Background: Chronic kidney disease (CKD) is a complex syndrome without a definitive treatment. For these patients, insulin resistance (IR) is associated with worse renal and patient outcomes. Until now, no predictive model using machine learning (ML) has been reported on IR in CKD patients. Methods: The CKD population studied was based on results from the National Health and Nutrition Examination Survey (NHANES) of the USA from 1999 to 2012. The homeostasis model assessment of IR (HOMA-IR) was used to assess insulin resistance. We began the model building process via the ML algorithm (random forest (RF), eXtreme Gradient Boosting (XGboost), logistic regression algorithms, and deep neural learning (DNN)). We compared different receiver operating characteristic (ROC) curves from different algorithms. Finally, we used SHAP values (SHapley Additive exPlanations) to explain how the different ML models worked. Results: In this study population, 71,916 participants were enrolled. Finally, we analyzed 1,229 of these participants. Their data were segregated into the IR group (HOMA IR > 3, n = 572) or non-IR group (HOMR IR ≤ 3, n = 657). In the validation group, RF had a higher accuracy (0.77), specificity (0.81), PPV (0.77), and NPV (0.77). In the test group, XGboost had a higher AUC of ROC (0.78). In addition, XGBoost also had a higher accuracy (0.7) and NPV (0.71). RF had a higher accuracy (0.7), specificity (0.78), and PPV (0.7). In the RF algorithm, the body mass index had a much larger impact on IR (0.1654), followed by triglyceride (0.0117), the daily calorie intake (0.0602), blood HDL value (0.0587), and age (0.0446). As for the SHAP value, in the RF algorithm, almost all features were well separated to show a positive or negative association with IR. Conclusion: This was the first study using ML to predict IR in patients with CKD. Our results showed that the RF algorithm had the best AUC of ROC and the best SHAP value differentiation. This was also the first study that included both macronutrients and micronutrients. We concluded that ML algorithms, particularly RF, can help determine risk factors and predict IR in patients with CKD.

Consuming a Mediterranean-style diet (MED) is helpful for primary prevention of atherosclerotic cardiovascular disease (ASCVD). However, few studies have compared mortality in ASCVD subjects with different degrees of adherence to the MED diet or have evaluated the contributions of individual diet components. We analyzed National Health and Nutrition Examination Survey (NHANES) participants with a history of coronary heart disease (CHD) or stroke (N = 2052) in a period from 1999 to 2010. Their individual vital status was linked to the National Death Index till the end of 2011. The level of adherence to the MED diet was quantified using a 9-point evaluation score (aMED score). Cox regression models were used to compare the different levels of adherence to the MED diet, and contributions of individual components of the MED diet on cardiovascular, cancer, and all-cause mortality. Among the 2052 subjects with CHD or stroke, 29.0% (596 of 2052) died after a median follow-up of 5.6 years. In Cox regression analysis, higher absolute aMED score (HR 0.798, p = 0.0079) or above median aMED score (score 4–9) (HR 0.646, p = 0.0013) was negatively associated with all-cause mortality. Among various components of the MED diet, intake of more whole grains or nuts was significantly associated with a lower all-cause mortality. In contrast, a higher aMED score was not associated with less cardiovascular mortality. In a secondary analysis that excluded deaths within 2 years of the NHANES study entry, the above median aMED score (score 4–9) was negatively associated with both all-cause and cardiovascular mortality. In conclusion, subjects with a history of CHD or stroke adhering better to the MED diet in the NHANES study had a lower all-cause mortality during follow-ups. Consuming more whole grains or nuts had a lower all-cause mortality. The protective effect of adherence to the MED diet on decreasing cardiovascular mortality was seen only after excluding those who died within first two years of the NHANES study entry.

We investigated the associations of adherence to the Mediterranean diet with all-cause and cardiovascular mortality in patients with heart failure. We analyzed the National Health and Nutrition Examination Survey (NHANES) participants from 1999 to 2010, with their vital status confirmed through to the end of 2011. The alternate Mediterranean Diet Index (aMED) was used to assess study participants’ adherence to the Mediterranean diet according to information on dietary questionnaires. We conducted weighted Cox proportional hazards regression models to determine the associations of adherence to the Mediterranean diet (aMED ≥ median vs. <median) with all-cause and cardiovascular mortality in participants with a history of heart failure. A total of 832 participants were analyzed, and the median aMED was 3. After a median follow-up of 4.7 years, 319 participants had died. aMED ≥ 3 (vs. <3) was not associated with a lower risk of all-cause (adjusted HR 0.797, 95% CI 0.599–1.059, p = 0.116) and cardiovascular (adjusted HR 0.911, 95% CI 0.539–1.538, p = 0.724) mortality. The findings were consistent across several subgroup populations. Among the components of aMED, a lower intake of red/processed meat was associated with a higher risk of mortality (adjusted HR 1.406, 95% CI 1.011–1.955, p = 0.043). We concluded that adherence to the Mediterranean diet was not associated with a lower risk of all-cause and cardiovascular mortality in participants with a history of heart failure. The higher risk of mortality associated with a lower intake of red/processed meat deserves further investigation.

Aim: This study aimed to compare mortality risks across uric acid (UA) levels between non-diabetes adults and participants with diabetes and to investigate the association between hyperuricemia and mortality risks in low-risk adults. Methods: We analyzed data from adults aged >18 years without coronary heart disease and chronic kidney disease (n = 29,226) from the National Health and Nutrition Examination Survey (1999–2010) and the associated mortality data (up to December 2011). We used the Cox proportional hazards models to examine the risk of all-cause and cause-specific (cardiovascular disease (CVD) and cancer) mortality at different UA levels between adults with and without diabetes. Results: Over a median follow-up of 6.6 years, 2069 participants died (495 from CVD and 520 from cancers). In non-diabetes adults at UA ≥ 5 mg/dL, all-cause and CVD mortality risks increased across higher UA levels (p-for-trend = 0.037 and 0.058, respectively). The lowest all-cause mortality risk in participants with diabetes was at the UA level of 5–7 mg/dL. We set the non-diabetes participants with UA levels of <7 mg/dL as a reference group. Without considering the effect of glycemic control, the all-cause mortality risk in non-diabetes participants with UA levels of ≥7 mg/dL was equivalent to risk among diabetes adults with UA levels of <7 mg/dL (hazard ratio = 1.44 vs. 1.57, p = 0.49). A similar result was shown in CVD mortality risk (hazard ratio = 1.80 vs. 2.06, p = 0.56). Conclusion: Hyperuricemia may be an indicator to manage multifaceted cardiovascular risk factors in low-risk adults without diabetes, but further studies and replication are warranted.