The purpose of the current study was to investigate associations between spicy food intake and serum lipids levels in Chinese rural population.

Information on spicy food flavour and intake frequency was obtained using a two-item questionnaire survey. Dietary data were collected using a validated thirteen-item FFQ. Fasting blood samples were collected and measured for total cholesterol (TC), TAG, HDL-cholesterol and LDL-cholesterol. Multivariate linear and logistic regression models were employed to examine the relationship between spicy food and serum lipids levels according to the spicy food flavour and intake frequency, respectively.

A cross-sectional study in Henan Province.

38 238 participants aged 18–79 years old.

Spicy flavour and intake frequency were consistently associated with decreased TC and non-HDL-cholesterol levels but mildly associated with elevated TAG levels. Each level increment in spicy flavour was inversely associated with high TC (OR: 0·91; 95 % CI 0·88, 0·93) and high non-HDL-cholesterol (OR: 0·88; 95 % CI 0·85, 0·91) but positively associated with high TAG (OR: 1·04; 95 % CI 1·01, 1·07). Similarly, 1-d increment in spicy food intake frequency was also inversely associated with high TC (OR: 0·92; 95 % CI 0·91, 0·94) and high non-HDL-cholesterol (OR: 0·91; 95 % CI 0·89, 0·93) but positively associated with high TAG (OR: 1·04; 95 % CI 1·02, 1·06).

Spicy food intake was mildly associated with increased risk of abnormal TAG level, significantly associated with decreased risk of abnormal TC and non-HDL levels. Spicy food intake may be contribute to the management of lipid levels.

We aimed to investigate whether dietary patterns were associated with the risk of type 2 diabetes mellitus (T2DM) or pre-diabetes in adults of rural area in Henan.

Cross-sectional study. Principal component analysis was used to identify dietary patterns, while multivariate logistic regression analysis and restricted cubic spline regression models were used to analyse the association between dietary patterns and both pre-diabetes and T2DM.

Rural area of Henan province, China.

A total of 38 779 adults aged 18–79 years were recruited from the Henan rural cohort study as the subjects.

The prevalence of pre-diabetes and T2DM in rural Henan was 6·8 % and 9·4 %, respectively. A total of three dietary patterns were assessed in the present study. Dietary pattern I with a high intake of red meat and white meat; dietary pattern II with a high intake of grains, nuts, milk and eggs and dietary pattern III with a high intake of vegetables, staple food and fruits. The highest quintile (Q5) of pattern III could reduce 32·7 % risk of pre-diabetes. The Q5 of pattern II showed a 15·5 % decreased risk of T2DM, in a U-shaped dose–response manner; meanwhile, the Q5 of pattern III was significantly associated with reduced risks of T2DM (OR: 0·582, 95 % CI (0·497, 0·682)).

Pattern III is beneficial for reducing risk of pre-diabetes or T2DM. Though a higher consumption of ‘grains-nuts-egg’ may associate with a reduced risk of T2DM, excessive intakes should be avoided. This study may provide a reference for the prevention of diabetes on dietary precautions.

The present study aimed to evaluate the validity and reproducibility of a thirteen-item FFQ regarding identification of dietary conditions in a rural population in China.

A reproducibility study repeated the first FFQ (FFQ1) approximately 4 weeks later (FFQ2). A validity study evaluated the mean of three consecutive 24 h diet recalls as the reference measure.

Cross-sectional study.

Residents of a rural area in Henan Province, which is located in the central region of China.

A total of 295 individuals participated in the reproducibility study. In addition, 123 people agreed to participate in the validity study. Spearman’s correlation coefficients between the two FFQ ranged from 0·06 (vegetables) to 0·58 (eggs). Spearman’s correlation coefficients between the two methods of collection ranged from 0·01 for cereal to 0·49 for staple foods. The mean of the intraclass correlation coefficients of the two FFQ (FFQ1 v. FFQ2) was 0·19. Bland–Altman analysis indicated good agreement for most food groups across the range of intake for the two studies.

The study demonstrated that our FFQ design could be used as a representative tool to conduct a dietary evaluation of a rural population.