underlying cause of death (ICD-10) as well as medical records for inpatient and outpatient diagnoses from the municipal or prefectural Health Commission were linked to ZHACS deterministically using each participant’s 18-digit national ID number. Information on deaths and healthcare utilization during the preceding calendar year was extracted every December from local vital status and medical insurance registries. Measurements at each wave of data collection. Living participants enrolled before 2017 were extracted a single fasting blood sample in 2017. Individuals recruited from 2018 onward were bled once, at their baseline interview. PSQI, the Pittsburgh Sleep Quality Index; ADL, the Activity of Daily Living scale; FRAIL, the 5-item frailty scale; MMSE, the Chinese version Mini-Mental State Examination; PHQ-9, the Patient Health Questionnaire-9; ADAS-cog subscale, the Alzheimer’s Disease Assessment Scale-Cognitive Subscale. Where feasible, mental health, depression, physical symptoms, and sleep were measured using established assessments and questionnaires. Cognitive function was assessed using the Chinese version of the Mini-Mental State Examination (MMSE) 12, which evaluates time and place orientation, immediate and delayed recall, prospective memory, verbal fluency, letter cancelation, numerical ability, and health literacy. Depressive symptoms were assessed using Patient Health Questionnaire-9 (PHQ-9), using a cutoff score of ≥4 to define clinically relevant depressive symptoms 13. The Activity of Daily Living (ADL) scale was used to assess the ability to care for oneself in daily life 14. And a 5-item frailty scale, FRAIL, was conducted to evaluate participants for frailty 15. Moreover, the Pittsburgh Sleep Quality Index (PSQI) was used to evaluate sleep patterns 16. In addition, for those with cognitive impairment (MMSE scores of ≤17 for participants who were illiterate, ≤20 for those with primary education only, or ≤24 for people with higher than primary education), the ADAS-Cog subscale was given. All the questionnaire data were collected through computer-assisted personal interviews administered by trained interviewers. Physical examinations and performance tests were conducted by GPs during the same visit. A robust quality control system was in place and many interviewers met with the same respondents over several waves of data collection. Supplementary modules have been incorporated at various points in the history of ZHACS to address additional topics. On average, the interviews take approximately 60 min to complete. In 2017, a 6 ml fasting venous blood sample was collected from all living participants. From 2018 onward, a single blood sample was obtained from each newly enrolled cohort member. The collection rate was 93.6% for all participants. All specimens were centrifuged, aliquoted into plasma and blood-cell fractions, and stored at −80°C for long-term biobanking. Thirty peer-reviewed articles have been published using the ZHACS data, covering a broad range of issues. Here, we outline a selection of findings that illustrate some of the ways in which this study can be exploited. The baseline data were analyzed for 25 111 participants, with an average age of 67.81 ± 7.45 years, of whom 53.49% were women and nearly half were illiterate. The proportions of smoking and alcohol consumption among men were 40.11% and 43.93%, respectively; both were much higher than the rates for women. The prevalence of cognitive impairment and depressive symptoms was 14.25% and 8.39%, respectively (Table 4). The baseline characteristics of all participants. Total score of health self-assessment is 100. Chronic diseases are the predominant health burden among adults aged 60 years and older, with prevalence increasing significantly with age and showing notable sex disparities. The most common conditions included hypertension, cognitive impairment, diabetes, being overweight, and depressive symptoms 17–19. Multimorbidity increased markedly with age. In the 60–69 age group, prevalence was 20.26% in men and 26.79% in women, increasing to 32.47% and 43.12%, respectively, in those aged 70–79. This pattern underscores the growing complexity of care and increased healthcare utilization among older adults. Cognitive impairment was consistently more common in women than men across all the age groups, with overall prevalence rates of 16.56% and 11.60%, respectively (P < .001). The gap widened with age, reaching 43.02% in women versus 31.75% in men aged 80 and above. Depression also showed a female predominance, particularly in the 60–69 years range. However, this sex difference was diminished in the oldest cohort (80+ years), where the prevalence rates converged. Sarcopenia emerged as a key concern, affecting approximately 10% of adults aged 60 and above globally, with prevalence exceeding 30% among those over 80. This condition significantly compromises mobility, independence, and quality of life and represents a critical target for geriatric intervention 20. The ZHACS is at the forefront of longitudinal research concerning China’s aging population. Using a multidisciplinary approach, it evaluates both the health and social aspects of an individual’s life. This highlights the complex interplay between society and health through the assessment of symptoms, subjective perceptions, medical diagnoses, and biomarkers. Its compatibility with other age-related studies allows international comparisons. The advanced age of the study cohort resulted in a high prevalence of diseases and comorbidities, enabling an in-depth longitudinal analysis of health outcomes. These analyses were enhanced using objective health markers, physical and cognitive performance measures, and biological indicators. Similar to many general panel studies, ZHACS has some limitations. It lacks an in-depth exploration of the specific health outcomes and psychosocial processes seen in hypothesis-driven research. Ethnic minorities are underrepresented, and creating representative oversamples is costly. Attrition is an ongoing problem, and efforts are being made to address it. Since the study started when the participants were 60 years old, early life information was retrospective. Sample refreshment has been crucial to maintaining a proper proportion of 60-year-olds and to allow for cross-cohort comparisons. Finally, this study was confined to eastern China. Currently, this database is only available to a limited extent. Researchers interested in ZHACS are invited to contact Xue Gu (e-mail: email protected) to discuss potential collaborations and specific study ideas. Collaborators with reasonable requests will be granted access to the data upon successful application. This study was approved by the Zhejiang Provincial Center for Disease Control and Prevention Medical Science Research Ethics Committee. We would like to thank the participants. We would also like to thank the Zhejiang Health Commission for invaluable support. We thank all the staff from CDC, community health service centers, and township hospitals in all project sites. Chen Wu, Fan He, Yuanyuan Xiao, and Junfen Lin applied for original funding, planned the establishment of the cohort, and contributed to the ongoing development of the cohort. Fudong Li, Xue Gu, and Tao Zhang contributed to the ongoing development of the cohort. Le Xu performed the data management and analyses. Xinyi Wang and Tao Zhang contributed to data acquisition. Yujia Zhai, and Mengna Wu provided governance oversight. Xu Gu has operational oversight of the cohort and drafted the article. All authors were involved in the development of this article, including revision and final approval. Supplementary material is available at IJE online. None declared. The ZHACS is supported by grants from Zhejiang Provincial Research of China Science of China Zhejiang Provincial Science of China Science Research of Zhejiang and Zhejiang Provincial Research We used and to English and was for study data or access can be in the of the Where can out