Orthopaedic surgery · 2026
Кардиометаболическая многомерность увеличивает риск разлома гипса: продольное коотертное исследование на основе CHARLS.
Название и аннотация: Машинный перевод · не проверен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Tian P, Du S, Tian R, Zhu W, Wang C, Wang K, Yang P, Huang X
Аннотация
В качестве сопутствующего явления Cardiometabolic Multimbidity (CMM) two или более заболеваний сердца, сахарного диабета, инсульта и гипертонии. Предыдущие исследования показали связь между кардиометаболическими заболеваниями и хрупкими трещинами; однако связь между ШМ и переломами бедра у китайского населения остается неясной. Таким образом, цель данного исследования заключается в изучении этой ассоциации в китайской когорте в целях информирования о стратегиях предотвращения переломов. 2011 в адрес 2020. Участники от 2011 Первоначально была включена группа базовых обследований. Впоследствии отдельные лица были последовательно исключены, если они находились под стражей 45 Возраст, неполная исходная информация по ШМ, имеет историю перелома бедра, потеряна для последующей деятельности или отсутствуют данные о смешениях. Для оценки взаимосвязи между ШМ и риском перелома бедра были проведены анализ выживаемости Каплан-Мейер, оценка пропорциональной опасности кокса, анализ подгрупп и анализ чувствительности. 6314 Возрастной состав участников 45 Включены годы и возраст, из которых 544 с ШМом. Больше а 9-year последующий период, 287 аварийные переломы бедренной части4.55%) были идентифицированы. Среди них: 36 У участников был диагностирован ШМ на исходном уровне, в то время как 251 Нет. Частота перелома бедра была значительно выше у участников с ШМ, чем у тех, кто не имеет ШМ (см.13% против. 8%, p = 0.015). После полной корректировки для смешателей многовариантная коксная регрессия показала, что ШМ ассоциируется с а 70% повышенный риск перелома бедра (HR = 1.70, 95% CI: 1.318-2.47; p = 0.005). Анализ, проведенный подгруппами, показал, что возраст и история падений являются существенными факторами, изменяющими последствия. Связь между ШМ и переломом бедра была более заметной среди участников в соответствии 60 лет (Р для взаимодействия = 0.048) и тех, кто имеет историю падений (P для взаимодействия = 0.014Эти выводы свидетельствуют о том, что ШМ повышает риск перелома бедра, особенно среди относительно молодых людей и лиц с историей падений.
Оригинал публикации
Cardiometabolic Multimorbidity Increases the Risk of Hip Fracture: A Longitudinal Cohort Study Based on CHARLS.
BackgroundCardiometabolic Multimorbidity (CMM) is defined as the co-occurrence of two or more conditions among heart disease, diabetes mellitus, stroke, and hypertension. Previous studies have shown associations between cardiometabolic diseases and fragility fractures; however, the relationship between CMM and hip fractures remains unclear in the Chinese population. This study therefore aims to investigate this association in a Chinese cohort to inform fracture prevention strategies.MethodsThis prospective cohort study used data from the China Health and Retirement Longitudinal Study (CHARLS) collected from 2011 to 2020. Participants from the 2011 baseline survey cohort were initially included. Subsequently, individuals were sequentially excluded if they were under 45 years of age, had incomplete baseline CMM information, had a history of hip fracture, lost to follow-up, or had missing data on confounders. Kaplan-Meier survival analysis, Cox proportional hazards regression, subgroup analyses, and sensitivity analyses were performed to evaluate the association between CMM and the risk of hip fracture.ResultsA total of 6314 participants aged 45 years and older were included, of whom 544 had CMM. Over a 9-year follow-up period, 287 incident hip fractures (4.55%) were identified. Among these, 36 participants had been diagnosed with CMM at baseline, whereas 251 had not. The incidence of hip fracture was significantly higher in participants with CMM than in those without CMM (13% vs. 8%, p = 0.015). After full adjustment for confounders, multivariable Cox regression showed that CMM was associated with a 70% increased risk of hip fracture (HR = 1.70, 95% CI: 1.318-2.47; p = 0.005). Subgroup analyses indicated that age and history of falls were significant effect modifiers. The association between CMM and hip fracture was more pronounced in participants under 60 years old (P for interaction = 0.048) and those with a history of falls (P for interaction = 0.014).ConclusionThese findings suggest that CMM increases the risk of hip fracture, particularly among relatively younger individuals and those with a history of falls.
Перевод полного текста: Перевод в очереди
Читать найденный полный текст
https://creativecommons.org/licenses/by-nc/4.0/ This is an open access article under the terms of the http://creativecommons.org/licenses/by-nc/4.0/ License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.
