A comparison of machine learning algorithms and traditional regression-based statistical modeling for predicting hypertension incidence in a Canadian population

Abstract

Risk prediction models are frequently used to identify individuals at risk of developing hypertension. This study evaluates different machine learning algorithms and compares their predictive performance with the conventional Cox proportional hazards (PH) model to predict hypertension incidence using survival data. This study analyzed 18,322 participants on 24 candidate features from the large Alberta’s Tomorrow Project (ATP) to develop different prediction models. To select the top features, we applied five feature selection methods, including two filter-based: a univariate Cox p-value and C-index; two embedded-based: random survival forest and least absolute shrinkage and selection operator (Lasso); and one constraint-based: the statistically equivalent signature (SES). Five machine learning algorithms were developed to predict hypertension incidence: penalized regression Ridge, Lasso, Elastic Net (EN), random survival forest (RSF), and gradient boosting (GB), along with the conventional Cox PH model. The predictive performance of the models was assessed using C-index. The performance of machine learning algorithms was observed, similar to the conventional Cox PH model. Average C-indexes were 0.78, 0.78, 0.78, 0.76, 0.76, and 0.77 for Ridge, Lasso, EN, RSF, GB and Cox PH, respectively. Important features associated with each model were also presented. Our study findings demonstrate little predictive performance difference between machine learning algorithms and the conventional Cox PH regression model in predicting hypertension incidence. In a moderate dataset with a reasonable number of features, conventional regression-based models perform similar to machine learning algorithms with good predictive accuracy.

Machine learning-based models to predict the conversion of normal blood pressure to hypertension within 5-year follow-up

“…Two factors, WHR and SBP, among the top predictive factors in our study, were in line with the top predictive factors in a similar and robust study conducted in Canada [ 47 ]. This conformity could indicate a percentage of similarity among different populations in predicting future hypertension.…”

“…Based on our ML model, individuals at high risk for developing hypertension can be recommended to modify their lifestyles and behaviors (such as physical activity, dietary changes, smoking cessation, and alcohol consumption) to avoid hypertension and prevent all associated dangerous complications and costs [ 47 ]. It is further recommended to employ new ML models in various geographical regions where there is a wide diversity in hypertension risk factors, as each model may reveal new predictive factors for hypertension [ 28 ].…”

“…Furthermore, we conducted hyperparameter tuning to optimize the performance of our models. Moreover, our study has a significantly larger number and scope of variables compared to most studies conducted in this field, including the Canadian study [ 47 ].…”

Precision Hypertension

“…[18][19][20] However, a study designed to test 5 AI algorithms found that their predictive power was no different to that of traditional regression-based statistical modeling. 21 AI is also likely to be important in the design and analysis of clinical trials. [22][23][24] As described earlier, RCTs have informed the current approach to hypertension management, with recommended treatment guidelines being based on their mean results and expert consensus.…”

“…18–20 However, a study designed to test 5 AI algorithms found that their predictive power was no different to that of traditional regression-based statistical modeling. 21…”