Using food network analysis to understand meal patterns in pregnant women with high and low diet quality

Association between dietary intake networks identified through a Gaussian graphical model and the risk of cancer: a prospective cohort study

PURPOSE

In this study, we aimed to investigate the association between dietary communities identified by a Gaussian graphical model (GGM) and cancer risk.

METHODS

We performed GGM to identify the dietary communities in a Korean population. GGM-derived communities were then scored and investigated for their association with cancer incidence in the entire population as well as in the 1:1 age- and sex-matched subgroup using a Cox proportional hazards model. In the sensitivity analysis, GGM-derived communities were compared to dietary patterns (DPs) that were identified by principal component analysis (PCA) and reduced rank regression (RRR).

RESULTS

During a median time to follow-up of 6.6 years, 397 cancer cases were newly diagnosed. The GGM identified 17 and 16 dietary communities for the total and matched populations, respectively. For each one-unit increase in the standard deviation of the community-specific score of the community that was composed of dairy products and bread, there was a reduced risk of cancer according to the fully adjusted model (HR: 0.80, 95% CI: 0.66-0.96). In the matched population, the third tertile of the community-specific score of the community composed of poultry, seafood, bread, cakes and sweets, and meat by-products showed a significantly reduced risk of cancer compared to that of the lowest tertile in the fully adjusted model (HR: 0.66, 95% CI: 0.50-0.86, p-trend = 0.002).

CONCLUSION

We found that the GGM-identified community composed of dairy products and bread showed a reduced risk of cancer. Further population-based prospective studies should be conducted to examine possible associations of dietary intake and specific cancer types.