The objective of this study was to determine the short-term and long-term effects of a nutrition intervention in using 37years of follow-up data. The Linxian Dysplasia Population Nutrition Intervention Trial was a randomized, double-blind, placebo-controlled trial with 7years of intervention and 30years of follow-up. The Cox proportional hazard model was used for analyses. Subgroup analyses were conducted in age and sex subgroups, and the 30years of follow-up were divided into two 15-year early and late periods. The results at 37years did not indicate any effects on mortality from cancers or other diseases. In the first 15years, the intervention decreased the overall risk of gastric cancer deaths in all participants (hazard ratio [HR], 0.76; 95% confidence interval [CI], 0.58-1.00) and in the subgroup participants younger than 55 years (HR, 0.64; 95% CI, 0.43-0.96). In addition, in the group younger than 55 years (HR, 0.58; 95% CI, 0.35-0.96), the intervention decreased the risk of death from other diseases; and, in the group aged 55 years and older (HR, 0.75; 95% CI, 0.58-0.98), the intervention reduced the risk of death from heart disease. There were no significant results in the later 15years, which indicated the disappearance of the intervention effect. Comparing demographic characteristics between those who died during the two periods, the participants who died later included more women, had a higher education level, had a lower smoking rate, were younger, and also more had a mild degree of esophageal dysplasia, representing a better lifestyle and health condition. Long-term follow-up indicated no effect of nutrition on deaths in a population with esophageal squamous dysplasia, further supporting the significance of continuous nutritional intervention for cancer protection. The pattern of protective effect of a nutrition intervention on gastric cancer in patients with esophageal squamous dysplasia was similar to that in the general population. Participants who died in the later period had more protective factors than those who died in the earlier period, contributing to the obvious effect of the intervention in early stage disease.
Read full abstract