Abbreviations
body mass index
China Health and Retirement Longitudinal Study
confidence intervals
cardiometabolic diseases
cardiometabolic Multimorbidity
C‐reactive protein
cardiovascular Disease
Global Burden of Disease Study
hazard ratio
standard deviation
Strengthening the Reporting of Observational Studies in Epidemiology
total cholesterol
triglyceride
Introduction
With the acceleration of global population aging and marked changes in lifestyle, multimorbidity of chronic diseases has become a major challenge for public health. Among these conditions, Cardiometabolic Multimorbidity (CMM) has attracted considerable attention from the academic community due to its high prevalence. CMM is defined as the coexistence of two or more cardiometabolic diseases (CMDs) in a single individual, including hypertension, diabetes mellitus, coronary heart disease, and stroke. This definition is based on the landmark study by Di Angelantonio et al. [ 1 ], which analyzed data from 689,300 participants and confirmed a multiplicative effect of cardiometabolic multimorbidity. Compared with individuals without these diseases, the hazard ratio (HR) for all‐cause mortality among patients with any two coexisting conditions ranged from 3.5 to 3.8, substantially higher than the simple sum of the risks associated with a single disease. The HR increased to 6.9 among those with three concurrent diseases. These results underscore the clinical relevance and pathophysiological plausibility of using “coexistence of at least two diseases” as the diagnostic threshold for CMM. In essence, CMM reflects a state of synergistic dysregulation of cardiovascular and metabolic functions, rather than a mere accumulation of separate diseases. Numerous large‐scale population cohort studies have revealed a comorbid relationship between Cardiovascular Disease (CVD) and metabolic diseases, with metabolic abnormalities increasing the risk of developing CVD [ 2 , 3 ]. In the United States, the prevalence of CMM is 14.4% among individuals aged 45 years and older and exceeds 30% in those aged 65 years and above [ 4 ]. In China, the 7‐year cumulative incidence of CMM among middle‐aged and older adults is as high as 22.01%, with the coexistence of hypertension and diabetes mellitus being the most common combination [ 5 ]. Notably, the health hazards of CMM far exceed those of a single CVD. Compared with older adults with one CVD, patients with CMM have a 1.89‐fold higher risk of all‐cause mortality and a 12‐year shorter life expectancy [ 6 , 7 ]. Owing to the high medical expenditures and impaired quality of life associated with it, CMM has become one of the key drivers aggravating the global burden of disease.
Among fragility fractures, hip fracture has emerged as one of the most concerning skeletal health crises in the context of global aging due to its high rates of disability and mortality, as well as its devastating impact on older adults' health [ 8 ]. Hip fracture refers specifically to non‐traumatic fractures of the proximal femur, including the femoral neck, intertrochanteric, and subtrochanteric regions [ 9 ]. According to authoritative estimates from the Global Burden of Disease Study (GBD), approximately 1.83 million new hip fracture cases were reported worldwide in 2019, and this number is projected to reach around 6 million by 2050, with individuals aged 65 years and older accounting for more than 80% of all cases [ 10 ]. In China, the annual incidence of hip fracture among people aged 60 years and older is 1.2‰, and the incidence is significantly higher in females than in males [ 11 ]. Even more concerning, the prognosis of hip fracture is poor: approximately 20%–30% of patients die from complications such as pulmonary infection and deep vein thrombosis within 1 year after the fracture [ 12 ]. Unfortunately, fewer than 50% of patients can regain their pre‐fracture functional status, while 30%–40% face prolonged bed confinement and loss of independent living ability [ 12 ].
Previous studies have indicated a close association between hypertension and orthopedic disorders. A study by Du et al. [ 13 ] demonstrated that a history of hypertension is linked to an increased risk of future fractures. In addition, the prevalence of osteoporosis in patients with hypertension is significantly higher than that in non‐hypertensive individuals [ 14 ]. Multiple studies have shown a positive correlation between diabetes mellitus and fractures. A meta‐analysis involving 11,358 patients revealed that type 2 diabetes mellitus increases the risk of hip fractures by 34% [ 15 ]. With respect to CVD, a cross‐sectional study indicated that bone fractures may be associated with an elevated risk of myocardial infarction during long‐term follow‐up among older adults, with this association being more pronounced in individuals aged over 50 years [ 16 ]. Stroke is also a risk factor for fractures; a prospective cohort study showed that, following a stroke, trabecular bone thickness and density decrease, along with reduced bone strength [ 17 ].
Although the above evidence has clarified the association between individual CMDs and fracture risk, these studies are limited to single‐disease analyses and cannot fully capture real‐world clinical scenarios. In routine clinical practice, middle‐aged and older adults frequently experience multiple coexisting CMDs. Notably, multimorbidity is not merely a simple accumulation of individual diseases; rather, it exerts a distinct synergistic amplification effect, and its overall adverse health consequences far exceed the sum of the risks attributable to each single condition [ 18 ]. At present, few domestic or international studies have treated CMM as an independent exposure factor. Systematic evidence on the comprehensive effects of multimorbidity on hip fracture remains scarce, and high‐quality prospective cohort data investigating the underlying synergistic mechanisms are still lacking. Examining only a single chronic disease fails to explain the additional fracture risk conferred by multiple coexisting conditions and cannot meet clinical needs for stratified prevention and management of high‐risk populations. Therefore, based on a nationally representative prospective cohort, this study focuses on hip fracture, a highly fatal and disabling outcome, to systematically examine the association between CMM and hip fracture risk. This study seeks to address the limitations of previous research and provide reliable evidence to support comprehensive clinical prevention and intervention strategies for hip fracture.
