-
Clinical Cardiology Oct 2023The purpose of this study was to develop and validate a machine learning (ML) based prediction model for the risk of heart failure (HF) in patients with prediabetes or... (Randomized Controlled Trial)
Randomized Controlled Trial
Development and validation of a prediction model based on machine learning algorithms for predicting the risk of heart failure in middle-aged and older US people with prediabetes or diabetes.
BACKGROUND
The purpose of this study was to develop and validate a machine learning (ML) based prediction model for the risk of heart failure (HF) in patients with prediabetes or diabetes.
METHODS
We used 3527 subjects aged 40 years and older with a prior diagnosis of prediabetes or diabetes from the National Health and Nutrition Examination Survey (NHANES) from 2007 to 2018. The search for independent risk variables linked to HF was conducted using univariate and multivariable logistic regression analysis. The 3527 subjects were randomly divided into training set and validation set in a 7:3 ratio. Five ML models were built on the training set using five ML algorithms, including random forest (RF), and then validated on the validation set. Receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis and Bootstrap resampling method were used to measure the predictive performance of the five ML models.
RESULTS
Multivariate logistic regression analysis showed that age, poverty-to-income ratio, myocardial infarction condition, coronary heart disease condition, chest pain condition, and glucose-lowering medication use were independent predictors of HF. By comparing the performance of the five ML models, the RF model (AUC = 0.978) was the best prediction model.
CONCLUSIONS
The risk of HF in middle-aged and elderly patients with prediabetes or diabetes can be accurately predicted using ML models. The best prediction performance is presented by RF model, which can assist doctors in making clinical decisions.
Topics: Aged; Middle Aged; Humans; Adult; Prediabetic State; Nutrition Surveys; Diabetes Mellitus; Heart Failure; Algorithms; Machine Learning; Risk Factors
PubMed: 37519220
DOI: 10.1002/clc.24104 -
Wiener Klinische Wochenschrift Jan 2023Metabolic diseases dramatically affect the life of men and women from infancy up to old age in different and manifold ways and are a major challenge for the healthcare...
Metabolic diseases dramatically affect the life of men and women from infancy up to old age in different and manifold ways and are a major challenge for the healthcare system. The treating physicians are confronted with the different needs of women and men in the clinical routine. Gender-specific differences affect pathophysiology, screening, diagnostic and treatment strategies of diseases as well as the development of complications and mortality rates. Impairments in glucose and lipid metabolism, regulation of energy balance and body fat distribution and therefore the associated cardiovascular diseases, are greatly influenced by steroidal and sex hormones. Furthermore, education, income and psychosocial factors play an important role in the development of obesity and diabetes differently in men and women. Males appear to be at greater risk of diabetes at a younger age and at a lower body mass index (BMI) compared to women but women feature a dramatic increase in the risk of diabetes-associated cardiovascular diseases after the menopause. The estimated future years of life lost owing to diabetes is somewhat higher in women than men, with a higher increase in vascular complications in women but a higher increase of cancer deaths in men. In women prediabetes or diabetes are more distinctly associated with a higher number of vascular risk factors, such as inflammatory parameters, unfavourable changes in coagulation and higher blood pressure. Women with prediabetes and diabetes have a much higher relative risk for vascular diseases. Women are more often morbidly obese and less physically active but may have an even greater benefit in health and life expectation from increased physical activity than men. In weight loss studies men often showed a higher weight loss than women; however, diabetes prevention is similarly effective in men and women with prediabetes with a risk reduction of nearly 40%. Nevertheless, a long-term reduction in all cause and cardiovascular mortality was so far only observed in women. Men predominantly feature increased fasting blood glucose levels, women often show impaired glucose tolerance. A history of gestational diabetes or polycystic ovary syndrome (PCOS) as well as increased androgen levels and decreased estrogen levels in women and the presence of erectile dysfunction or decreased testosterone levels in men are important sex-specific risk factors for the development of diabetes. Many studies showed that women with diabetes reach their target values for HbA, blood pressure and low-density lipoprotein (LDL)-cholesterol less often than their male counterparts, although the reasons are unclear. Furthermore, sex differences in the effects, pharmacokinetics and side effects of pharmacological treatment should be taken more into consideration.
