Development, validation and visualization of a web-based nomogram for predicting risk of new-onset diabetes after … – Nature.com

This study was conducted using medical record data from four centers in three provinces and cities of China and has thus pioneered the construction of a clinical prediction model for NODAP and implemented risk stratification. This study represents the first attempt to establish a nomogram and an online web-based risk calculator for predicting the risk of NODAP. At a median follow-up of 4.6 years (IQR, 2.38.6 years), the incidence of NODAP was established at 46.52%. The risk of NODAP can be appropriately evaluated through FPG level, LDL-C level, hypertension, family history of diabetes and use of diuretics. Our study is the first to report the long-term NODAP incidence rate in Chinese patients with ACS post-PCI, suggesting a high risk of new-onset diabetes that necessitates widespread attention. The prediction model we developed underwent internal and external validation with results, indicating good predictability.

Our predictive model revealed that the long-term incidence of NODAP exceeds the reported incidence of type 2 diabetes in the Chinese population15. Having a history of ACS might serve as an independent risk factor for new-onset diabetes, aligning with results drawn from other cohort study16. Nearly half of patients who undergo PCI must manage diabetes-related medical burden, thus underscoring the importance of appropriate and timely prediction of NODAP. Currently, there are more than 30 prediction models for the incidence of diabetes, each based on different populations17, with some models predicting the risk of new-onset diabetes in patients with coronary heart disease8. However, only one previous study has focused on the risk factors and incidence of NODAP in the Asian population6, with a lack of NODAP prediction models. Our study used rigorous methodology to tackle challenges such as missing data, model building, and internal and external validation, providing a clearer understanding of the groups at high-risk of NODAP.

Glycolipid metabolism levels were identified as significant risk factors for predicting the risk of NODAP, a conclusion that aligns with the findings of several prior studies that developed models to predict new diabetes onset8,18,19. Chun et al. determined through univariate and multivariate regression analyses that the risk factors for NODAP include FPG level100mg/dL, TG level150mg/dL, high body mass index (BMI), and high-intensity statins and that FPG is more influential than other factors6. FPG levels are significantly associated with the risk of diabetes20, and each 1 mg/dL increase in FPG leads to a 9% increase in the risk of diabetes, not only in patients with impaired fasting glucose, but also even in patients with normal FPG levels (9099mg/dL)21. 2 h postprandial glucose (2hPG) is also an indicator for detecting blood glucose levels, but studies have shown that the inclusion of 2hPG and insulin resistance in the prediction model did not significantly improve the predictive accuracy of the model22,23, and therefore NODAP risk prediction by incorporating a convenient and economical FPG is reliable and practical.

There is a consensus that higher LDL-C levels lead to an increased risk of cardiovascular events24,25, and LDL-C levels are the primary therapeutic target for lipid-lowering therapy in patients with coronary artery disease. Meanwhile, LDL-C has also been found to be associated with an increased risk of diabetes26,27, and the mechanism may be related to abnormal cholesterol metabolism affecting pancreatic -cell membrane function and pancreatic cholesterol deposition, which leads to pancreatic -cell dysfunction affecting glucose metabolism28,29,30. The prediction model in this study showed that LDL-C was also a very strong predictor of NODAP, and LDL-C levels were also significantly associated with the onset of NODAP, so controlling LDL-C levels in patients has the additional benefit of preventing and controlling the development of NODAP in addition to the reduced cardiovascular benefit31, but the specific mechanism of the effect of LDL-C on NODAP has yet to be investigated.

Family history of diabetes is a well-known risk factor for type 2 diabetes32, and the same results were obtained in this study for NODAP. Family history increases the risk of diabetes mellitus, which is thought to be mediated by a combination of genetic, environmental, and lifestyle pathways. Genetic susceptibility to diabetes mellitus has been demonstrated by several genome-wide association studies (GWAS), which have linked susceptibility location to pancreatic -cell dysfunction, insulin resistance, and other factors33.

In addition, this study showed that NODAP risk was strongly associated with a history of hypertension rather than with transient values of BP. Hypertension and diabetes mellitus share multiple metabolic syndrome phenotypes including higher body mass index. abdominal obesity, hyperinsulinemia and hypertriglyceridemia34,35, and pathological processes such as dysregulation of reninangiotensinaldosterone system (RAAS), insulin resistance, and inflammation in hypertension can contribute to diabetes mellitus36,37. Therefore, our findings suggest that closer attention should be paid to glycemic indicators in patients undergoing PCI with hypertension.

