Prediction of dementia based on older adults’ sleep disturbances using machine learning: a controlled experiment.

Abstract: Background. Sleep disturbances can indicate an increased risk of dementia. This study examines whether machine learning can predict this association and which sleep disturbance factors impact dementia. Methods. To assess the association, we use five machine learning algorithms (gradient boosting, logistic regression, gaussian naive Bayes, random forest and support vector machine) and data on the older population (60+, n=4175) in Sweden from the Swedish National Study on Ageing and Care – Blekinge (SNAC-B). The algorithms use 16 features from SNAC-B, which are on personal and sleep disturbance factors. Further, each algorithm uses 10-fold stratified cross-validation to obtain their results, which consist of the Brier score for checking accuracy and the feature importance for examining the risk factors for dementia.  Results. Logistic regression found an association between dementia and sleep disturbances. However, it is slight for the given features. Gradient boosting was the most accurate algorithm with 92.9% accuracy, 0.926 f1-score, 0.974 ROC AUC and 0.056 Brier score. The significant risk factors were different in each machine-learning algorithm. If the person sleeps more than two hours during the day, their sex, education level, age, waking up during the night and if the person snores are the variables that most consistently have the highest feature importance across algorithms. Conclusions. There is an association between sleep disturbances and dementia, which machine learning algorithms can predict. Furthermore, the risk factors for dementia vary across the algorithms, but sleep disturbances can predict dementia.