Subjects and study design
A total of 533 volunteers, without sex restrictions, were recruited from the Third People’s Hospital of Baiyin City and the Jingyuan County Hospital of Traditional Chinese Medicine nearby community. Of these, 179 participants were selected for the study, including 12 hospital administrative staff, 102 healthcare professionals, 20 unemployed or retired persons, 13 migrant workers, and 32 individuals from other occupations, and all signed written informed consent forms. Baseline surveys (F0) were conducted at both hospitals in June 2021 and September 2021, with follow-up visits (F2) in August 2021 and November 2021. The study adhered to medical ethics standards and was reviewed and approved by the Ethics Committee of the School of Public Health of Lanzhou University (Approval No. IRB19122001). Inclusion criteria were as follows: (1) age 20–70 years, voluntary participation. Exclusion criteria included: (1) refusal to sign the informed consent form; (2) pregnant or lactating women, allergic; (3) patients with serious diseases of the heart, liver, kidney, haematopoietic system and psychiatric patients; (4) use of drugs that severely impair gastrointestinal function within 3 months; (5) patients using e.g. NSAIDs or anticoagulants; (6) patients with severe ulcers of the peptic system; (7) self-reported patients with atrophic gastritis; (8) patients who are unable to cooperate with the completion of investigations or tests.
Laboratory tests
After fasting for at least 8 h, 5 ml of venous blood was collected. The levels of inflammatory cytokines IFN-γ, TNF-α, CRP, IL-17A, IL-1β, IL-8 and IL-18 were measured by ELISA kit (Human ELISA Kit; Elabscience, Wuhan, Hubei, CHN). The coefficient of variation of inflammatory cytokines above was 11.25, 9.37%, 12.89, 5.34, 10.53, 9.34 and 9.32% respectively. Plasma levels of PG I and PG II were tested with a time-resolved fluorescence immunoassay (Pepsinogen Quantitative Determination Kit; Jiangyuan industrial Technology, Wuxi, Jiangsu, CHN), and the ratio of PG I to PG II (PGR) was calculated. The coefficient of variation of PG I and PG II was 10.90 and 9.24%, respectively. The change in inflammatory cytokines is defined as the difference between the inflammatory cytokines levels at follow-up and the baseline levels (F2-F0).
Changes in lifestyle components and combined adverse lifestyle scores
Data on lifestyle factors were collected using a self-administered questionnaire and a face-to-face survey conducted by uniformly trained investigators. The combined adverse lifestyle score was calculated by summing the lifestyle scores across six dimensions7,19 (Table 1), with a maximum possible score of 8. A higher score indicated a more adverse lifestyle.
The lifestyle factor change score was determined by subtracting the baseline lifestyle score from the follow-up lifestyle score. A positive value indicated a worsening lifestyle, while a negative value indicated an improvement (e.g. + 1 for a change from normal to abnormal BMI and – 1 for a change from abnormal to normal). The change in the combined adverse lifestyle score was calculated similarly, ranging from − 8 to + 8. A higher positive score indicated a deterioration in lifestyle. For further details, please refer to the eMethods in the Supplements.
Covariates
Our analysis included several key covariates to ensure a comprehensive understanding of the factors influencing inflammation: sex (male/female), age, education level (junior high school and below, high school and polytechnic school, junior college, bachelor or above), annual household income (less than or equal to 20,000 yuan, over 20,000 to 60,000 yuan, over 60,000 to 100,000 yuan, over 100,000 yuan), monthly alcohol intake, prevalence of 20 common diseases (including gastric ulcer, acute gastritis, chronic gastritis, duodenal ulcer, cholecystitis, hepatitis, fatty or alcoholic liver, enteritis, pancreatitis, reflux oesophagitis, kidney disease, prostate disease, rhinitis, pharyngitis, asthma, rash or urticaria, lupus erythematosus, cardiovascular disease, hyperthyroidism, gout), use of Bacillus licheniformis and other probiotics (e.g. bifidobacterium, lactobacillus capsules) in the last month, use of medications damaging to the stomach (e.g. dexamethasone, erythromycin, aspirin, phenylbutazone) in the last month, use of cold or painkillers in the last month, use of vitamin and plant antioxidant-based nutritional supplements, use of anti-inflammatory drugs (e.g. omeprazole, amoxicillin, clarithromycin, levofloxacin, metronidazole, tetracycline, furazolidone, citric acid, bismuth gum, ranitidine) in the last month, and changes in blood pressure, blood glucose, HDL-C, and LDL-C in the last month at the time of follow-up.
Statistical analysis
The sample size for this study was calculated using CRP levels as the primary indicator, based on previous research showing CRP levels of 1.2 ± 2.7 pg/mL and 2.5 ± 2.5 pg/mL in different lifestyle groups7. With a test power of 95%, a two-sided α of 0.05, and accounting for a 20% follow-up loss rate, the required sample size was determined to be 174 participants.
Continuous variables were described using means and standard deviations or medians and interquartile ranges, while categorical variables were described using frequencies and composition ratios. Histograms and Q-Q plots were used to assess the approximate normality of all measures. CRP indicators with a right-skewed distribution were LN-transformed to improve normality. Levels of inflammatory cytokines in different lifestyle groups at baseline were compared using independent samples t-tests, ANOVA, and LSD-t tests (see Table S2 in the Supplements).
The relationship between changes in lifestyle and changes in seven inflammatory cytokines over time was analyzed using multiple linear regression. Changes in the inflammatory cytokines served as the dependent variables, while changes in lifestyle factors and combined adverse lifestyle changes were the independent variables. Model 1 adjusted for sex and age, and also included baseline levels of each variable to account for their impact on observed changes. Model 2 further adjusted for all covariates. When analyzing changes in single lifestyle factors, adjustments were made for changes in other lifestyle factors. Since smoking status did not change at follow-up, the total number of previous smoking packs was included as a covariate. Additionally, a mixed-effects linear model in Model 2 evaluated the relationship between independent variables and cytokines progression, with time included as a covariate. Random intercepts for study participants were included to reflect differences in cytokine levels associated with varying lifestyle levels and changes. Continuous variables such as metabolic equivalent (MET) of daily activity, BMI, sleep time, alcohol intake, and smoking (packs) were used to fit restricted cubic spline (RCS) curves.
Several sensitivity analyses were conducted. First, weighted lifestyle scores were calculated based on the standardized β effect values for each lifestyle factor in multiple linear regression Model 2, using the formula: i = [βi/(∑βi)] × 5; (2) Each lifestyle component of the combined adverse lifestyle score was excluded in turn from the analysis. (3) Subjects with potential atrophic gastritis lesions (pepsinogen PG I ≤ 60 ng/mL or PGR ≤ 6, based on the normal reference range provided in the kit) and those using anti-inflammatory medications were excluded separately. Multiple linear regression model 2 also examined interactions between changes in different lifestyle factors. All statistical tests were two-tailed, with a significance level set at P ≤ 0.05, using IBM SPSS version 25.0 (IBM Corp., Armonk, NY, USA).
Ethics approval
This study was conducted in accordance with the Declaration of Helsinki, and approved by the Ethics Committee of the School of Public Health of Lanzhou University (Approval No. IRB19122001).
Consent to participate
Informed consent was obtained from all individual participants included in the study. Details that might disclose the identity of the subjects under study have been omitted.