Methods
Study Population
The China Health and Retirement Longitudinal Study (CHARLS) is a nationally representative cohort study in China, designed to collect comprehensive health‐related data from residents aged 45 years and older in mainland China. During the survey, CHARLS staff members extensively collected information on participants' demographic characteristics, socioeconomic status, health‐related behaviors, and a range of acute and chronic conditions. First launched in 2011, CHARLS used a stratified, multistage sampling approach to survey populations across 28 provinces in China, covering 150 counties and including 17,708 individuals from more than 10,000 households [ 19 ].
The CHARLS study was approved by the Biomedical Ethics Review Committee of Peking University (approval number IRB00001052‐11015), and the entire research process was conducted in strict accordance with the Declaration of Helsinki. All research procedures involving human participants were implemented in compliance with the 2013 version of the Declaration of Helsinki and the relevant national and institutional ethical guidelines, ensuring that the study met the highest ethical standards. In addition, written informed consent was obtained from all participants in this study, guaranteeing that they voluntarily participated after fully understanding the research content. Meanwhile, the preparation of this study report strictly followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [ 20 ].
Participants from the 2011 baseline survey were enrolled in this study. The following groups were sequentially excluded: (1) 407 participants younger than 45 years; (2) 412 participants with undiagnosable CMM at baseline; (3) 1528 participants with a history of hip fracture at baseline; (4) 5678 participants lost to follow‐up; and (5) 3369 participants with missing data on confounding variables (Figure 1 ).
Selection process of research subjects. CMM, Cardiometabolic Multimorbidity.
Diagnosis of Hip Fracture
During subsequent follow‐up interviews, researchers asked participants the question, “Have you experienced a hip fracture since the last interview?” using a self‐report approach. To ensure the accuracy of the reported information, interviewers provided participants with a detailed explanation of the specific location of the hip joint, and hip fracture status was ultimately recorded as a binary outcome of “yes” or “no”. To further improve data reliability, participants were required to complete two self‐assessments of their health status during the interview: the first at the beginning of this section and the second at the end of this section, with the questions in both self‐assessments presented in a random order [ 21 ].
Definition of CMM
CMM refers to the concurrent presence of two or more diseases related to the cardiovascular and metabolic domains, specifically including hypertension, diabetes mellitus, heart disease, and stroke [ 22 ]. To determine whether an individual had any of the aforementioned conditions, researchers used the following questioning methods: “Has a doctor ever told you that you have hypertension?” “Have you been diagnosed with diabetes mellitus, or do you have elevated blood glucose levels (including impaired glucose tolerance and elevated fasting blood glucose)?” “Do you have heart disease (e.g., myocardial infarction, coronary heart disease, angina pectoris, congestive heart failure, or other types of heart disease)?” and “Have you ever been diagnosed with stroke (including cerebral infarction and intracerebral hemorrhage) by a doctor?”
Covariates
A structured questionnaire was used to collect baseline sociodemographic characteristics and health‐related information on participants, with the aim of thoroughly characterizing their sociodemographic profiles and health status. Sociodemographic variables included sex, age, education level, marital status, and residence. Lifestyle factors comprised smoking and alcohol consumption. Health‐related variables included body mass index (BMI), kidney disease, dyslipidemia, emotional disorders, history of falls, triglyceride (TG) levels, total cholesterol (TC) levels, and C‐reactive protein (CRP) levels.
Statistical Analysis
Participants were categorized into two groups according to whether they developed a new‐onset hip fracture. For comparisons of continuous variables, the t ‐test was used, with results reported as mean and standard deviation (SD). For categorical variables, the chi‐square test was applied, with results presented as number of cases (n) and percentages (%). Kaplan–Meier survival analysis and the log‐rank test were used to compare cumulative risks between groups. CMM was entered as a binary variable in a multivariable Cox regression model to estimate the hazard ratio (HR) and 95% confidence intervals (CIs) for the association between CMM and hip fracture. Covariates were added in four sequential steps in the Cox model to examine the independent association between CMM and incident hip fracture. Model 0 was the unadjusted crude model. Model 1 adjusted for age, gender, educational level, place of residence, and marital status. Model 2 further adjusted for smoking, alcohol consumption, and BMI based on Model 1. Model 3 additionally adjusted for kidney disease, dyslipidemia, emotional problems, falls, TG, TC, and CRP. Subgroup analyses and interaction tests were conducted to assess potential effect modification by covariates. Specifically, multiplicative interaction terms between each covariate and CMM were constructed and included in the model, and the likelihood ratio test was used to compare models with interaction terms against the original model.
To evaluate the robustness of the findings, sensitivity analyses were conducted. First, hip fracture cases occurring within 2 years of recruitment were excluded. Second, multiple imputation was applied to missing covariates. Third, propensity score matching was performed at a 1:3 ratio based on the presence or absence of hip fracture. Nearest‐neighbor matching was used, with a caliper set to 0.02.
A two‐sided p ‐value of less than 0.05 was considered statistically significant. All statistical analyses were performed using R version 4.4.4 ( http://www.r‐project.org ).
