Development of a novel risk prediction and risk stratification score for polycystic ovary syndrome

Harshal Deshmukh, Maria Papageorgiou, Eric S. Kilpatrick, Stephen Atkin, Thozhukat Sathyapalan

Research output: Contribution to journalArticle

Abstract

Objective: The aim of this study was to develop a simple phenotypic algorithm that can capture the underlying clinical and hormonal abnormalities to help in the diagnosis and risk stratification of polycystic ovary syndrome (PCOS). Methods: The study consisted of 111 women with PCOS fulfilling the Rotterdam diagnostic criteria and 67 women without PCOS. A Firth's penalized logistic regression model was used for independent variable section. Model optimism, discrimination and calibration were assessed using bootstrapping, area under the curve (AUC) and Hosmer-Lemeshow statistics, respectively. The prognostic index (PI) and risk score for developing PCOS were calculated using independent variables from the regression model. Results: Firth penalized logistic regression model with backward selection identified four independent predictors of PCOS namely free androgen index [β 0.30 (0.12), P = 0.008], 17-OHP [β = 0.20 (0.01), P = 0.026], anti-mullerian hormone [AMH; β = 0.04 (0.01) P < 0.0001] and waist circumference [β = 0.08 (0.02), P < 0.0001]. The model estimates indicated high internal validity (minimal optimism on 1000-fold bootstrapping), good discrimination ability (bias corrected c-statistic = 0.90) and good calibration (Hosmer-Lemeshow χ2 = 3.7865). PCOS women with a high-risk score (q1 + q2 + q3 vs q4) presented with a worse metabolic profile characterized by a higher 2-hour glucose (P = 0.01), insulin (P = 0.0003), triglycerides (P = 0.0005), C-reactive protein (P < 0.0001) and low HDL-cholesterol (P = 0.02) as compared to those with lower risk score for PCOS. Conclusions: We propose a simple four-variable model, which captures the underlying clinical and hormonal abnormalities in PCOS and can be used for diagnosis and metabolic risk stratification in women with PCOS.

Original languageEnglish
JournalClinical Endocrinology
DOIs
Publication statusAccepted/In press - 1 Jan 2018

Fingerprint

Polycystic Ovary Syndrome
Logistic Models
Calibration
Anti-Mullerian Hormone
Aptitude
Metabolome
Waist Circumference
C-Reactive Protein
HDL Cholesterol
Androgens
Area Under Curve
Triglycerides
Insulin
Glucose

Keywords

  • 17-OHP
  • AMH
  • FAI
  • PCOS
  • risk score

ASJC Scopus subject areas

  • Endocrinology, Diabetes and Metabolism
  • Endocrinology

Cite this

Development of a novel risk prediction and risk stratification score for polycystic ovary syndrome. / Deshmukh, Harshal; Papageorgiou, Maria; Kilpatrick, Eric S.; Atkin, Stephen; Sathyapalan, Thozhukat.

In: Clinical Endocrinology, 01.01.2018.

Research output: Contribution to journalArticle

Deshmukh, Harshal ; Papageorgiou, Maria ; Kilpatrick, Eric S. ; Atkin, Stephen ; Sathyapalan, Thozhukat. / Development of a novel risk prediction and risk stratification score for polycystic ovary syndrome. In: Clinical Endocrinology. 2018.
@article{7073e8691d674702a5e60b8aa3b41262,
title = "Development of a novel risk prediction and risk stratification score for polycystic ovary syndrome",
abstract = "Objective: The aim of this study was to develop a simple phenotypic algorithm that can capture the underlying clinical and hormonal abnormalities to help in the diagnosis and risk stratification of polycystic ovary syndrome (PCOS). Methods: The study consisted of 111 women with PCOS fulfilling the Rotterdam diagnostic criteria and 67 women without PCOS. A Firth's penalized logistic regression model was used for independent variable section. Model optimism, discrimination and calibration were assessed using bootstrapping, area under the curve (AUC) and Hosmer-Lemeshow statistics, respectively. The prognostic index (PI) and risk score for developing PCOS were calculated using independent variables from the regression model. Results: Firth penalized logistic regression model with backward selection identified four independent predictors of PCOS namely free androgen index [β 0.30 (0.12), P = 0.008], 17-OHP [β = 0.20 (0.01), P = 0.026], anti-mullerian hormone [AMH; β = 0.04 (0.01) P < 0.0001] and waist circumference [β = 0.08 (0.02), P < 0.0001]. The model estimates indicated high internal validity (minimal optimism on 1000-fold bootstrapping), good discrimination ability (bias corrected c-statistic = 0.90) and good calibration (Hosmer-Lemeshow χ2 = 3.7865). PCOS women with a high-risk score (q1 + q2 + q3 vs q4) presented with a worse metabolic profile characterized by a higher 2-hour glucose (P = 0.01), insulin (P = 0.0003), triglycerides (P = 0.0005), C-reactive protein (P < 0.0001) and low HDL-cholesterol (P = 0.02) as compared to those with lower risk score for PCOS. Conclusions: We propose a simple four-variable model, which captures the underlying clinical and hormonal abnormalities in PCOS and can be used for diagnosis and metabolic risk stratification in women with PCOS.",
keywords = "17-OHP, AMH, FAI, PCOS, risk score",
author = "Harshal Deshmukh and Maria Papageorgiou and Kilpatrick, {Eric S.} and Stephen Atkin and Thozhukat Sathyapalan",
year = "2018",
month = "1",
day = "1",
doi = "10.1111/cen.13879",
language = "English",
journal = "Clinical Endocrinology",
issn = "0300-0664",
publisher = "Wiley-Blackwell",

