Relation of insulin resistance with social-demographics, adiposity and behavioral factors in non-diabetic adult Canadians

The association between the Canadian active living environments index and glucose metabolism in a Canadian national population study

Biomarkers of glucose metabolism may reflect insulin resistance, a risk factor for diabetes and cardiovascular disease (CVD). Neighborhoods conducive to a physically active lifestyle have the potential to improve these biomarkers. We examined cross-sectional associations between walkability and blood biomarkers of glucose metabolism in 29,649 Canadian Health Measures Survey (CHMS) participants. We used generalized linear mixed models with sampling weights adjusted for province, participants' age, sex, annual household income and educational attainment, cigarette smoking, environmental tobacco smoke, alcohol consumption, and exposure to ambient fine particulate air pollution (PM2.5). A higher value of the Canadian Active Living Environments Index, a measure of neighborhood walkability, equivalent to the magnitude of its interquartile range (IQR) of 2.4 was significantly associated with percentage differences of -0.48 (95% confidence interval (CI): 0.63, -0.32), -3.17 (95%CI: 5.27, -1.08), -3.88 (95%CI: 6.38, -1.38), and -3.36 (95%CI: 5.25, -1.47) in HbA1C, fasting insulin, HOMA-IR, and HOMA-β, respectively, for all CHMS participants. No significant effects were observed in those ≤16 years old. Canadians living in neighborhoods that facilitate active living have more favorable biomarkers of glucose metabolism, suggesting that the built environment has the potential to improve risk factors for diabetes and CVD in adults.

Triglyceride/high-density lipoprotein ratio as a predictor for insulin resistance in a sample of healthy Iraqi adults

Insulin resistance (IR) is a common pathophysiological condition associated with many metabolic diseases, including obesity, prediabetes, type 2 diabetes mellitus (T2DM), and cardiovascular disease. The widely used homeostasis model assessment for IR (HOMA-IR) is usually used to estimate IR. However, this model cannot be used for screening IR due to several logistic difficulties, such as costs and insulin instability, which are essential for measurement. Thus, finding feasible alternatives is of paramount importance. This study aimed to evaluate the value of triglyceride/high-density lipoprotein-cholesterol (TG/HDL-c) ratio in predicting IR in healthy adult individuals. The study involved 83 euglycemic non-diabetic adults (≥45 years old). Lipid profile, fasting insulin, fasting blood sugar (FBS), and glycated hemoglobin were measured for all participants. The TG/HDL-c ratio was calculated by dividing TG by HDL. Insulin resistance was calculated using the HOMA-IR formula. The receiver operating characteristic (ROC) curve was used to evaluate the predictive value of the TG/HDL-c ratio. The prevalence of IR among healthy adult Iraqis was 28.92%. Subjects in the IR group showed a higher TG/HDL ratio than the insulin-sensitive group (3.69±0.68 versus 2.71±1.0) with a significant difference. The area under the curve (AUC) for this ratio was 0.849, 95% CI= 0.763-0.935, p<0.002. The sensitivity and specificity of the test were 83% and 81%, respectively, at a best cut-off value of 3.1 (TG/HDL ratio). The prevalence of IR among healthy adult Iraqis was 28.92%. Triglyceride/HDL-c ratio had a very good predictive value for IR.

Altered Visceral Adipose Tissue Predictors and Women’s Health: A Unicenter Study

(1) Background: The excess visceral adipose tissue (VAT) accumulation in women may reflect an early or advanced state of a metabolic disorder and a higher risk of cardiovascular disease than other types of obesity. This study aimed to determine the predictor variables (demographic information, anthropometric data, and blood biomarkers) for changes in VAT in adult women. (2) Methods: This cross-sectional study was conducted with women aged 18–59 years attending nutritional consultation at the Centro Universitário de Brasília (CEUB)’s nutrition school clinic, Brazil. All participants’ medical records were reviewed throughout the study and data of interest were collected. Various anthropometric measurements and biochemical exams were performed and analyzed in a univariate logistic regression model to identify the possible risk factors predictors for the presence of altered VAT. (3) Results: Our logistic regression model considered body mass index (BMI) greater than 25 kg/m², lipid accumulation product (LAP), and waist–hip ratio (WHR) as predictors of altered VAT. (4) Conclusion: LAP has a robust predictive capacity for changes in visceral fat in adult women, followed by WHR and BMI, making these variables effective in assessing the risk for changes in visceral fat and their inclusion essential in the individual and collective clinical practice.

Quantifying the Homeostatic Model Assessment of Insulin Resistance to Predict Mortality in Multi-organ Dysfunction Syndrome

Background

Insulin resistance is an integral component of a multi-organ dysfunction syndrome (MODS) associated with increased mortality. We determined a cutoff value for the homeostatic model assessment of insulin resistance (HOMA-IR) during an ICU admission that could predict 28-day mortality of nondiabetic MODS patients.

