Design, Methodology, and Preliminary Results of the Follow-up of a Population-Based Cohort Study in Rural Area of Northern China: Handan Eye Study
The Handan Eye Study (HES) represents a critical population-based cohort investigation focusing on the epidemiology of ocular diseases among rural populations in northern China. Conducted in Yongnian County, Hebei Province, the study addresses the scarcity of data on major eye conditions in rural China, where over half of the nation’s population resides. While previous studies have primarily emphasized urban populations, HES aims to fill this gap by providing longitudinal insights into the incidence, progression, and risk factors of vision-threatening diseases such as glaucoma, cataract, age-related macular degeneration (AMD), diabetic retinopathy (DR), and refractive errors. This article details the design, methodology, and preliminary findings from the follow-up phase of HES, conducted between 2012 and 2013, six years after the baseline survey.
Study Design and Methodology
HES follow-up retained the population-based cohort design of its baseline phase. The original cohort comprised 6,830 adults aged ≥30 years recruited in 2006 through cluster sampling across 13 villages in Yongnian County. The follow-up phase aimed to evaluate the 6-year cumulative incidence of ocular diseases, assess disease progression, and identify risk factors for visual impairment and mortality.
Participant Recruitment and Follow-up Rate
Of the 6,830 baseline participants, 5,394 (85.3%) completed the follow-up examinations. Exclusions included 507 deceased individuals and 929 lost to follow-up. Reasons for loss included work-related absence (53.4%), refusal (27.0%), severe physical/mental illness (12.6%), and loss of contact (6.8%). The high follow-up rate minimized potential attrition bias, though differences between groups were analyzed to assess data robustness.
Data Collection and Measurements
Data collection involved comprehensive questionnaires and ocular examinations administered by trained personnel. Key components included:
- Questionnaires: Expanded from baseline to include socio-demographics, medical history, behavioral factors, and validated scales: Mini-Mental State Examination (MMSE), EuroQol-5D, SF-8, near-vision-related quality of life (NVR-QOL), and a 15-item visual quality questionnaire.
- Ocular Examinations: Autorefraction, best-corrected visual acuity (BCVA), intraocular pressure (IOP), anterior segment optical coherence tomography (OCT), slit-lamp biomicroscopy, fundus photography, and visual field testing. New additions included color vision assessment, near-vision measurement, and trachoma screening.
Disease Definitions
Standardized diagnostic criteria were applied:
- Glaucoma: Based on optic disc morphology, visual field defects, and IOP >21 mmHg.
- Cataract: Defined using the Lens Opacities Classification System III (LOCS III).
- AMD: Graded using criteria from the Blue Mountains Eye Study, distinguishing early and late stages.
- Diabetic Retinopathy: Classified via fundus photography using a modified Early Treatment Diabetic Retinopathy Study (ETDRS) scale.
- Refractive Changes: Myopic or hyperopic shifts were defined as spherical equivalent refraction (SER) changes of ≥0.5 diopters (D).
Statistical Analysis
Analyses compared three groups: follow-up (n=5,394), loss to follow-up (n=929), and deceased (n=507). Continuous variables were analyzed using ANOVA or Kruskal-Wallis tests; categorical variables used chi-square tests. Multivariate logistic regression identified mortality risk factors. Inter-class correlation coefficients (ICCs) and Kappa statistics evaluated inter-operator consistency, while Cronbach’s alpha and cumulative variance assessed scale reliability and validity.
Key Findings
Baseline Characteristics and Group Comparisons
The deceased group was significantly older (66.5 ± 10.3 years) than the follow-up (51.4 ± 11.1 years) and loss-to-follow-up groups (50.2 ± 14.0 years) (P<0.001). Males constituted 59.0% of the deceased, compared to 44.6% and 49.3% in the other groups (P<0.001). The deceased also had lower educational attainment (65.9% completed middle school vs. 85.5% in follow-up), lower BMI (24.05 vs. 24.55 kg/m²), worse BCVA (0.37 ± 0.57 logMAR vs. 0.08 ± 0.28), and higher prevalence of hypertension (36.7%), diabetes (8.2%), and heart disease (10.9%) (P<0.05 for all).
Factors Associated with Mortality
Multivariate analysis identified age (OR=1.091 per year, 95% CI: 1.074–1.108), male gender (OR=0.317, 95% CI: 0.224–0.448), and poorer BCVA (OR=0.282, 95% CI: 0.158–0.503) as significant predictors of mortality. These findings underscore the interplay between aging, gender, visual impairment, and systemic health in rural populations.
Loss-to-Follow-up Bias Assessment
The follow-up and loss-to-follow-up groups differed minimally in age, gender, BMI, systolic blood pressure, and SER, but showed no significant differences in BCVA, IOP, or axial length. This suggested low risk of attrition bias affecting ocular disease estimates.
Data Quality and Reliability
- Scale Validation: Cronbach’s alpha for EuroQol-5D, SF-8, NVR-QOL, visual quality, and MMSE scales ranged from 0.63 to 0.90, with cumulative variances ≥0.61, confirming reliability.
- Inter-Operator Consistency: ICCs for IOP, BCVA, and refractive measurements exceeded 0.80. Kappa coefficients for glaucoma, AMD, and DR diagnoses were ≥0.69, indicating high diagnostic agreement.
Discussion
The HES follow-up achieved a high retention rate (85.3%) and rigorous quality control, ensuring robust data for longitudinal analysis. The study’s focus on rural populations addresses a critical gap in ocular epidemiology, particularly for aging-related diseases like cataract and AMD. The deceased group’s distinct profile highlighted socio-demographic and health disparities influencing mortality, emphasizing the need for targeted healthcare interventions.
Strengths and Limitations
Strengths include the large sample size, standardized protocols, and comprehensive disease definitions. Limitations include the exclusion of younger adults (<36 years) and potential underrepresentation of rare diseases. However, the study’s design facilitates future analyses of incidence trends and risk factors.
Conclusion
The HES follow-up provides valuable insights into the ocular health of rural Chinese adults, demonstrating the feasibility of longitudinal studies in resource-limited settings. By linking systemic health factors with visual outcomes, the study informs public health strategies to reduce preventable blindness and mortality. Future analyses will explore disease-specific incidence rates and refine risk prediction models.
doi.org/10.1097/CM9.0000000000000418
Was this helpful?
0 / 0