Although a history of hypertension and a family history of diabetes are non-modifiable factors, evidence from several randomized controlled trials has demonstrated that the onset of diabetes can also be effectively delayed by improving lifestyle38,39. Therefore, our findings suggest that closer attention should be paid to the glycemic indexes of patients undergoing PCI with a family history of hypertension and diabetes, and consideration should be given to reducing the risk of NODAP through stricter lifestyle control.

Patients undergoing PCI are more likely to have hypertension and poor cardiac function and have more complex medications. Therefore, in our study we also considered the influence of prescribed medications, including antihypertensive drugs, psychotropic drugs, diuretics, and various types and intensities of statins on NODAP. We ultimately identified significant correlations between diuretics and NODAP, aligning with the outcomes of several clinical studies40,41. The mechanism by which diuretics affect glucose metabolism is mostly thought to be an indirect effect on insulin secretion due to diuretic-induced hypokalemia42, but its benefits in terms of reducing the occurrence of cardiovascular events are much greater43, and therefore it is still possible to use this drug for treatment with a combination of potassium-preserving diuretics or increased potassium supplementation and thus improvement of glucose metabolism, although the exact mechanism has not yet been fully elucidated44. This suggests that in patients at high risk of NODAP who require diuretic therapy, combined potassium-preserving measures may be considered to reduce the risk of morbidity. Previously, diuretics have also been found to have a dose-related effect on diabetes, and small doses of diuretics may not increase insulin resistance or insulin release45,46. We have not conducted further studies on the dose to be explored in the future. While previous studies on patients post-PCI have indicated that statin treatment increases the risk of NODAP by 27%11, and high-intensity statin treatment increases the risk of NODAP by 48%6, our study did not establish any association between statins or high-intensity statins and an increased risk of NODAP. Considering that the cardiovascular benefits of statin drugs considerably outweigh the adverse effects of new-onset diabetes47,48, statin drugs remain the primary choice for patients post-PCI.

In this study, we revealed that NODAP is a consequence of the combined effects of factors such as genetics, metabolism, and medication. We classified the risk of NODAP into low (11.339.7%), moderate (39.852.8%), and high (52.990.0%) risk. This stratification allows for improved formulation of post-PCI lifestyle and preventive antidiabetic regimens. Specifically, in the clinic, we can provide active NODAP preventive treatment to patients in advance based on their risk prediction results combined with diabetes prevention and treatment guidelines49. For example, if a patient is screened as intermediate risk, it is recommended that he/she should first control modifiable NODAP risk factors, including improving lifestyle, controlling glucose and lipid levels, etc. If he/she is screened as high risk, it is recommended that he/she should carry out intensive lifestyle interventions, including dietary control, exercise and avoiding the use of diuretics as much as possible. This study provides a precise risk calculator for patients with NODAP as well as prompts healthcare workers, especially those in the cardiovascular field, to pay attention to new-onset diabetes and underscores the urgent need for proactive NODAP prevention.

Despite its strengths, there are several limitations to this study: first, it relies on data from hospitals across three provinces and cities in China, with data from the chosen regions being both limited and unevenly spread, thus there may be selection bias. The long follow-up duration of the study may introduce potential attrition bias, and the precise onset date of diabetes could not be determined. Second, this research is a database-oriented retrospective study; therefore, there may be reporting inaccuracies or missing variables, such as height, weight and waist circumference reflecting obesity and type of ACS, number of diseased vessels, and number of stents implanted reflecting severity of the condition, as well as the inevitable recollection bias of the case data. And it was also not possible to obtain variable characteristics of lifestyle, dietary patterns, and frequency of exercise that may also affect NODAP. The lack of comprehensiveness of the variables covered also resulted in an inability to effectively compare with previous diabetes prediction models. Third, there are many differences between the training and validation cohort population characteristics, which may be related to the fact that the two parts of the cohort came from different provinces, different hospitals, and different severity of the disease, etc. More research centers, larger sample sizes, and more research variables need to be included to further externally validate the model efficacy. Additionally, the laboratory indicators used for diagnosing new-onset diabetes were dependent on glycated, fasting, and random blood glucose levels, which may not fully represent the long-term blood glucose control. Future research using oral glucose tolerance test results could potentially offer higher diagnostic sensitivity and precision. Lastly, we have only established a NODAP prediction and risk stratification system, and have not studied the prognosis of patients at different risks with different treatment modalities, nor have we been able to determine a causal relationship. Further studies are needed to investigate the effects of different management strategies on the long-term prognosis of patients undergoing PCI and the specific mechanisms linking these factors to NODAP.

Read more here:
Development, validation and visualization of a web-based nomogram for predicting risk of new-onset diabetes after ... - Nature.com