ABSTRACT
The systematic assessment of cognitive performance of older people without cognitive complaints is controversial and unfeasible. Identifying individuals at higher risk of cognitive impairment could optimize resource allocation. We aimed to develop and test machine learning models to predict cognitive impairment using variables obtainable in primary care settings. In this cross-sectional study, we included 8,291 participants of the baseline assessment of the ELSA-Brasil study, who were aged between 50 and 74 years and were free of dementia. Cognitive performance was assessed with a neuropsychological battery and cognitive impairment was defined as global cognitive z-score below 2 standard deviations. Variables used as input to the prediction models included demographics, social determinants, clinical conditions, family history, lifestyle, and laboratory tests. We developed machine learning models using logistic regression, neural networks, and gradient boosted trees. Participants' mean age was 58.3±6.2 years, 55% were female. Cognitive impairment was present in 328 individuals (4%). Machine learning algorithms presented fair to good discrimination (areas under the ROC curve between 0.801 and 0.873). Extreme Gradient Boosting presented the highest discrimination, high specificity (97%), and negative predictive value (97%). Seventy-six percent of the individuals with cognitive impairment were included among the highest ranked individuals by this algorithm. In conclusion, we developed and tested a machine learning model to predict cognitive impairment based on primary care data that presented good discrimination and high specificity. These characteristics could support the detection of patients who would not benefit from cognitive assessment, facilitating the allocation of human and economic resources.
ABSTRACT
Sarcopenia and sleep problems share common physiopathology. We aimed to investigate the association of sleep disturbances with sarcopenia and its defining components in Brazilian middle-aged and older adults. In this cross-sectional analysis of the second wave of the ELSA-Brasil study, we included data from 7948 participants aged 50 years and older. Muscle mass was evaluated by bioelectrical impedance analysis and muscle strength by hand-grip strength. Sarcopenia was defined according to the Foundation for the National Institutes of Health criteria. Sleep duration and insomnia complaint were self-reported. Short sleep duration was considered as ≤6 h/night and long sleep duration as >8 h/night. High risk of obstructive sleep apnea (OSA) was assessed using the STOP-Bang questionnaire. Possible confounders included socio-demographic characteristics, lifestyle, clinical comorbidities, and use of sedatives and hypnotics. The frequencies of sarcopenia, low muscle mass, and low muscle strength were 1.6, 21.1, and 4.1%, respectively. After adjustment for possible confounders, high risk of OSA was associated with low muscle mass (OR=2.17, 95%CI: 1.92-2.45). Among obese participants, high risk of OSA was associated with low muscle strength (OR=1.68, 95%CI: 1.07-2.64). However, neither short nor long sleep duration or frequent insomnia complaint were associated with sarcopenia or its defining components. In conclusion, high risk of OSA was associated with low muscle mass in the whole sample and with low muscle strength among obese participants. Future studies are needed to clarify the temporal relationship between both conditions.