Diet and Disease: Long-Term Nutritional Impact Model

Establishing causality in nutrition research is a significant challenge. According to editorial research analyzed by Groundwork, While randomised controlled trials (RCTs) provide reliable evidence, long-term dietary intervention studies with disease endpoints are often impractical. Short-term RCTs can instead identify intermediate traits that may lie on the causal pathway between diet and disease.

The Limitations of Mendelian Randomisation

Mendelian randomisation (MR) is an epidemiological approach that uses genetic variants as proxies for modifiable exposures to estimate the effects of lifelong differences in exposure on disease risk. However, the utility of MR is limited for complex dietary patterns because genetic variants typically reflect biological mechanisms rather than specific diets.

A Two-Step Framework for Integrating RCTs and MR

We propose a two-step framework integrating dietary RCTs with MR to infer potential lifetime effects of dietary interventions. The first step involves using RCT data to identify molecular traits altered by an intervention. The second step involves using MR to evaluate whether these traits are associated with long-term disease risk.

Step 1: Identifying Molecular Traits Altered by an Intervention

In the first step, RCT data are used to identify molecular traits altered by an intervention. This can be achieved by measuring circulating proteins, gene expression, or other biomarkers before and after the intervention. The Diabetes Remission Clinical Trial (DiRECT) is a prime example of this approach. DiRECT measured circulating proteins and diabetes remission in response to a dietary intervention.

Step 2: Evaluating Associations with Long-Term Disease Risk

In the second step, MR is used to evaluate whether the molecular traits identified in the first step are associated with long-term disease risk. This involves using genetic variants associated with the molecular traits to estimate the effects of lifelong differences in exposure on disease risk.

Demonstrating the Framework Using DiRECT

We demonstrate this framework using the Diabetes Remission Clinical Trial (DiRECT). DiRECT measured circulating proteins and diabetes remission in response to a dietary intervention. Using protein data alone, 216 of 4,601 proteins changed following the intervention. MR was then used to evaluate whether these proteins were associated with long-term diabetes risk. The results showed that 10 proteins were associated with diabetes risk using MR.

Comparing MR Estimates with Observed Protein-Remission Associations

We then compared these MR estimates with observed protein-remission associations from DiRECT. The broad agreement between the two (r{approx}-0.645, R2=0.416) supports this framework as a useful approach for estimating long-term effects of dietary interventions.

The Implications of This Framework

This framework has significant implications for the field of nutrition research. By integrating RCTs and MR, researchers can estimate the long-term effects of dietary interventions on disease risk. This can help to inform public health recommendations and guide the development of effective dietary interventions.

Actionable Takeaways

1. Integrate RCTs and MR: Use RCT data to identify molecular traits altered by an intervention, and then use MR to evaluate whether these traits are associated with long-term disease risk.
2. Measure Biomarkers: Measure circulating proteins, gene expression, or other biomarkers before and after the intervention to identify molecular traits altered by the intervention.
3. Use Genetic Variants: Use genetic variants associated with the molecular traits to estimate the effects of lifelong differences in exposure on disease risk.

The proposed two-step framework integrating RCTs and MR has the potential to transform the field of nutrition research. By estimating the long-term effects of dietary interventions on disease risk, researchers can inform public health recommendations and guide the development of effective dietary interventions. However, further research is needed to validate this framework and explore its applications in different disease areas.