Materials and methods

In this prospective, outcome assessor blinded cohort design, we evaluated 82 such patients for fasting blood glucose (FBG)/insulin levels (FIL) during an ICU admission and followed their outcome for 28 days. The primary outcome variable was the HOMA-IR score calculated from the above variables. The statistical tool included receiver operating characteristic curve, Youden index, and correlation and regression analysis.

Results

Overall, 38 patients succumbed to their illness. The optimal cutoff value for HOMA-IR was ≥1.61 (area under curve: 0.684, sensitivity: 36.8%, specificity: 95.5%). The 28-day survival was significantly lower (p = 0.001) at HOMA-IR threshold ≥1.61 (odds ratio: 12.25, hazard ratio: 2.98). The mean HOMA-IR among survivors vs nonsurvivors was 0.76 ± 0.61 and 1.38 ± 1.14, respectively (p = 0.004). Except for FIL and FBG, HOMA-IR values did not correlate with any other baseline or outcome parameters (demographics, APACHE II/sequential organ failure assessment score, vasopressor needs, or ICU/hospital stay). On comparing these parameters across the HOMA-IR threshold, only FIL and the hospital stay varied significantly. Most of the outcome parameters, however, varied significantly among nonsurvivors vs survivors.

Conclusion

The HOMA-IR is a significant predictor of mortality in MODS. Its cutoff value may assist in determining a reference range for critically ill patients. Its routine use in the light of other disease severity scores may serve in their better prognostication.

How to cite this article

Sama S, Jain G, Kant R, Bhadoria AS, Naithani M, Kumar A. Quantifying the Homeostatic Model Assessment of Insulin Resistance to Predict Mortality in Multi-organ Dysfunction Syndrome. Indian J Crit Care Med 2021;25(12):1364–1369.

Simple Method to Predict Insulin Resistance in Children Aged 6-12 Years by Using Machine Learning

BACKGROUND

Due to the increasing insulin resistance (IR) in childhood, rates of diabetes and cardiovascular disease may rise in the future and seriously threaten the healthy development of children. Finding an easy way to predict IR in children can help pediatricians to identify these children in time and intervene appropriately, which is particularly important for practitioners in primary health care.

PATIENTS AND METHODS

Seventeen features from 503 children 6-12 years old were collected. We defined IR by HOMA-IR greater than 3.0, thus classifying children with IR and those without IR. Data were preprocessed by multivariate imputation and oversampling to resolve missing values and data imbalances; then, recursive feature elimination was applied to further select features of interest, and 5 machine learning methods-namely, logistic regression (LR), support vector machine (SVM), random forest (RF), extreme gradient boosting (XGBoost), and gradient boosting with categorical features support (CatBoost)-were used for model training. We tested the trained models on an external test set containing information from 133 children, from which performance metrics were extracted and the optimal model was selected.

RESULTS

After feature selection, the numbers of chosen features for the LR, SVM, RF, XGBoost, and CatBoost models were 6, 9, 10, 14, and 6, respectively. Among them, glucose, waist circumference, and age were chosen as predictors by most of the models. Finally, all 5 models achieved good performance on the external test set. Both XGBoost and CatBoost had the same AUC (0.85), which was highest among those of all models. Their accuracy, sensitivity, precision, and F1 scores were also close, but the specificity of XGBoost reached 0.79, which was significantly higher than that of CatBoost, so XGBoost was chosen as the optimal model.

CONCLUSION

The model developed herein has a good predictive ability for IR in children 6-12 years old and can be clinically applied to help pediatricians identify children with IR in a simple and inexpensive way.

[Consumption of red and processed meat, insulin resistance, and diabetes in the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil)Consumo de carnes rojas y de carne procesada, resistencia a la insulina y diabetes en el estudio longitudinal de salud del adulto (ELSA-Brasil)]

Objetivo.

Investigar a associação entre consumo de carne vermelha e processada e a ocorrência de novos casos de resistência insulínica (RI) e diabetes mellitus (DM) em participantes do Estudo Longitudinal de Saúde do Adulto (ELSA-Brasil).

Métodos.

Estudo de coorte com 15 105 servidores públicos (idade: 35 a 74 anos). Dados bioquímicos, antropométricos, socioeconômicos e de estilo de vida foram coletados na linha de base (2008–2010) e na segunda onda (2012–2014). O consumo de carnes (g/dia) foi estimado por questionário de frequência alimentar. Para categorizar baixo, médio e alto consumo as variáveis independentes foram divididas em tercis. DM foi diagnosticado como glicemia de jejum ≥ 126 mg/dL, glicose pós-sobrecarga ≥ 200 mg/dL ou hemoglobina glicada ≥ 6,5. RI foi determinada pelo índice HOMA-IR com pontos de corte construídos a partir do percentil 75 da amostra.

Resultados.

