Scholarlink

<p>The Body Roundness Index (BRI) has recently been proposed as an effective metric for assessing obesity-related health risks, but its association with female infertility remains insufficiently studied. This study aims to evaluate the predictive power of BRI in comparison to traditional obesity measures (waist circumference (WC) and body mass index (BMI)) for predicting female infertility using data from the National Health and Nutrition Examination Survey (NHANES). A total of 2962 women aged 18–45 years were included in this cross-sectional analysis. Multivariable logistic regression models were employed to examine the relationship between BRI and infertility, adjusting for demographic, behavioral, and metabolic factors. In addition, Generalized Additive Models (GAM) and Restricted Cubic Spline (RCS) regression were used to investigate potential non-linear associations, and machine learning algorithms were applied to assess the predictive performance and identify significant features. Our results demonstrated a significant positive association between BRI and female infertility. In the fully adjusted model, each unit increase in BRI was associated with 18% higher odds of infertility (OR = 1.18, <i>p</i> = 0.032). A dose-response relationship was also observed across BRI quartiles, with women in the highest quartile (Q4) exhibiting 125% higher odds of infertility compared to those in the lowest quartile. Machine learning analysis further confirmed the robustness of BRI in predicting infertility risk, with the XGBoost model providing the highest area under the curve (AUC = 0.935). These findings highlight BRI as a superior predictor for female infertility compared to traditional obesity measures, suggesting its potential for improving clinical risk stratification in reproductive health.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Evaluating the predictive power of the body roundness index (BRI) versus traditional obesity measures for female infertility: a hybrid approach using regression models and machine learning algorithms on cross-sectional data

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月10日

查看全文
  • Seyed Sobhan Bahreiny,
  • Mohammad-Navid Bastani,
  • Razieh Kazemzadeh,
  • Reza Mohammadpour Fard,
  • Mojtaba Sarvestani

摘要

The Body Roundness Index (BRI) has recently been proposed as an effective metric for assessing obesity-related health risks, but its association with female infertility remains insufficiently studied. This study aims to evaluate the predictive power of BRI in comparison to traditional obesity measures (waist circumference (WC) and body mass index (BMI)) for predicting female infertility using data from the National Health and Nutrition Examination Survey (NHANES). A total of 2962 women aged 18–45 years were included in this cross-sectional analysis. Multivariable logistic regression models were employed to examine the relationship between BRI and infertility, adjusting for demographic, behavioral, and metabolic factors. In addition, Generalized Additive Models (GAM) and Restricted Cubic Spline (RCS) regression were used to investigate potential non-linear associations, and machine learning algorithms were applied to assess the predictive performance and identify significant features. Our results demonstrated a significant positive association between BRI and female infertility. In the fully adjusted model, each unit increase in BRI was associated with 18% higher odds of infertility (OR = 1.18, p = 0.032). A dose-response relationship was also observed across BRI quartiles, with women in the highest quartile (Q4) exhibiting 125% higher odds of infertility compared to those in the lowest quartile. Machine learning analysis further confirmed the robustness of BRI in predicting infertility risk, with the XGBoost model providing the highest area under the curve (AUC = 0.935). These findings highlight BRI as a superior predictor for female infertility compared to traditional obesity measures, suggesting its potential for improving clinical risk stratification in reproductive health.