Identifying non-adult attention-deficit/hyperactivity disorder individuals using a stacked machine learning algorithm using administrative data population registers in a universal healthcare system

Background

This research project aims to build a Machine Learning algorithm (ML) to predict first-time ADHD diagnosis, given that it is the most frequent mental disorder for the non-adult population.

Methods

We used a stacked model combining 4 ML approaches to predict the presence of ADHD. The dataset contains data from population health care administrative registers in Catalonia comprising 1,225,406 non-adult individuals for 2013-2017, linked to socioeconomic characteristics and dispensed drug consumption. We defined a measure of proper ADHD diagnoses based on medical factors.

Results

We obtained an AUC of 79.6% with the stacked model. Significant variables that explain the ADHD presence are the dispersion across patients' visits to healthcare providers; the number of visits, diagnoses related to other mental disorders and drug consumption; age, and sex.

Conclusions

ML techniques can help predict ADHD early diagnosis using administrative registers. We must continuously investigate the potential use of ADHD early detection strategies and intervention in the health system.

Early prey intake of a short-finned pilot whale (Globicephala macrorhynchus Gray, 1846, Cetacea: Delphinidae) in the Canary Islands

This study reveals early prey eating by a short-finned pilot whale (Globicephala macrorhynchus Gray, 1846, Cetacea: Delphinidae) in the Canary Islands. Stomach contents, trophic markers, skin isotopic ratios of nitrogen (δ15N:15N/14N) and carbon (δ13C:13C/12C), and fatty acid profiles of the blubber of a short-finned pilot whale of 213 cm size euthanized in free-ranging conditions were analyzed. A total of 15 species of oegopsid squids, mostly diel vertical mesopelagic migrant species of the families Enoploteuthidae, Ommastrephidae, and Histioteuthidae, as well as mother's milk, were identified in the stomach contents. Asperoteuthis acanthoderma (Lu, 1977, Cephalopoda: Chiroteuthidae) was found as first time in this area, suggesting the possibility of its presence on both sides of the subtropical Atlantic, extending its current known distribution. The δ15N value (11.55‰) was higher than expected based on the size range of squid ingested, but lower than that of adult pilot whales, suggesting that mother's milk intake has a significant effect on these values in calves. Similarly, the δ13C values (-17.99‰) were shifted to those of adult pilot whales rather than the ingested squids, also due to the ingestion of high-fat breast milk. The fatty acid (FA) composition of blubber showed a clear stratification. Long-chain polyunsaturated fatty acids (LC-PUFA) were mainly present in the inner layer, while most relevant ≤C20 monounsaturated fatty acids (MUFA) were more abundant in the outer layer.

Fibroblast growth factor 23 independently predicts adverse outcomes after an acute coronary syndrome

AIMS

Abnormalities of mineral metabolism (MM) have been related to cardiovascular disorders. There are no reports on the prognostic role of MM after an acute coronary syndrome (ACS). We aim to assess the prognostic role of MM after an ACS.

METHODS AND RESULTS

Plasma levels of components of MM [fibroblast growth factor 23 (FGF23), calcidiol, parathormone, klotho, and phosphate], high-sensitivity C-reactive protein, and N-terminal-pro-brain natriuretic peptide were measured in 1190 patients at discharge from an ACS. The primary outcome was a combination of acute ischaemic events, heart failure (HF) and death. Secondary outcomes were the separate components of the primary outcome. Age was 61.7 ± 12.2 years, and 77.1% were men. Median follow-up was 5.44 (3.03-7.46) years. Two hundred and ninety-four patients developed the primary outcome. At multivariable analysis FGF23 (hazard ratio, HR 1.18 [1.08-1.29], P < 0.001), calcidiol (HR 0.86 [0.74-1.00], P = 0.046), previous coronary or cerebrovascular disease, and hypertension were independent predictors of the primary outcome. The predictive power of FGF23 was homogeneous across different subgroups of population. FGF23 (HR 1.45 [1.28-1.65], P < 0.001) and parathormone (HR 1.06 1.01-1.12]; P = 0.032) resulted as independent predictors of HF. FGF23 (HR 1.21 [1.07-1.37], P = 0.002) and calcidiol (HR 0.72 [0.54-0.97), P = 0.028) were independent predictors of death. No biomarker predicted acute ischaemic events. FGF23 predicted independently the primary outcome in patients with estimated glomerular filtration rate > 60 mL/min/1.73 m2 .

CONCLUSIONS

FGF23 and other components of MM are independent predictors of HF and death after an ACS. This effect is homogeneous across different subgroups of population, and it is not limited to patients with chronic kidney disease.