Effect of model updating strategies on the performance of prevalent diabetes risk prediction models in a mixed-ancestry population of South Africa

Recalibration of a Non-Laboratory-Based Risk Model to Estimate Pre-Diabetes/Diabetes Mellitus Risk in Primary Care in Hong Kong

INTRODUCTION/OBJECTIVES

A non-laboratory-based pre-diabetes/diabetes mellitus (pre-DM/DM) risk prediction model developed from the Hong Kong Chinese population showed good external discrimination in a primary care (PC) population, but the estimated risk level was significantly lower than the observed incidence, indicating poor calibration. This study explored whether recalibrating/updating methods could improve the model's accuracy in estimating individuals' risks in PC.

METHODS

We performed a secondary analysis on the model's predictors and blood test results of 919 Chinese adults with no prior DM diagnosis recruited from PC clinics from April 2021 to January 2022 in HK. The dataset was randomly split in half into a training set and a test set. The model was recalibrated/updated based on a seven-step methodology, including model recalibrating, revising and extending methods. The primary outcome was the calibration of the recalibrated/updated models, indicated by calibration plots. The models' discrimination, indicated by the area under the receiver operating characteristic curves (AUC-ROC), was also evaluated.

RESULTS

Recalibrating the model's regression constant, with no change to the predictors' coefficients, improved the model's accuracy (calibration plot intercept: -0.01, slope: 0.69). More extensive methods could not improve any further. All recalibrated/updated models had similar AUC-ROCs to the original model.

CONCLUSION

The simple recalibration method can adapt the HK Chinese pre-DM/DM model to PC populations with different pre-test probabilities. The recalibrated model can be used as a first-step screening tool and as a measure to monitor changes in pre-DM/DM risks over time or after interventions.

Consumption of Plant Foods and Its Association with Cardiovascular Disease Risk Profile in South Africans at High-Risk of Type 2 Diabetes Mellitus

We assessed the distribution and association of cardiovascular disease (CVD) risk factors by plant foods consumption in individuals at high-risk for type 2 diabetes mellitus. This cross-sectional study utilized baseline data of 693 participants in the South African Diabetes Prevention Programme. Participants underwent a physical examination, biochemical analysis, and dietary assessment using a single non-quantified 24-h recall. Group comparisons were conducted to explore the distribution and associations of common CVD risk factors by plant foods consumption. The mean age of the participants was 51 years, with 81% being females. Consumers of yellow-coloured vitamin A-rich vegetables and tubers and maize had significantly lower systolic blood pressure, fasting insulin, low-density lipoprotein cholesterol, triglycerides, and fibrinogen levels. Cereals consumption increased the likelihood of obesity (OR = 1.72 95% CI [1.09, 2.70] p = 0.019) while the consumption of white roots and tubers decreased the likelihood of obesity (AOR = 0.64 95% CI [0.41, 1.00] p = 0.048). This study reported the consumption of some healthy plant foods with lower levels of, and decreased risk for, some CVD risk factors. A further in-depth investigation is needed to understand these associations.

Marginal versus conditional odds ratios when updating risk prediction models

Risk prediction models often need to be updated when applied to new settings. A simple updating method involves fixed odds ratio transformation of predicted risks to adjust the model for outcome prevalence in the new setting. When a sample from the target population is available, the gold standard is to use a logistic regression model to estimate this odds ratio. A simpler method has been proposed that calculates this odds ratio from the prevalence estimates in the original and new samples. We show that the marginal odds ratio estimated in this way is generally closer to one than the correct (conditional) odds ratio; thus, the simpler method should be avoided when individual-level data are available. When prevalence estimates are the only information at hand, we suggest an approximate method for recovering the conditional odds ratio from the variance of predicted risks in the development sample. Brief simulations and examples show that this approach reduces undercorrection, often substantially.

Examining social determinants of undiagnosed diabetes in Namibia and South Africa using a behavioral model of health services use

AIMS

To examine factors associated with undiagnosed diabetes in Namibia and South Africa.

METHODS

This study used the most recent Demographic and Health Surveys (DHS) from Namibia (2013) and South Africa (2016). This study focused on adults at 35-64 years old. Using Andersen's Behavioral Model, potential contributing factors were categorized into predisposing factors (sex and education), enabling factors (wealth, health insurance, and residence), and a need factor (age, BMI, and high blood pressure). Separate multivariable logistic regression models were used to examine factors associated with undiagnosed diabetes in Namibia (N = 242) and South Africa (N = 525).

RESULTS

In Namibia, higher odds of having undiagnosed diabetes were associated with rural residence (adjusted odds ratio (aOR) = 2.21) and age younger than 45 years old (aOR = 3.20). In South Africa, odds of having undiagnosed diabetes were higher among the poorest-to-poorer group than it was in the richer-to-richest group (aOR = 2.33). In both countries, having high blood pressure was associated with lower odds of having undiagnosed diabetes (aOR = 0.31 in Namibia; aOR = 0.21 in South Africa).

DISCUSSION

Different enabling and need factors were associated with undiagnosed diabetes in these two countries, which implies potentially-different mechanisms driving the high prevalence of undiagnosed diabetes, as well as the needs for different solutions.