Scholarlink

Background <p>Conventional epidemiological studies often rely on average associations for single social determinants, which may obscure heterogeneity across jointly defined social positions. This study applied an intersectionality-informed Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA) approach to examine disparities in multimorbidity prevalence and burden among Chinese adults.</p> Methods <p>We analyzed 25,409 participants from the 2023–2024 China National Health Survey. Intersectional strata were defined a priori by six social-position dimensions: age, sex, region, residence, educational attainment, and occupational status, yielding 127 observed strata. Multimorbidity was defined as having two or more of 14 chronic conditions, and burden was assessed using the Chinese Multimorbidity Weighted Index (CMWI). MAIHDA models were fitted separately for multimorbidity and CMWI to estimate stratum-specific predictions, variance partition coefficients (VPCs), proportional changes in variance (PCVs), and discriminatory accuracy. Sensitivity analyses examined age restriction, outcome definition, survey-site clustering, lifestyle adjustment, and birthplace-based residence classification.</p> Results <p>Overall, multimorbidity prevalence was 22.96%, and mean CMWI was 1.32. Predicted multimorbidity prevalence ranged from 3.36% to 53.37%, and predicted CMWI ranged from 0.19 to 2.95. Higher predicted values were concentrated among older adults, northern residents, participants with lower education, and those who were farmers, unemployed, or retired. Intersectional strata accounted for 16.92% of total variation in multimorbidity and 13.74% in CMWI; discriminatory accuracy for multimorbidity was moderate (AUC = 0.75). Age explained the largest share of between-stratum variance, followed by occupational status and region. Additive main effects explained most between-stratum heterogeneity (PCVs: 82.69% for multimorbidity and 88.37% for CMWI), although residual stratum-specific deviations persisted in a subset of strata. Sensitivity analyses supported the robustness of the findings.</p> Conclusions <p>Multimorbidity prevalence and burden were unevenly distributed across intersectionally defined social positions in China. Most heterogeneity reflected additive social-position gradients, particularly age, occupation, and region, with smaller residual deviations beyond additivity. These findings support proportionate universalism: universal chronic disease prevention and management should be strengthened, with additional support directed toward socially and regionally disadvantaged groups with greater burden.</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

Social determinants of intersectional inequalities in multimorbidity and its burden in China: a multilevel analysis of individual heterogeneity and discriminatory accuracy (MAIHDA)

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月21日

查看全文
  • Zhili Chen,
  • Yaoda Hu,
  • Huijing He,
  • Qiong Ou,
  • Yawen Liu,
  • Tan Xu,
  • Ji Tu,
  • Zhiming Lu,
  • Zichao Wang,
  • Da Xu,
  • Qing Wang,
  • Zhujie Ran,
  • Mengwei Zhang,
  • Guangliang Shan

摘要

Background

Conventional epidemiological studies often rely on average associations for single social determinants, which may obscure heterogeneity across jointly defined social positions. This study applied an intersectionality-informed Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA) approach to examine disparities in multimorbidity prevalence and burden among Chinese adults.

Methods

We analyzed 25,409 participants from the 2023–2024 China National Health Survey. Intersectional strata were defined a priori by six social-position dimensions: age, sex, region, residence, educational attainment, and occupational status, yielding 127 observed strata. Multimorbidity was defined as having two or more of 14 chronic conditions, and burden was assessed using the Chinese Multimorbidity Weighted Index (CMWI). MAIHDA models were fitted separately for multimorbidity and CMWI to estimate stratum-specific predictions, variance partition coefficients (VPCs), proportional changes in variance (PCVs), and discriminatory accuracy. Sensitivity analyses examined age restriction, outcome definition, survey-site clustering, lifestyle adjustment, and birthplace-based residence classification.

Results

Overall, multimorbidity prevalence was 22.96%, and mean CMWI was 1.32. Predicted multimorbidity prevalence ranged from 3.36% to 53.37%, and predicted CMWI ranged from 0.19 to 2.95. Higher predicted values were concentrated among older adults, northern residents, participants with lower education, and those who were farmers, unemployed, or retired. Intersectional strata accounted for 16.92% of total variation in multimorbidity and 13.74% in CMWI; discriminatory accuracy for multimorbidity was moderate (AUC = 0.75). Age explained the largest share of between-stratum variance, followed by occupational status and region. Additive main effects explained most between-stratum heterogeneity (PCVs: 82.69% for multimorbidity and 88.37% for CMWI), although residual stratum-specific deviations persisted in a subset of strata. Sensitivity analyses supported the robustness of the findings.

Conclusions

Multimorbidity prevalence and burden were unevenly distributed across intersectionally defined social positions in China. Most heterogeneity reflected additive social-position gradients, particularly age, occupation, and region, with smaller residual deviations beyond additivity. These findings support proportionate universalism: universal chronic disease prevention and management should be strengthened, with additional support directed toward socially and regionally disadvantaged groups with greater burden.