}

TY - JOUR

T1 - Development of a novel risk prediction and risk stratification score for polycystic ovary syndrome

AU - Deshmukh, Harshal

AU - Papageorgiou, Maria

AU - Kilpatrick, Eric S.

AU - Atkin, Stephen

AU - Sathyapalan, Thozhukat

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Objective: The aim of this study was to develop a simple phenotypic algorithm that can capture the underlying clinical and hormonal abnormalities to help in the diagnosis and risk stratification of polycystic ovary syndrome (PCOS). Methods: The study consisted of 111 women with PCOS fulfilling the Rotterdam diagnostic criteria and 67 women without PCOS. A Firth's penalized logistic regression model was used for independent variable section. Model optimism, discrimination and calibration were assessed using bootstrapping, area under the curve (AUC) and Hosmer-Lemeshow statistics, respectively. The prognostic index (PI) and risk score for developing PCOS were calculated using independent variables from the regression model. Results: Firth penalized logistic regression model with backward selection identified four independent predictors of PCOS namely free androgen index [β 0.30 (0.12), P = 0.008], 17-OHP [β = 0.20 (0.01), P = 0.026], anti-mullerian hormone [AMH; β = 0.04 (0.01) P < 0.0001] and waist circumference [β = 0.08 (0.02), P < 0.0001]. The model estimates indicated high internal validity (minimal optimism on 1000-fold bootstrapping), good discrimination ability (bias corrected c-statistic = 0.90) and good calibration (Hosmer-Lemeshow χ2 = 3.7865). PCOS women with a high-risk score (q1 + q2 + q3 vs q4) presented with a worse metabolic profile characterized by a higher 2-hour glucose (P = 0.01), insulin (P = 0.0003), triglycerides (P = 0.0005), C-reactive protein (P < 0.0001) and low HDL-cholesterol (P = 0.02) as compared to those with lower risk score for PCOS. Conclusions: We propose a simple four-variable model, which captures the underlying clinical and hormonal abnormalities in PCOS and can be used for diagnosis and metabolic risk stratification in women with PCOS.

AB - Objective: The aim of this study was to develop a simple phenotypic algorithm that can capture the underlying clinical and hormonal abnormalities to help in the diagnosis and risk stratification of polycystic ovary syndrome (PCOS). Methods: The study consisted of 111 women with PCOS fulfilling the Rotterdam diagnostic criteria and 67 women without PCOS. A Firth's penalized logistic regression model was used for independent variable section. Model optimism, discrimination and calibration were assessed using bootstrapping, area under the curve (AUC) and Hosmer-Lemeshow statistics, respectively. The prognostic index (PI) and risk score for developing PCOS were calculated using independent variables from the regression model. Results: Firth penalized logistic regression model with backward selection identified four independent predictors of PCOS namely free androgen index [β 0.30 (0.12), P = 0.008], 17-OHP [β = 0.20 (0.01), P = 0.026], anti-mullerian hormone [AMH; β = 0.04 (0.01) P < 0.0001] and waist circumference [β = 0.08 (0.02), P < 0.0001]. The model estimates indicated high internal validity (minimal optimism on 1000-fold bootstrapping), good discrimination ability (bias corrected c-statistic = 0.90) and good calibration (Hosmer-Lemeshow χ2 = 3.7865). PCOS women with a high-risk score (q1 + q2 + q3 vs q4) presented with a worse metabolic profile characterized by a higher 2-hour glucose (P = 0.01), insulin (P = 0.0003), triglycerides (P = 0.0005), C-reactive protein (P < 0.0001) and low HDL-cholesterol (P = 0.02) as compared to those with lower risk score for PCOS. Conclusions: We propose a simple four-variable model, which captures the underlying clinical and hormonal abnormalities in PCOS and can be used for diagnosis and metabolic risk stratification in women with PCOS.

KW - 17-OHP

KW - AMH

KW - FAI

KW - PCOS

KW - risk score

UR - http://www.scopus.com/inward/record.url?scp=85056716785&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85056716785&partnerID=8YFLogxK

U2 - 10.1111/cen.13879

DO - 10.1111/cen.13879

M3 - Article

JO - Clinical Endocrinology

JF - Clinical Endocrinology

SN - 0300-0664

ER -