Topics: Pregnancy; Female; Male; Humans; Prediabetic State; Obesity, Morbid; Cardiovascular Diseases; Risk Factors; Weight Loss; Diabetes, Gestational
PubMed: 37101049
DOI: 10.1007/s00508-023-02185-5 -
Journal of the American Board of Family... 2020Prediabetes is increasing in prevalence and is associated with risk of developing diabetes, heart disease, stroke, and retinopathy. Clinicians have limited tools to...
BACKGROUND
Prediabetes is increasing in prevalence and is associated with risk of developing diabetes, heart disease, stroke, and retinopathy. Clinicians have limited tools to facilitate prediabetes discussions within primary care visits.
PURPOSE
1) Develop a Patient and Stakeholder Advisory Committee (PASAC) to design, evaluate, and revise a prediabetes shared decision aid, and 2) evaluate the feasibility and experience of implementing the tool within primary care practice.
METHODS
A prediabetes decision aid (double-sided infographic with decision questions) was created by a PASAC that included patients, primary care clinicians, diabetes educators, endocrinologists, and pharmacists. Five clinicians within 3 primary care practices tested the prediabetes tool with 50 adult patients with prediabetes. Patients completed 2 surveys immediately after the office visit and 6 weeks later. Clinicians and PASAC members completed a postintervention survey.
RESULTS
The prediabetes shared decision aid was created through a deliberative process over 3 PASAC meetings. Ninety-six percent of patients felt the tool prepared them to decide on a diabetes prevention plan, and 100% of clinicians would use the tool again and felt the tool did not extend visit length.
DISCUSSION
It was feasible to cocreate a prediabetes shared decision aid within a PASAC and implement the tool within a primary care setting. Patients and clinicians reported a prediabetes discussion, which may mitigate rates of progression to diabetes and associated complications. Future research should evaluate which of the intervention components most effectively promotes discussion of prediabetes within a primary care setting.
Topics: Adult; Decision Support Techniques; Diabetes Mellitus; Humans; Pharmacists; Prediabetic State; Primary Health Care
PubMed: 32179609
DOI: 10.3122/jabfm.2020.02.190070 -
Clinical and Molecular Hepatology Jul 2022Nonalcoholic fatty liver disease (NAFLD) is closely associated with diabetes. The cumulative impact of both diseases synergistically increases risk of adverse events....
BACKGROUND/AIMS
Nonalcoholic fatty liver disease (NAFLD) is closely associated with diabetes. The cumulative impact of both diseases synergistically increases risk of adverse events. However, present population analysis is predominantly conducted with reference to non-NAFLD individuals and has not yet examined the impact of prediabetes. Hence, we sought to conduct a retrospective analysis on the impact of diabetic status in NAFLD patients, referencing non-diabetic NAFLD individuals.
METHODS
Data from the National Health and Nutrition Examination Survey 1999-2018 was used. Hepatic steatosis was defined with United States Fatty Liver Index (US-FLI) and FLI at a cut-off of 30 and 60 respectively, in absence of substantial alcohol use. A multivariate generalized linear model was used for risk ratios of binary outcomes while survival analysis was conducted with Cox regression and Fine Gray model for competing risk.
RESULTS
Of 32,234 patients, 28.92% were identified to have NAFLD. 36.04%, 38.32% and 25.63% were non-diabetic, prediabetic and diabetic respectively. Diabetic NAFLD significantly increased risk of cardiovascular disease (CVD), stroke, chronic kidney disease, all-cause and CVD mortality compared to non-diabetic NAFLD. However, prediabetic NAFLD only significantly increased the risk of CVD and did not result in a higher risk of mortality.
CONCLUSION
Given the increased risk of adverse outcomes, this study highlights the importance of regular diabetes screening in NAFLD and adoption of prompt lifestyle modifications to reduce disease progression. Facing high cardiovascular burden, prediabetic and diabetic NAFLD individuals can benefit from early cardiovascular referrals to reduce risk of CVD events and mortality.