Results
Population Characteristics
After applying inclusion and exclusion criteria, 6314 participants aged 45 years and older were enrolled. Over a 9‐year follow‐up period, 287 new hip fracture cases (4.55%) were identified. The incidence of hip fracture was significantly higher among participants with CMM than among those without CMM (13% vs. 8%, p = 0.015). Baseline prevalence of CMM was significantly higher in the new‐onset hip fracture group than in the control group ( p = 0.02); however, there was no significant difference in the baseline CMM subgroup combinations between the new‐onset hip fracture group and the control group (all p > 0.05). In addition, participants in the new‐onset hip fracture group were more likely to be older and illiterate, single and residing in rural areas, to be underweight with kidney disease, and to have a history of falls (all p < 0.05). By contrast, there were no significant differences between the two groups in terms of sex, smoking and alcohol consumption history, blood lipid levels, emotional disorders, or CRP levels (all p > 0.05) (Table 1 ).
TABLE 1 Characteristics stratified by hip fracture in this study. Variables Total ( n = 6314) Control ( n = 6027) Hip fracture ( n = 287)
p
CMM, n (%) 0.02 No 5770 (91) 5519 (92) 251 (87) Yes 544 (9) 508 (8) 36 (13) Heart disease + stroke, n (%) 0.401 No 6282 (99) 5995 (99) 287 (100) Yes 32 (1) 32 (1) 0 (0) Heart disease + hypertension, n (%) 0.317 No 5968 (95) 5701 (95) 267 (93) Yes 346 (5) 326 (5) 20 (7) Heart disease + diabetes mellitus, n (%) 0.164 No 6237 (99) 5956 (99) 281 (98) Yes 77 (1) 71 (1) 6 (2) Stroke + hypertension, n (%) 1 No 6238 (99) 5954 (99) 284 (99) Yes 76 (1) 73 (1) 3 (1) Stroke + diabetes mellitus, n (%) 0.479 No 6300 (100) 6014 (100) 286 (100) Yes 14 (0) 13 (0) 1 (0) Hypertension + diabetes mellitus, n (%) 0.472 No 6147 (97) 5870 (97) 277 (97) Yes 167 (3) 157 (3) 10 (3) Heart disease + stroke + hypertension, n (%) 0.625 No 6290 (100) 6003 (100) 287 (100) Yes 24 (0) 24 (0) 0 (0) Heart disease + stroke + diabetes mellitus, n (%) 1 No 6309 (100) 6022 (100) 287 (100) Yes 5 (0) 5 (0) 0 (0) Stroke + hypertension + diabetes mellitus, n (%) 1 No 6304 (100) 6017 (100) 287 (100) Yes 10 (0) 10 (0) 0 (0) Heart disease + hypertension + diabetes mellitus, n (%) 0.293 No 6263 (99) 5980 (99) 283 (99) Yes 51 (1) 47 (1) 4 (1) Heart disease + stroke + hypertension + diabetes mellitus, n (%) 1 No 6310 (100) 6023 (100) 287 (100) Yes 4 (0) 4 (0) 0 (0) Sex, n (%) 0.436 Female 3522 (56) 3355 (56) 167 (58) Male 2792 (44) 2672 (44) 120 (42) Age (years), (mean (SD)) 58.29 (8.47) 58.13 (8.38) 61.51 (9.77) < 0.001 Education, n (%) < 0.001 High school or above 616 (10) 595 (10) 21 (7) Illiterate 1792 (28) 1676 (28) 116 (40) Junior high school or below 3906 (62) 3756 (62) 150 (52) Marital status, n (%) < 0.001 Married 5712 (90) 5475 (91) 237 (83) Single 602 (10) 552 (9) 50 (17) Residence place, n (%) 0.001 Rural 4302 (68) 4081 (68) 221 (77) Urban 2012 (32) 1946 (32) 66 (23) Smoke, n (%) 0.31 No 4469 (71) 4274 (71) 195 (68) Yes 1845 (29) 1753 (29) 92 (32) Drink, n (%) 0.519 No 3778 (60) 3612 (60) 166 (58) Yes 2536 (40) 2415 (40) 121 (42) BMI (kg/m 2 ), n (%) 0.001 18.5~24 3295 (52) 3131 (52) 164 (57) < 18.5 371 (6) 343 (6) 28 (10) 24~28 1881 (30) 1818 (30) 63 (22) > 28 767 (12) 735 (12) 32 (11) Kidney disease, n (%) 0.039 No 5986 (95) 5722 (95) 264 (92) Yes 328 (5) 305 (5) 23 (8) Dyslipidemia, n (%) 0.161 No 5738 (91) 5470 (91) 268 (93) Yes 576 (9) 557 (9) 19 (7) Emotional problems, n (%) 0.351 No 6254 (99) 5971 (99) 283 (99) Yes 60 (1) 56 (1) 4 (1) Falls, n (%) < 0.001 No 5305 (84) 5090 (84) 215 (75) Yes 1009 (16) 937 (16) 72 (25) TG (mg/dL), (mean (SD)) 133.73 (111.89) 134.27 (113.46) 122.50 (70.67) 0.082 TC (mmol/L), (mean (SD)) 5.03 (1.00) 5.03 (1.00) 4.97 (0.98) 0.361 CRP (mg/L), (mean (SD)) 2.43 (6.58) 2.43 (6.62) 2.58 (5.74) 0.69 Abbreviations: BMI, body mass index; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive Protein; SD, standard deviation; TC, Total Cholesterol; TG, Triglyceride.