Homens e participantes com menor renda e escolaridade relataram maior consumo de carne vermelha e processada. Maior consumo de carne processada (último tercil, > 27,1 g/dia) associou-se a novos casos de RI em homens (OR = 1,68; IC95%: 1,31 a 2,16) e mulheres (OR = 1,23; IC95%: 1,00 a 1,52). Alto consumo de carne vermelha aumentou em 40% (IC95%: 1,04 a 1,96) a chance de novos casos de DM em homens.

Conclusões.

O consumo elevado de carne vermelha e processada teve impacto negativo na saúde dos participantes. O consumo moderado de carnes pode ser recomendado para a população em geral e para prevenção do DM.

High waist-to-hip ratio levels are associated with insulin resistance markers in normal-weight women

AIM

To assess the association between high waist-to-hip ratio (WHR) levels and insulin resistance (IR) or hyperinsulinemia after oral glucose tolerance test (OGTT) in a sample of normal-weight women.

METHODS

We conducted an analytical cross-sectional study in euthyroid non-diabetic women, who attended the outpatient service of a private clinic in Lima-Peru from 2012 to 2016. Participants were divided in two groups according to the presence or absence of high WHR levels, IR or hyperinsulinemia after OGTT. We considered WHR values > 0.85 as high levels. IR was defined as a Homeostasis Model Assessment (HOMA-IR) value > 2.39 and hyperinsulinemia after OGTT as a serum insulin value ≥ 80μU/mL after 120 min of 75-g glucose intake. We elaborated crude and adjusted Poisson generalized linear models to evaluate the association between high WHR levels and IR or hyperinsulinemia after OGTT and reported the prevalence ratio (PR) with their respective 95% confidence intervals (95%CI).

RESULTS

We analyzed the data of 248 euthyroid, non-diabetic and normal-weight women. The prevalence of high WHR levels was 68.9% (n = 171) while the prevalence of IR and hyperinsulinemia after OGTT was 25% (n = 62) and 15.3% (n = 38), respectively. WHR values were positively correlated with HOMA-IR (r = 0.307; p < 0.001) and serum insulin after OGTT (r = 0.260; p < 0.001). In the adjusted model, high WHR levels were associated with both IR (aPR = 2.63; 95%CI: 1.39-5.01) and hyperinsulinemia after OGTT (aPR = 2.35; 95%CI: 1.03-5.38).

CONCLUSION

High WHR levels were associated with both IR markers used in our study, appearing to be a useful anthropometric indicator to assess IR in euthyroid normal-weight women without type 2 diabetes mellitus.

Neural, Hormonal, and Cognitive Correlates of Metabolic Dysfunction and Emotional Reactivity

OBJECTIVE

Prediabetes and type 2 diabetes (i.e., hyperglycemia) are characterized by insulin resistance. These problems with energy metabolism may exacerbate emotional reactivity to negatively valenced stimuli and related phenomena such as predisposition toward negative affect, as well as cognitive deficits. Higher emotional reactivity is seen with hyperglycemia and insulin resistance. However, it is largely unknown how metabolic dysfunction correlates with related neural, hormonal, and cognitive outcomes.

METHODS

Among 331 adults from the Midlife in the United States study, eye-blink response (EBR) we cross sectionally examined to gauge reactivity to negative, positive, or neutrally valenced pictures from international affect picture system stimuli proximal to an acoustic startle probe. Increased EBR to negative stimuli was considered an index of stress reactivity. Frontal alpha asymmetry, a biomarker of negative affect predisposition, was determined using resting electroencephalography. Baseline urinary cortisol output was collected. Cognitive performance was gauged using the Brief Test of Adult Cognition by telephone. Fasting glucose and insulin characterized hyperglycemia or the homeostatic model assessment of insulin resistance.

RESULTS

Higher homeostatic model assessment of insulin resistance corresponded to an increased startle response, measured by EBR magnitude, for negative versus positive stimuli (R = 0.218, F(1,457) = 5.48, p = .020, euglycemia: M(SD) = .092(.776), hyperglycemia: M(SD) = .120(.881)). Participants with hyperglycemia versus euglycemia showed greater right frontal alpha asymmetry (F(1,307) = 6.62, p = .011, euglycemia: M(SD) = .018(.167), hyperglycemia: M(SD) = -.029(.160)), and worse Brief Test of Adult Cognition by telephone arithmetic performance (F(1,284) = 4.25, p = .040, euglycemia: M(SD) = 2.390(1.526), hyperglycemia: M(SD) = 1.920(1.462)). Baseline urinary cortisol (log10 μg/12 hours) was also dysregulated in individuals with hyperglycemia (F(1,324) = 5.09, p = .025, euglycemia: M(SD) = 1.052 ± .332, hyperglycemia: M(SD) = .961 (.362)).

CONCLUSIONS

These results suggest that dysmetabolism is associated with increased emotional reactivity, predisposition toward negative affect, and specific cognitive deficits.