Topics: Cardiovascular Diseases; Diabetes Mellitus; Humans; Non-alcoholic Fatty Liver Disease; Nutrition Surveys; Prediabetic State; Prevalence; Retrospective Studies; Risk Factors
PubMed: 35585687
DOI: 10.3350/cmh.2022.0096 -
Journal of the American Board of Family... 2021Lifestyle modification can significantly reduce the risk of developing diabetes. However, the effect of prediabetes status awareness on perceptions of health and health...
INTRODUCTION
Lifestyle modification can significantly reduce the risk of developing diabetes. However, the effect of prediabetes status awareness on perceptions of health and health behaviors are mixed. We used the 2015 to 2016 National Health and Nutrition Examination Survey (NHANES) data to evaluate the association of prediabetes status awareness with self-reported perception of health and health behaviors.
METHODS
A secondary analysis was performed on the 2015 to 2016 NHANES, which is a biannual, cross-sectional survey designed to be representative of the civilian, noninstitutionalized United States population. Survey participants were interviewed at home and invited to a mobile examination center to undergo examinations and laboratory measurements. Inclusion and exclusion criteria were applied to create prediabetes-aware and prediabetes-unaware groups. The groups were propensity-score matched based on Body Mass Index, A1c, race, and education. Measures of health perception and health behaviors were analyzed, including feeling at risk for developing diabetes, general health status, perception of weight, desire to lose weight, dietary behaviors, and physical activity behaviors.
RESULTS
Participants who were prediabetes aware were more likely to report a perceived risk of diabetes and to consider themselves overweight. Prediabetes awareness was not associated with any difference in dietary or physical activity behaviors.
CONCLUSIONS
Although participants who were aware of their prediabetes status were more likely to report a perceived threat of developing diabetes, they did not report increased engagement in health behaviors. Future research can better understand how these aspects of a health belief model affect engagement in health behaviors for people with prediabetes.
Topics: Cross-Sectional Studies; Diabetes Mellitus; Humans; Nutrition Surveys; Perception; Prediabetic State; United States
PubMed: 33452101
DOI: 10.3122/jabfm.2021.01.200146 -
Metabolic Syndrome and Related Disorders Apr 2021To characterize the associations of clinical risk factors, lifestyle factors, and wellness measures with prediabetes and new type 2 diabetes mellitus (T2DM) diagnosis...
To characterize the associations of clinical risk factors, lifestyle factors, and wellness measures with prediabetes and new type 2 diabetes mellitus (T2DM) diagnosis in Hispanic adults and guide primary prevention. Biobank enrolled 3733 Hispanic adults from Phoenix, AZ, United States, from 2013 to 2018. This analysis included participants with euglycemia, prediabetes, or new T2DM diagnosis (, no prior T2DM diagnosis) at enrollment. Participants completed a baseline questionnaire on cardiometabolic risk factors and wellness measures and provided biometric measurements. The associations of factors and measures with odds (95% confidence interval) of prediabetes and new T2DM diagnosis were analyzed in logistic regression models. Among 3299 participants with euglycemia ( = 1301), prediabetes ( = 1718), and new T2DM diagnosis ( = 280) at enrollment, 72% were women ( = 2376/3299). In adjusted models, most cardiometabolic risk factors were positively associated with prediabetes and new T2DM diagnosis, with stronger associations for new T2DM diagnosis. Obesity (body mass index ≥30 kg/m vs. lower) was associated with higher odds of new T2DM diagnosis (3.14 [2.30-4.28]; < 0.01) than prediabetes versus euglycemia (1.96 [1.66-2.32]; < 0.01) and Interaction ( = 0.01). Similarly, waist circumference, family history of diabetes, and average systolic and diastolic blood pressure were associated with higher odds of new T2DM diagnosis versus euglycemia than prediabetes versus euglycemia. Using stepwise logistic regression modeling, a parsimonious model of age, family history of diabetes, waist circumference, diastolic blood pressure, passive tobacco exposure, and self-rated general health were associated with new T2DM diagnosis versus euglycemia. In Hispanic adults, modifiable cardiometabolic and lifestyle factors were associated with prediabetes and new T2DM diagnosis. Personalized interventions targeting these factors and measures could guide T2DM primary prevention efforts among Hispanic adults.