Characteristics stratified by hip fracture in this study.
Abbreviations: BMI, body mass index; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive Protein; SD, standard deviation; TC, Total Cholesterol; TG, Triglyceride.
Longitudinal Association Between CMM and New‐Onset Hip
During the 9‐year follow‐up period, 287 new hip fracture cases (4.55%) were identified. Among them, 36 participants were diagnosed with hip fracture who had CMM at baseline, and 251 had no CMM at baseline. Kaplan–Meier survival analysis showed that the cumulative incidence of hip fracture was significantly higher in the CMM group than in the control group (log‐rank p = 0.015) (Figure 2 ).
The Kaplan–Meier curves of cumulative risk for hip fractures between participants With CMM and Without CMM at baseline. CMM, Cardiometabolic Multimorbidity.
Table 2 presents the Cox regression results assessing the association between CMM and new‐onset hip fracture. During the 9‐year follow‐up period, in the unadjusted Model 0, compared with the non‐CMM group, CMM was associated with a 54% higher incidence of hip fracture (HR = 1.54, 95% CI: 1.09–2.18; p = 0.02). In Model 1, adjusted for demographic factors, the incidence of hip fracture was 48% higher in the CMM group (HR = 1.48, 95% CI: 1.04–2.11; p = 0.03). In Model 2, which further adjusted for smoking, alcohol consumption, and BMI, CMM was associated with a 60% higher incidence of hip fracture (HR = 1.60, 95% CI: 1.12–2.29; p = 0.01). After full adjustment for demographic factors, lifestyle factors, and health‐related factors, CMM remained an independent risk factor, with a 70% increased risk of hip fracture (HR = 1.70, 95% CI: 1.18–2.47; p = 0.005).
TABLE 2 The results of Cox regression on the association between CMM and hip fracture.
Hip fracture~CMM
Crude model Model 1 Model 2 Model 3 HR (95% CI)
p
HR (95% CI)
p
HR (95% CI)
p
HR (95% CI)
p
Without CMM ref ref ref ref With CMM 1.54 (1.09, 2.18) 0.02 1.48 (1.04, 2.11) 0.03 1.60 (1.12, 2.29) 0.01 1.70 (1.18, 2.47) 0.005
Note: Model 0: No covariate was adjusted. Model 1: Adjusted for sex, age, education attainment, Marital status, and residence place. Model 2: Further adjusted for smoking status, drinking status, and BMI based on Model 1. Model 3: Further adjusted for kidney disease, dyslipidemia, emotional problems, falls, TG, TC, and CRP on Model 2. Abbreviations: BMI, body mass index; CI, confidence interval; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive protein; HR, hazard ratio; ref., reference; TC, total cholesterol; TG, triglyceride.
The results of Cox regression on the association between CMM and hip fracture.
Hip fracture~CMM
Note: Model 0: No covariate was adjusted.
Model 1: Adjusted for sex, age, education attainment, Marital status, and residence place.
Model 2: Further adjusted for smoking status, drinking status, and BMI based on Model 1.
Model 3: Further adjusted for kidney disease, dyslipidemia, emotional problems, falls, TG, TC, and CRP on Model 2.
Abbreviations: BMI, body mass index; CI, confidence interval; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive protein; HR, hazard ratio; ref., reference; TC, total cholesterol; TG, triglyceride.
Subgroup Analysis
Figure 3 shows the results of the subgroup analyses. Our findings suggested that age and history of falls were significant effect modifiers. Among participants aged under 60 years, CMM was associated with a 162.6% increased incidence of hip fracture (HR = 2.626, 95% CI: 1.539–4.483; p < 0.001). In contrast, among participants aged 60 years and older, no significant association was observed between CMM and hip fracture ( p = 0.327), and the difference between subgroups was statistically significant ( p for interaction = 0.048). In the subgroup with a history of falls, CMM was associated with a 229.2% higher incidence of hip fracture (HR = 3.292, 95% CI: 1.807–5.997; p < 0.0001). However, among participants without a history of falls, no significant association was found between CMM and hip fracture ( p = 0.374), and the between‐subgroup difference was statistically significant ( p for interaction = 0.014).
The subgroup analysis of association between CMM and hip fracture. CMM, Cardiometabolic Multimorbidity.
