Association of Triglyceride to HDL-Cholesterol ratio with Cognitive impairment among elderly people: A cross-sectional study in Southwestern Uganda
Description
This dataset was generated as part of a community-based cross-sectional study conducted in Rukiga District, Southwestern Uganda, from February to March 2025. The study aimed to investigate the association between the triglyceride to high-density lipoprotein cholesterol (TG/HDL-C) ratio and cognitive impairment among elderly individuals aged 60 years and above. Our central hypothesis was that elevated TG/HDL-C ratio, a marker of metabolic dysregulation, is independently associated with cognitive impairment in older adults. Data were collected from 272 participants selected via multistage cluster sampling. Each participant underwent a standardized cognitive assessment using the Mini-Mental State Examination (MMSE), with scores below 24 indicating cognitive impairment. In total, 183 individuals (67.3%) screened positive, with 95 having mild and 88 severe impairment. The dataset includes variables across multiple domains: sociodemographic factors (e.g., age, gender, education level), clinical conditions (e.g., diabetes, hypertension, depression), anthropometric data (e.g., waist circumference, BMI), lifestyle behaviors (e.g., smoking, alcohol use, physical activity), and laboratory-measured serum lipid profiles (TG, HDL-C, LDL-C, TC), collected after 8–12 hours of fasting. TG/HDL-C ratio and other lipoprotein ratios were computed. Blood samples were processed using enzymatic methods on a HumaStar 200 analyzer with appropriate quality controls. Statistical analysis included descriptive statistics, nonparametric tests, and logistic regression. Notably, a high TG/HDL-C ratio was independently associated with cognitive impairment after adjusting for confounders (aOR for highest tertile = 3.38; p = 0.031). Age, lower educational attainment, and depression were also significant predictors. This dataset provides crucial insight into the metabolic and psychosocial determinants of cognitive impairment in a low-resource setting. It is particularly valuable for researchers exploring blood-based biomarkers for early detection of neurocognitive disorders and can inform public health interventions targeting aging populations in sub-Saharan Africa. The dataset is structured to enable both univariate and multivariate analyses and is suitable for epidemiological, clinical, or machine learning-based investigations into geriatric cognitive health.
Files
Steps to reproduce
The dataset was obtained through a community-based cross-sectional study conducted from February 15 to March 29, 2025, in Rukiga District, Southwestern Uganda, targeting individuals aged 60 years and above. Participants were selected using the WHO cluster probability sampling method with support from Village Health Teams (VHTs). Data were collected using a structured researcher-administered questionnaire to capture sociodemographic, behavioral, and clinical variables. Cognitive function was assessed using the Mini-Mental State Examination (MMSE), with scores <24 indicating impairment. Depression and physical activity were measured using the Patient Health Questionnaire-9 (PHQ-9) and the International Physical Activity Questionnaire Short Form (IPAQ-SF), respectively. Anthropometric measurements (waist and hip circumferences) were taken using a Seca Ergonomic Measuring Tape, and blood pressure was recorded using a digital sphygmomanometer. After overnight fasting (8–12 hours), venous blood samples were collected in red-top vacutainers, transported under cold conditions, centrifuged at Mparo Health Centre IV, and analyzed for lipid profiles (TG, TC, HDL-C, LDL-C) using enzymatic methods on the HumaStar 200 automated analyzer with appropriate quality control measures. Data entry was done in Microsoft Excel, cleaned, and exported to Stata version 17 for analysis. Statistical procedures included normality testing (Shapiro-Wilk), Kruskal-Wallis tests, and logistic regression to explore associations. These protocols, instruments, and workflows were standard and reproducible, allowing replication in similar low-resource geriatric research settings.
Institutions
- Mbarara University of Science and Technology Faculty of Medicine