Topics: Adult; Diabetes Mellitus, Type 2; Female; Hispanic or Latino; Humans; Prediabetic State; Risk Factors; Waist Circumference
PubMed: 33439762
DOI: 10.1089/met.2020.0102 -
Frontiers in Endocrinology 2023Fatty liver, obesity, and dyslipidemia are associated with prediabetes or diabetes risk, and hyperuricemia co-exists. The present study evaluated the role of multiple...
BACKGROUND
Fatty liver, obesity, and dyslipidemia are associated with prediabetes or diabetes risk, and hyperuricemia co-exists. The present study evaluated the role of multiple mediators, namely, fatty liver, body mass index (BMI), and dyslipidemia, in the association between hyperuricemia and diabetes status.
METHODS
Baseline data from the ongoing Fuqing cohort (5,336 participants) were analyzed to investigate the association of hyperuricemia with diabetes status using a multinomial logistic regression model. Furthermore, causal mediation analysis with the weighting-based approach was performed to estimate hyperuricemia's total natural direct effect (tnde), total natural indirect effect (tnie), and total effect (te) on prediabetes and diabetes risk, mediating jointly fatty liver, BMI, and dyslipidemia.
RESULTS
In multinomial analysis without considering mediators' effects, hyperuricemia was associated with a higher risk of prediabetes only (odds ratio: 1.25; 95% CI: 1.09-1.43; < 0.001). When fatty liver, BMI, and dyslipidemia were considered as multiple mediators in the association, hyperuricemia was linked to both prediabetes [tnde: 1.11, 95% CI: 1.04-1.11; tnie: 1.07, 95% CI: 1.05-1.09; and overall proportion mediated (pm): 42%, 95% CI: 27%-73%] and diabetes risk (tnde: 0.96, 95% CI: 0.82-1.14; tnie: 1.25, 95% CI: 1.18-1.33; and pm: 100%, 95% CI: 57%-361%). Hyperuricemia showed significant tnde, te, and tnie, mediated by fatty liver jointly with dyslipidemia (pm = 17%) or BMI (pm = 35%), on prediabetes risk.
CONCLUSION
Hyperuricemia could increase prediabetes or diabetes risk, partially mediated by fatty liver, BMI, and dyslipidemia. Fatty liver is the crucial mediator in the association between hyperuricemia and prediabetes.
Topics: Humans; Prediabetic State; Mediation Analysis; Risk Factors; Hyperuricemia; Diabetes Mellitus; Fatty Liver; Dyslipidemias
PubMed: 37124731
DOI: 10.3389/fendo.2023.1133515 -
Cardiovascular Diabetology Jan 2024Abnormal lipid metabolism poses a risk for prediabetes. However, research on lipid parameters used to predict the risk of prediabetes is scarce, and the significance of...
BACKGROUND
Abnormal lipid metabolism poses a risk for prediabetes. However, research on lipid parameters used to predict the risk of prediabetes is scarce, and the significance of traditional and untraditional lipid parameters remains unexplored in prediabetes. This study aimed to comprehensively evaluate the association between 12 lipid parameters and prediabetes and their diagnostic value.
METHODS
This cross-sectional study included data from 100,309 Chinese adults with normal baseline blood glucose levels. New onset of prediabetes was the outcome of concern. Untraditional lipid parameters were derived from traditional lipid parameters. Multivariate logistic regression and smooth curve fitting were used to examine the nonlinear relationship between lipid parameters and prediabetes. A two-piecewise linear regression model was used to identify the critical points of lipid parameters influencing the risk of prediabetes. The areas under the receiver operating characteristic curve estimated the predictive value of the lipid parameters.
RESULTS
A total of 12,352 participants (12.31%) were newly diagnosed with prediabetes. Following adjustments for confounding covariables, high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol were negatively correlated with prediabetes risk. Conversely, total cholesterol, triglyceride (TG), lipoprotein combine index (LCI), atherogenic index of plasma (AIP), non-HDL-C, atherogenic coefficient, Castelli's index-I, remnant cholesterol (RC), and RC/HDL-C ratio displayed positive correlations. In younger adults, females, individuals with a family history of diabetes, and non-obese individuals, LCI, TG, and AIP exhibited higher predictive values for the onset of prediabetes compared to other lipid profiles.