Sensitivity Analysis
To enhance the credibility of the findings, three sensitivity analyses were performed. As shown in Table 3 , first, we excluded new‐onset fracture cases occurring within 2 years to reduce the likelihood of reverse causation; the results indicated that in the fully adjusted Model 3, the association between CMM and hip fracture remained statistically significant (HR = 1.65, 95% CI: 1.10–2.48; p = 0.01). Second, to reduce potential bias arising from missing covariate data, multiple imputation was performed; after full adjustment, the association between CMM and hip fracture remained robust (HR = 1.52, 95% CI: 1.11–2.07; p = 0.01). Third, after 1:3 propensity score matching between participants with hip fracture and those without hip fracture, baseline characteristics showed no significant differences (Table S1 ). Cox regression results further confirmed a significant positive association between CMM and hip fracture (HR = 1.53, 95% CI: 1.08–2.18; p = 0.02).
TABLE 3 Sensitivity analysis. Hip fracture~CMM Crude model Model 1 Model 2 Model 3 HR (95% CI)
p
HR (95% CI)
p
HR (95% CI)
p
HR (95% CI)
p
Sensitivity 1 Without CMM ref ref ref ref With CMM 1.48 (1.01, 2.17) 0.04 1.44 (0.98, 2.11) 0.06 1.57 (1.06, 2.32) 0.02 1.65 (1.10, 2.48) 0.01 Sensitivity 2 Without CMM ref ref ref ref With CMM 1.41 (1.05, 1.89) 0.02 1.39 (1.03, 1.87) 0.03 1.47 (1.09, 1.99) 0.01 1.52 (1.11, 2.07) 0.01 Sensitivity 3 Without CMM ref With CMM 1.53 (1.08, 2.18) 0.02
Note: Sensitivity 1: New‐onset hip fractures within 2 years were excluded. Sensitivity 2: Multiple imputation for missing variables. Sensitivity 3: A 1:3 propensity score matching was conducted between participants with hip fracture and without hip fracture. Model 0: No covariate was adjusted. Model 1: Adjusted for sex, age, education attainment, Marital status, and residence place. Model 2: Further adjusted for smoking status, drinking status, and BMI based on Model 1. Model 3: Further adjusted for kidney disease, dyslipidemia, emotional problems, falls, TG, TC, and CRP on Model 2. Abbreviations: BMI, body mass index; CI, confidence interval; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive protein; HR, hazard ratio; ref., reference; TC, total cholesterol; TG, triglyceride.
Sensitivity analysis.
Note: Sensitivity 1: New‐onset hip fractures within 2 years were excluded.
Sensitivity 2: Multiple imputation for missing variables.
Sensitivity 3: A 1:3 propensity score matching was conducted between participants with hip fracture and without hip fracture.
Model 0: No covariate was adjusted.
Model 1: Adjusted for sex, age, education attainment, Marital status, and residence place.
Model 2: Further adjusted for smoking status, drinking status, and BMI based on Model 1.
Model 3: Further adjusted for kidney disease, dyslipidemia, emotional problems, falls, TG, TC, and CRP on Model 2.
Abbreviations: BMI, body mass index; CI, confidence interval; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive protein; HR, hazard ratio; ref., reference; TC, total cholesterol; TG, triglyceride.
Discussion
Main Findings of This Study
This study investigated the association between hip fracture and CMM among 6314 middle‐aged and older adults aged 45 years and older. The results showed that CMM was an independent risk factor for hip fracture, with participants with CMM experiencing a 70% higher incidence of hip fracture. Moreover, the association between CMM and hip fracture was stronger among relatively younger individuals and among those with a history of falls. These findings suggest that controlling the number of chronic cardiovascular and metabolic diseases may reduce the synergistic effect of multimorbidity on the risk of hip fracture.
Comparison With Previous Studies
This study offers compelling evidence that CMM increases the risk of hip fracture. Although research on the association between CMM and hip fracture has remained limited to date, multiple studies examining the links between individual CMM components and fractures have indirectly corroborated the robustness of the conclusions drawn here. In a case–control study, Gerber et al. [ 23 ] showed that ischemic heart disease was significantly associated with a higher risk of hip fracture, with this relationship being especially pronounced in elderly women. Wu et al. [ 24 ] further broadened the scope of the association between CVD and hip fracture. Their findings indicated that heart failure increased the risk of hip fracture by 2.89‐fold, while hip fracture also increased the risk of heart failure by 2.74‐fold, suggesting a bidirectional relationship between heart disease and fracture. As a core component of CMM, diabetes mellitus is likewise strongly supported by extensive evidence linking it to hip fracture. The 2025 clinical practice guideline issued by the American Diabetes Association explicitly states that patients with type 2 diabetes mellitus have a 1.79‐fold higher risk of hip fracture even when bone mineral density is normal, and that their lifetime fracture risk is 40%–70% greater than that of individuals without diabetes [ 25 ]. This clearly indicates that the adverse effects of diabetes mellitus on skeletal health are not mediated solely by reduced bone mineral density; they may also involve deterioration in bone quality, such as disruption of bone microstructure and abnormal bone mineralization. Using large‐scale UK Biobank data, Curtis et al. [ 26 ] found a positive association between the duration of diabetes mellitus and fracture risk, with patients who had type 2 diabetes for more than 10 years experiencing a substantially higher risk of fracture. Stroke, another major CVD event related to CMM, has also been confirmed to be associated with an increased hip fracture risk. A Korean cohort study reported that stroke patients had an overall fracture risk 40% higher than the general population, with a 2.42‐fold increase in hip fracture risk [ 27 ]. However, these studies have largely focused on single disease components. In contrast, this study is the first to directly demonstrate that the co‐occurrence of these conditions in the form of multimorbidity produces a synergistic and amplifying effect, resulting in a substantial increase in hip fracture risk. These findings therefore provide more comprehensive and direct evidence to inform the prevention and control of fracture risk in the context of multimorbidity.
