Diabetes mellitus is a significant global health concern, placing a substantial economic and health burden. The International Diabetes Federation estimates a substantial number of adults with diabetes, with a concerningly high proportion remaining undiagnosed. Early diagnosis and intervention are crucial for preventing complications. Dietary patterns play a vital role in diabetes risk and management. Studies have shown associations between dietary factors (e.g., Mediterranean diet, processed red meat, sugar-sweetened beverages) and diabetes risk. This study focuses on the trends in dietary patterns among US adults with diagnosed and undiagnosed diabetes over the past decade, using the Healthy Eating Index (HEI) and Dietary Inflammatory Index (DII) to assess dietary quality and inflammatory potential. The researchers aim to investigate the association between these dietary patterns and long-term mortality, contributing to a better understanding of lifestyle interventions for diabetes management, particularly for the undiagnosed population.

Existing literature highlights the strong link between diet and diabetes incidence. The Mediterranean diet, for instance, has been associated with a lower risk of diabetes, while diets high in processed red meat, sugar-sweetened beverages, and pro-inflammatory components have shown increased risk. The HEI, a measure of dietary quality based on the Dietary Guidelines for Americans, has been linked to lower risks of cardiovascular disease and diabetes in previous studies. However, its association with diabetes prognosis remains less explored. The DII, which assesses the dietary inflammatory potential, has also been connected to increased diabetes risk. Despite the established link between diet and diabetes incidence, there's a lack of data on the relationship between dietary patterns and diabetes prognosis, especially in the undiagnosed population. This research addresses this gap by examining the trends in dietary patterns and their association with long-term mortality using HEI and DII.

This retrospective cohort study utilized data from the National Health and Nutrition Examination Survey (NHANES) 2007-2018. Participants aged 20 years or older, without pregnancy, and with available 24-hour dietary interview data were included. Participants were categorized into three groups: no diabetes, undiagnosed diabetes, and diagnosed diabetes. Diabetes diagnosis was determined based on self-reported physician diagnosis, fasting glucose levels, oral glucose tolerance test results, HbA1c levels, and/or use of diabetes medication or insulin. The HEI-2015, comprising 13 food parameters, and the DII, incorporating various nutrients and food components, were used to evaluate dietary patterns. The HEI score was calculated using the Food Stamps participant HEI scoring approach and a mobile and simple HEI scoring algorithm. The DII score was calculated based on the Dietary Guidelines method reported by Shang et al. (2014). The primary endpoint was all-cause mortality, with secondary endpoints including death from malignant neoplasms, heart disease, chronic lower respiratory diseases, and cerebrovascular diseases. Survival analysis using age-adjusted weighted Cox proportional hazard regression models was conducted to examine the association between HEI/DII scores and mortality, adjusting for age, sex, education, BMI, smoking, hypertension, hyperlipidemia, and alcohol consumption. All analyses incorporated NHANES's complex survey design.

The study found a steadily increasing prevalence of diabetes in the US over the past decade, with a notable increase in undiagnosed diabetes. The overall HEI scores showed a downward trend across all three groups (no diabetes, undiagnosed diabetes, diagnosed diabetes). Participants with undiagnosed diabetes had significantly lower HEI scores compared to those with diagnosed diabetes. Both undiagnosed and diagnosed diabetes groups exhibited higher DII scores than the non-diabetes group. Survival analysis revealed a significant association between higher HEI scores and lower risk of all-cause mortality and death from heart disease. Conversely, higher DII scores were significantly associated with a higher risk of all-cause mortality and death from heart disease. These associations remained significant even after adjusting for potential confounders. The prevalence of undiagnosed diabetes increased from 8.43% in 2007-2008 to 11.65% in 2017-2018.

The study's findings highlight the worsening dietary patterns among US adults, particularly those with diabetes, both diagnosed and undiagnosed. The declining HEI scores suggest a potential decrease in the effectiveness of dietary interventions. The increased prevalence of undiagnosed diabetes emphasizes the critical need for early detection and intervention. The association between higher HEI scores and lower mortality risk underscores the importance of adhering to dietary guidelines for improved health outcomes. Conversely, the association between higher DII scores and higher mortality risk emphasizes the negative impact of pro-inflammatory diets. These findings support the need for improved dietary management strategies, focusing on reducing dietary inflammatory potential and improving overall dietary quality.

This study provides valuable insights into dietary patterns and their association with mortality in US adults with diabetes. The declining HEI scores and rising DII scores highlight the need for intensified dietary interventions and public health initiatives aimed at improving dietary habits. Future research should explore strategies to enhance dietary interventions and address the increasing prevalence of undiagnosed diabetes. Further investigation into the specific dietary components contributing to DII scores and their impact on inflammation and mortality is warranted. The limitations of 24-hour dietary recall data should also be considered in future studies.

The study's reliance on 24-hour dietary recall data may have limitations in accurately capturing long-term dietary patterns. The study did not assess other dietary patterns, such as the Mediterranean diet. The lack of inflammatory marker data prevents a direct assessment of the relationship between DII scores and inflammatory markers. The cross-sectional nature of the NHANES data limits causal inference.