The Nutrition and Health in Southwest China (NHSC) study: design, implementation, and major findings

China faces growing burdens from noncommunicable diseases (NCDs) amid rapid urbanization and westernization. Unhealthy diet and lifestyle are key contributors to NCDs, and Asians may have greater genetic susceptibility to type 2 diabetes than Western populations. Ethnic differences in NCD risk may reflect genetic susceptibility and interactions between genetic background and environmental factors such as diet and lifestyle. Few epidemiologic studies in China have jointly investigated environmental factors, genetic factors, and their interactions on NCD risk. The Nutrition and Health in Southwest China (NHSC) study was initiated to address this gap by prospectively examining how diet, lifestyle, genetic background, and their interactions influence obesity, diabetes, hypertension, cardiovascular diseases, and other NCDs among Chinese adults in Southwest China.

Study design: The NHSC is a population-based prospective cohort initiated in winter 2013 in Southwest China (Sichuan, Guizhou, Yunnan; region ~0.49 million km², ~144.1 million residents). Sampling was stratified by urban/rural location across 54 sites (23 communities, 31 villages) as of December 2018. At each site, a two-stage sampling (household then person) was used. From official household lists, 150 households were randomly selected. Eligible participants were adults aged 18–70 years who had lived at their current residence for at least 1 year. If a selected household member refused or was unavailable, a replacement household of similar composition was randomly selected within the site to maintain sample size and representativeness for chronic disease prevalence and distributions of age, gender, education, and economic status. Ethics approval was obtained from the Ethics Committee of Sichuan University; all participants provided written informed consent. Follow-up: Baseline data collection concluded in December 2018. Follow-up visits occur every 2 years. By the time of reporting, at least one follow-up had been completed for those enrolled in 2013–2016, and two follow-ups for those enrolled in 2013–2014. Data collection domains and instruments: - Questionnaires: demographics and socioeconomic status (age, residence, marital status, family size, personal monthly income, annual family income, education, employment), family history of chronic diseases, reproductive history (menarche, menopause), diet, eating behaviors, physical activity, sedentary behaviors, sleep, smoking, perceived work/life stress, depression (since 2015; HDRS-17), and mental health/cognition (since 2017; Mini-Mental State Examination). - Nutrition assessment: three 24-hour dietary recalls administered face-to-face by trained interviewers—one on registration day and two participant-selected nonconsecutive days within 10 days. Detailed recipes, food types/brands were recorded. Nutrient intakes were calculated via an in-house database based on China Food Composition tables. A 66-item interviewer-administered food frequency questionnaire (FFQ) assessed usual intake over the past year with frequencies from never to >5 times/day using standard serving sizes. Nutrient intakes were computed by multiplying consumption frequency by nutrient content and summing across items. - Eating behaviors: frequency, timing, and location of breakfast/lunch/dinner; foods commonly consumed; frequency of family dinners; snacking frequency and types; frequency of eating away from home; local dietary specialties. - Physical activity and sedentary behavior: a validated 38-item questionnaire assessed usual type, frequency, and duration of activities over the past 12 months; transportation mode to work; time spent in sedentary behaviors (television, computer, smartphone, reading, playing cards/mahjong). MVPA energy expenditure (MJ/day) derived from activity data. - Lifestyle factors: smoking status (current, former, never), cigarettes/day and duration for smokers; sleep duration and quality via Pittsburgh Sleep Quality Index (19 items, 7 components), napping frequency/duration; perceived work and life stress frequency and intensity (response options from not at all to extremely stressful). - Anthropometry: measured by trained staff with participants lightly clothed and barefoot. Height and weight (nearest 0.1 cm and 0.1 kg) using an Ultrasonic Weight and Height Instrument (DHM-30). Waist circumference measured midway between the lowest rib and iliac crest; hip circumference at maximal gluteal protrusion (nearest 0.1 cm). Skinfolds measured on the right side (nearest 0.1 mm) using a Holtain caliper. All anthropometric measurements performed twice. - Blood pressure: after 5–10 minutes rest, two measurements of systolic and diastolic BP taken on the right upper arm using a mercury sphygmomanometer. - Biochemical assessments: fasting (≥10 h) venous blood drawn at each visit into tubes with sodium fluoride or EDTA. Samples centrifuged and stored at 4°C for analysis. Biomarkers include glucose metabolism (fasting plasma glucose, HbA1c, insulin, HOMA2-IR, HOMA2-β), lipid profile (triglycerides, total cholesterol, HDL, LDL), serum calcium, telomere length, C-reactive protein, and superoxide dismutase. - Genomics: genomic DNA extracted from peripheral blood lymphocytes; aliquoted into five tubes and stored at −80°C in different locations. Genotyping performed using Illumina Infinium II technology (Human HAP300 panel) to assess single nucleotide polymorphisms. Quality assurance: Standardized 3-day training for site staff; detailed operations manual; portion-size estimation aided by standard bowls/plates/glasses and a photo book of snacks/beverages. Participants received free clinical examination results within a week to encourage completion.