Age and Fall History Are Potential Effect Modifiers
In the subgroup analyses, we identified age and a history of falls as significant effect modifiers. Prior epidemiological studies have consistently shown a strong relationship between age and fracture risk. A Turkish study involving more than 3.28 million fracture patients found that the overall fracture incidence among individuals aged 20–64 years was 0.94%, rising progressively with age; the average incidence peaked at 1.16% in the 55–64 age group, which was approximately 30% higher than in the 20–34 age group [ 28 ]. A retrospective study in Japan reported that the incidence of hip, distal radius, and proximal humerus fractures all increased with age, with the age‐related rise in risk being more pronounced in women [ 29 ]. In the present study, the association between CMM and hip fracture was stronger among individuals under 60 years of age. This finding likely reflects the early detrimental impact of CMM on skeletal health. Although relatively younger individuals would theoretically be in a stage of comparatively stable bone strength, the presence of CMM may trigger premature processes of bone loss and deterioration of bone microstructure. In addition, such changes are often overlooked in the absence of overt symptoms, leading to an accumulation of high hip fracture risk in this population before they reach advanced old age.
This study also found that among patients with CMM, those with a history of falls had a significantly higher risk of hip fracture. CMM increases hip fracture risk through multiple systemic pathophysiological pathways. On the one hand, conditions such as diabetes mellitus and hypertension can directly reduce bone mineral density and disrupt bone microstructure, making the hips inherently more fragile [ 25 , 30 ]. On the other hand, complications including neuropathy, cardiac insufficiency, and vascular sclerosis—induced by heart disease, stroke, and diabetes—impair balance, exercise tolerance, and postural control, thereby indirectly increasing the likelihood of falls [ 31 , 32 ]. A history of falls is a direct risk factor for hip fracture, as individuals with such a history commonly exhibit underlying vulnerabilities, including impaired balance and insufficient muscle strength [ 33 ]. When individuals with a history of falls are also diagnosed with CMM, CMM further exacerbates skeletal fragility and physical control impairment, producing a synergistic effect with the history of falls. Consequently, the association between CMM and hip fracture is stronger in the subgroup of patients with a history of falls.
Potential Mechanisms Underlying the Association Between CMM and Hip Fracture
From a pathophysiological perspective, first, sarcopenia—common among patients with CMM—reduces the buffering capacity of the lower‐limb muscles against falls and increases the risk of hip fracture [ 34 ]. Second, the association between CMM and hip fracture may be mediated by the synergistic effects of multiple pathways, including metabolic disorders, disruption of the bone microenvironment, and functional decline. Chronic hyperinsulinemia can impair bone formation via two mechanisms. On the one hand, insulin can competitively bind to insulin‐like growth factor‐1 receptors on the surface of osteoblasts, thereby suppressing osteoblast proliferation and collagen synthesis [ 35 ]. On the other hand, hyperglycemia promotes the accumulation of advanced glycation end products, which disrupt the cross‐linking structure of bone collagen fibers, reduce bone elasticity and shear resistance, and make the proximal femur more vulnerable to fracture after low‐energy trauma [ 36 ]. In addition, the chronic low‐grade inflammatory state associated with CMM further disturbs the balance between bone formation and bone resorption. Inflammatory factors can directly stimulate osteoclast proliferation and inhibit the Wnt/β‐catenin signaling pathway, thereby reducing osteoblast activity [ 37 ]. Finally, dyslipidemia—often comorbid with CMM—can impair vascular endothelial function, decrease skeletal blood supply, and further aggravate osteocyte apoptosis [ 38 ].
Clinical and Public Health Implications of the Present Findings
Based on the findings of this study, we propose three recommendations. First, in clinical diagnosis and management, bone mineral density testing should be included in the routine assessment of patients with CMM, and timely anti‐osteoporotic therapy should be initiated for those with osteopenia or osteoporosis. Second, regular screening and intervention should be implemented for fall‐related risk factors associated with CMM. For example, in patients with orthostatic hypotension, adjusting the dosage of antihypertensive drugs can help avoid excessive blood pressure fluctuations, and in patients with peripheral neuropathy, supplementing vitamins can improve neurological function. Third, at the public health level, health education regarding the co‐management of CMM and hip fracture should be strengthened. For CMM patients with a history of falls, home environment modifications and fall emergency training should be provided to reduce the risk of hip fracture.
Advantages and Limitations
This study has several strengths. First, for the first time, based on a large‐scale community cohort of the Chinese population, it clarifies the independent risk effect of CMM on hip fracture, thereby addressing the gap created by insufficient attention to middle‐aged and older Chinese adults. Second, the study conducted several sensitivity analyses, producing relatively robust results. Third, subgroup analyses were also performed, identifying two high‐risk subgroups—relatively younger individuals and those with a history of falls—thus offering a basis for population‐specific precision interventions. Fourth, this study comprehensively accounted for various potential confounders, and the results obtained were relatively robust.