CONCLUSION
Nonlinear associations were observed between untraditional lipid parameters and the risk of prediabetes. The predictive value of untraditional lipid parameters for prediabetes surpassed that of traditional lipid parameters, with LCI emerging as the most effective predictor for prediabetes.
Topics: Adult; Female; Humans; Prediabetic State; Retrospective Studies; Cross-Sectional Studies; Cholesterol, HDL; Triglycerides; China
PubMed: 38184606
DOI: 10.1186/s12933-023-02103-z -
Scientific Reports Feb 2018Polycystic ovary syndrome (PCOS) is associated with an increased risk of type 2 diabetes mellitus (T2DM) but its association with prediabetes and T2DM is unknown in...
Polycystic ovary syndrome (PCOS) is associated with an increased risk of type 2 diabetes mellitus (T2DM) but its association with prediabetes and T2DM is unknown in Qatar. A cross sectional analysis of 3,017 Qatari subjects from the Qatar Biobank, identified 749 women aged 18-40 years, 720 of whom were assessed by the National Institute for Health (NIH) Guidelines for PCOS. Prediabetes (HbA1c 5.7-6.4% and/or impaired fasting glucose (IFG): fasting plasma glucose (FPG) 100-125 mg/dL (5.6-6.9 mmol/L)), and T2DM (fasting plasma glucose > 125 mg/dL (≥7 mmol/L), and/or HbA1c ≥ 6.5%) were determined. The prevalence of prediabetes was 10.6% and the prevalence of undiagnosed diabetes was found to be 4.0% in the total population. Overall, 12.1% of 720 women had PCOS, of whom FPG and HbA1c were available in 62 women with PCOS: 19.4% had prediabetes and 9.7% had diabetes. An adverse cardiovascular risk profile for IFG women compared to normal women was found. Women with PCOS alone had a similar adverse cardiovascular profile as those with IFG alone and T2DM. Thus, the risk of prediabetes and diabetes is increased in Qatari women with PCOS, with an adverse cardiovascular risk profile similar to that seen in prediabetes and T2DM.
Topics: Adolescent; Adult; Blood Glucose; Body Mass Index; Cross-Sectional Studies; Diabetes Mellitus, Type 2; Fasting; Female; Glycated Hemoglobin; Humans; Polycystic Ovary Syndrome; Prediabetic State; Waist Circumference; Young Adult
PubMed: 29483674
DOI: 10.1038/s41598-018-21987-6 -
Frontiers in Endocrinology 2023Diabetes Mellitus is a public health problem associated with complications such as neuropathy; however, it has been proposed that these may begin to develop during... (Review)
Review
Diabetes Mellitus is a public health problem associated with complications such as neuropathy; however, it has been proposed that these may begin to develop during prediabetes and may also be present in persons with obesity. Diabetic peripheral neuropathy is the presence of signs and/or symptoms of peripheral nerve dysfunction in people living with diabetes, which increases the risk of developing complications and has a deleterious impact on quality of life. As part of the therapeutic protocol for diabetes, screening tests to identify peripheral neuropathy are suggested, however, there are no recommendations for people with prediabetes and obesity without symptoms such as pain, numbness, or paresthesias. Moreover, clinical screening tests that are usually used to recognize this alteration, such as tendon reflex, temperature sensation, and pressure and vibration perception, might be subjective as they depend on the evaluator's experience thus the incorrect application of these tests may not recognize the damage to small or large-nerve fibers. Recent evidence suggests that an objective study such as the impairment of the rate-dependent depression of the H-reflex could be used as a biomarker of spinal disinhibition and hence may provide more information on sensorimotor integration.
Topics: Humans; Prediabetic State; H-Reflex; Quality of Life; Diabetic Neuropathies; Obesity
PubMed: 37476495
DOI: 10.3389/fendo.2023.1206552