In the Journals

Obesity, dyslipidemia, inflammation: Potential psoriasis comorbidities of coronary burden

A machine learning program with clinical and coronary CT angiography data identified the top predictors of noncalcified coronary burden in psoriasis, which were markers related to obesity, dyslipidemia and inflammation, according to research in Journal of the American Academy of Dermatology.

“Machine learning algorithms open the opportunity to map multiple complex data variables to clinical outcomes that are crucial for the advancement of our understanding of cardiovascular disease risk factors,” Amit K. Dey, MD, postdoctoral fellow of the National Heart, Lung, and Blood Institute at NIH, and colleagues wrote.

Researchers utilized 263 consecutive patient records along with 92 phenotype variables measured at baseline to develop the machine learning algorithm using random forest ensembles as a method to remove excessive variables. Random forests were best suited for this analysis because they allow for decision trees that are unique to every data set, among other important characteristics, according to researchers.

“In this method, we begin by manually removing variables in the data set, and then we grow an ensembles of decision trees to measure variable importance by permutation,” Dey and colleagues wrote.

Patients underwent coronary CT angiography on the same day as blood draw and 1-year scans were performed using the same scanner and protocol.

Twenty-nine variables of the 92 initial variables were deemed redundant and were removed before analysis. Patients with psoriasis at baseline were middle-aged, predominantly men, with low cardiovascular risk by Framingham risk score and had mild to moderate skin disease.

The algorithm found that the top predictors of noncalcified coronary burden in patients with psoriasis were BMI, adiposity, waist-to-hip ratio, apolipoprotein A-I, lipoprotein particles, cholesterol efflux capacity, apolipoprotein B, erythrocyte sedimentation rate, high-sensitivity C-reactive protein, absolute immature granulocyte count, absolute monocyte count and white blood cells.

The analysis only included baseline values from the patients’ first visit, which was reported as a major limitation of the study.

The researchers concluded that the markers related to obesity, dyslipidemia and inflammation predict coronary artery burden and might be important comorbidities to treat in psoriasis. – by Abigail Sutton

 

 

Disclosures: Dey reports no relevant financial disclosures. Please see the study for all other authors’ relevant financial disclosures.

 

A machine learning program with clinical and coronary CT angiography data identified the top predictors of noncalcified coronary burden in psoriasis, which were markers related to obesity, dyslipidemia and inflammation, according to research in Journal of the American Academy of Dermatology.

“Machine learning algorithms open the opportunity to map multiple complex data variables to clinical outcomes that are crucial for the advancement of our understanding of cardiovascular disease risk factors,” Amit K. Dey, MD, postdoctoral fellow of the National Heart, Lung, and Blood Institute at NIH, and colleagues wrote.

Researchers utilized 263 consecutive patient records along with 92 phenotype variables measured at baseline to develop the machine learning algorithm using random forest ensembles as a method to remove excessive variables. Random forests were best suited for this analysis because they allow for decision trees that are unique to every data set, among other important characteristics, according to researchers.

“In this method, we begin by manually removing variables in the data set, and then we grow an ensembles of decision trees to measure variable importance by permutation,” Dey and colleagues wrote.

Patients underwent coronary CT angiography on the same day as blood draw and 1-year scans were performed using the same scanner and protocol.

Twenty-nine variables of the 92 initial variables were deemed redundant and were removed before analysis. Patients with psoriasis at baseline were middle-aged, predominantly men, with low cardiovascular risk by Framingham risk score and had mild to moderate skin disease.

The algorithm found that the top predictors of noncalcified coronary burden in patients with psoriasis were BMI, adiposity, waist-to-hip ratio, apolipoprotein A-I, lipoprotein particles, cholesterol efflux capacity, apolipoprotein B, erythrocyte sedimentation rate, high-sensitivity C-reactive protein, absolute immature granulocyte count, absolute monocyte count and white blood cells.

The analysis only included baseline values from the patients’ first visit, which was reported as a major limitation of the study.

The researchers concluded that the markers related to obesity, dyslipidemia and inflammation predict coronary artery burden and might be important comorbidities to treat in psoriasis. – by Abigail Sutton

 

 

Disclosures: Dey reports no relevant financial disclosures. Please see the study for all other authors’ relevant financial disclosures.