- Recruitment and response: 8612 adults invited; 686 declined (reasons: long duration or no interest); overall response rate 92%. Baseline cohort: 7926 adults completed questionnaires. - Demographics: Mean age 42.6 (SD 9.8) years; 53.9% female; 31.2% urban residents; 37.2% with >12 years of education (n=2948); 49.3% with annual family income >35,000 Yuan (n=3908). - Nutrition: Mean daily energy intake 1997 (525) kcal; macronutrient distribution: protein 16.6% (4.3), fat 30.6% (5.9), carbohydrate 52.8% (12.7). - Lifestyle: MVPA energy expenditure 2.7 (1.9) MJ/day; 62.8% reported recommended sleep duration (≥7 and ≤9 h/day); 21.6% current smokers (n=1712). - Anthropometry: Mean BMI 23.3 (2.7) kg/m²; overweight prevalence 41.7% (n=3305). Mean waist circumference 85.7 (9.6) cm; median percentage body fat 35.1% (Q1 31.2, Q3 42.6). - Biochemical and clinical: Fasting plasma glucose 5.3 mmol/L (Q1 4.6, Q3 5.7); HbA1c 5.5% (5.1, 5.8); insulin 6.2 µIU/mL (4.9, 9.8); HOMA2-IR 0.9 (0.6, 1.5); prediabetes prevalence 39.6% (n=3138); systolic BP 125.2 (15.6) mmHg; diastolic BP 75.3 (17.1) mmHg; total cholesterol 4.5 (4.1, 5.6); HDL-C 1.2 (1.1, 1.5); LDL-C 3.2 (2.8, 3.8); triglycerides 1.2 (0.8, 1.8); C-reactive protein 3.0 (1.3, 5.2) mg/L; superoxide dismutase 146.1 (19.2) U/mL; serum calcium 2.4 (2.2, 2.5) mmol/L; serum 25(OH)D 20.6 (15.0, 27.8) ng/mL. - Baseline analytical findings: Higher dietary energy density was associated with greater BMI, fat mass index, fat-free mass index, and percentage body fat. Higher dietary glycemic index, glycemic load, or higher serum 25(OH)D levels were associated with less favorable glucose homeostasis. Lower time spent watching television and healthier eating patterns were linked to indicators of slower cellular aging (e.g., longer telomere length).

The NHSC cohort was designed to elucidate how diet, lifestyle, and genetic background interact to influence the risk of NCDs in a large, diverse adult population in Southwest China. Early analyses of baseline data support the study’s central hypotheses: higher dietary energy density relates to greater adiposity, and higher dietary glycemic index/load relates to poorer glucose homeostasis. The observed associations between sedentary behavior (television viewing), healthy dietary patterns, and cellular aging markers suggest behavioral pathways that may modulate biological aging and NCD risk. By integrating comprehensive phenotyping, dietary assessment, biomarker profiling, and genomic data with repeated follow-ups, the NHSC provides a robust platform to investigate causal pathways, identify high-risk subgroups, and explore gene–environment interactions relevant to obesity, diabetes, hypertension, and cardiovascular diseases in this population.

The NHSC cohort establishes a comprehensive, prospective resource to study nutrition, lifestyle, genomics, and their interactions in relation to NCDs among adults in Southwest China. Baseline findings indicate that dietary energy density and carbohydrate quality are associated with adiposity and glucose homeostasis, and that sedentary time and overall diet quality may be related to cellular aging. Future research will leverage longitudinal follow-up and genomic data to clarify temporal relationships, dissect gene–diet and gene–lifestyle interactions, and inform targeted prevention strategies for NCDs in Chinese populations.