A recent study published in Engineering has shed new light on how the combinations of foods we eat over the long term, rather than just individual food items, can significantly influence cardiometabolic health. The research, led by a team of scientists from various institutions, including the First Affiliated Hospital of Nanjing Medical University and the Guangdong Provincial Institute of Public Health, analyzed data from two large cohorts: the National Health and Nutrition Examination Survey (NHANES) in the United States and the Guangdong Gut Microbiome Project (GGMP) in China. The findings suggest that the balance and imbalance of food intake, captured through long-term food pairing patterns, are independently associated with cardiometabolic traits and can modulate gut microbial functionalities.

The study introduced the concept of long-term food pairing patterns, which include additive food pairing patterns (AFPs) and subtractive food pairing patterns (SFPs). AFPs assume synergistic effects between two foods, while SFPs assume antagonistic effects. The researchers used normalized monthly consumption frequencies of individual foods to calculate these patterns. They identified 1759 and 306 cardiometabolic-related long-term food pairing patterns from the NHANES and GGMP cohorts, respectively, with a false discovery rate (FDR) of less than 0.05. Notably, by comparing the cardiometabolic association results with single foods and long-term food pairing patterns, around 80.8%/82.5% of these pairing foods were not individually associated with cardiometabolic traits, indicating that the combined effects of food pairings play a crucial role in health outcomes.

The study found that long-term food pairing patterns were weakly correlated with single food intake frequencies and various dietary indices, such as the Healthy Eating Index 2020 (HEI2020) and the Dietary Approaches to Stop Hypertension (DASH) index. This suggests that food pairing patterns provide unique insights beyond conventional dietary indicators. The researchers also observed that the associations between food pairing patterns and cardiometabolic traits were consistent across Eastern and Western populations at the hyper food classification group level, despite significant differences in individual food intake patterns.
The study further explored the mediating role of the gut microbiome in the relationship between long-term food pairing patterns and cardiometabolic traits. Mediation analysis revealed that 72.7% of long-term food pairing patterns affected cardiometabolic traits through 31 microbial genera, with Clostridium sensu stricto 1 playing a predominant role. The researchers found that microbes primarily mediated the impact of long-term food pairing patterns on cardiometabolic traits through their metabolic pathways, such as pyruvate fermentation to propanoate and ergothioneine biosynthesis pathways.

The findings highlight the importance of considering the balance and imbalance of food intake in dietary strategies aimed at improving cardiometabolic health. The study suggests that personalized dietary recommendations could benefit from incorporating long-term food pairing patterns, which may offer a more comprehensive understanding of how diet influences health compared to focusing solely on individual foods or dietary indices.

The study acknowledges several limitations, including the observational nature of the research and the potential for recall bias in food frequency questionnaires. Future research could explore the causal relationships between food pairing patterns, gut microbiome, and cardiometabolic health through longitudinal studies or controlled trials. Nonetheless, the study provides valuable insights into the potential role of long-term food pairing patterns in human health and opens new avenues for developing precision nutritional strategies.

The paper “Food pairing pattern independently associated with cardiometabolic traits beyond dietary indices across eastern and western populations,” is authored by Qiufeng Deng, Yuhao Sun, Mingxia Gu, Jingxiang Fu, Fengzhe Xu, Yuwen Jiao, Yan Zhou, Beining Ma, Lu Liu, Xiuchao Wang, Quanbin Dong, Tingting Wu, Huayiyang Zou, Jing Shi, Yifeng Wang, Yanhui Sheng, Liming Tang, Wei Sun, An Li, Shixian Hu, Jusheng Zheng, Yan He, Hongwei Zhou, Wei Wu, Xiangqing Kong, Lianmin Chen. Full text of the open access paper: https://doi.org/10.1016/j.eng.2025.02.015. For more information about Engineering, visit the website at https://www.sciencedirect.com/journal/engineering.