However, we acknowledge the following limitations. First, the participants were middle‐aged and older adults aged 45 years and above; therefore, the conclusions cannot be generalized to young and middle‐aged adults. In addition, the limited representation of adults aged 85 years and older may lead to an underestimation of the effect of CMM on hip fracture risk. Second, the diagnosis of CMM and hip fracture relied on self‐reports, without corroboration from clinical imaging or laboratory confirmation, which may have introduced misclassification bias. Third, although both urban and rural samples were included, coverage of underdeveloped western regions and remote rural areas was limited due to constraints related to sampling weights and sample size. Consequently, the conclusions may be more applicable to the central and eastern regions, and caution is warranted when extrapolating them to rural areas in western China. Fourth, the number of positive cases for specific CMM disease combinations was too small to construct a stable statistical model. Meanwhile, strict stratification by the number of comorbidities (two, three, and four conditions) would yield insufficient sample sizes within each subgroup, which limited further mechanistic investigation and clinical interpretation. Fifth, only 36 hip fracture events were documented among participants with CMM. Accordingly, subgroup analyses and interaction tests stratified by age, fall history, and other factors were constrained by limited statistical power. These findings should be considered exploratory and should not be overinterpreted or broadly generalized. Further large‐scale cohort studies with longer follow‐up are needed to validate these subgroup effects and enrich the relevant clinical evidence. Sixth, owing to data availability in the public CHARLS database, potential confounders such as glucocorticoids and anti‐osteoporosis medications could not be adjusted. The use of these medications may independently affect hip fracture risk and is also associated with the presence of CMM. Inadequate adjustment for these confounders may result in residual confounding bias and potentially distort the estimated association between CMM and hip fracture risk. Finally, as a prospective cohort study, this research cannot completely rule out reverse causation.
Conclusion
This study confirms that CMM is positively associated with hip fracture risk in middle‐aged and older Chinese adults, and that this association is stronger among relatively younger middle‐aged and older adults as well as among individuals with a history of falls.
Author Contributions
Peng Tian: conceptualization, writing – original draft, writing – review and editing, formal analysis, visualization, methodology, software. Weidong Zhu: conceptualization, investigation, writing – original draft, supervision. Shaoxiong Du: conceptualization, writing – original draft, writing – review and editing, formal analysis. Pei Yang: funding acquisition, project administration. Chunsheng Wang: supervision, validation, funding acquisition. Run Tian: formal analysis, validation, methodology, conceptualization. KunZheng Wang: funding acquisition, validation, supervision, project administration. Xin Huang: funding acquisition, project administration, data curation, resources.
Funding
The authors have nothing to report.
Ethics Statement
All waves of the CHARLS were reviewed and approved by the Institutional Review Board (IRB) of Peking University. Specifically, the IRB approval number assigned to the core household survey (with anthropometric measurements included) is IRB00001052‐11015, while that for the biomarker collection component is IRB00001052‐11014. All experimental procedures were carried out in strict compliance with the principles stipulated in the Declaration of Helsinki.
Consent
Written informed consent was secured from each participant prior to their enrollment in the surveys.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: Baseline characteristics after propensity score matching.
body mass index
China Health and Retirement Longitudinal Study
confidence intervals
cardiometabolic diseases
cardiometabolic Multimorbidity
C‐reactive protein
cardiovascular Disease
Global Burden of Disease Study
hazard ratio
standard deviation
Strengthening the Reporting of Observational Studies in Epidemiology
total cholesterol
triglyceride
Introduction
With the acceleration of global population aging and marked changes in lifestyle, multimorbidity of chronic diseases has become a major challenge for public health. Among these conditions, Cardiometabolic Multimorbidity (CMM) has attracted considerable attention from the academic community due to its high prevalence. CMM is defined as the coexistence of two or more cardiometabolic diseases (CMDs) in a single individual, including hypertension, diabetes mellitus, coronary heart disease, and stroke. This definition is based on the landmark study by Di Angelantonio et al. [ 1 ], which analyzed data from 689,300 participants and confirmed a multiplicative effect of cardiometabolic multimorbidity. Compared with individuals without these diseases, the hazard ratio (HR) for all‐cause mortality among patients with any two coexisting conditions ranged from 3.5 to 3.8, substantially higher than the simple sum of the risks associated with a single disease. The HR increased to 6.9 among those with three concurrent diseases. These results underscore the clinical relevance and pathophysiological plausibility of using “coexistence of at least two diseases” as the diagnostic threshold for CMM. In essence, CMM reflects a state of synergistic dysregulation of cardiovascular and metabolic functions, rather than a mere accumulation of separate diseases. Numerous large‐scale population cohort studies have revealed a comorbid relationship between Cardiovascular Disease (CVD) and metabolic diseases, with metabolic abnormalities increasing the risk of developing CVD [ 2 , 3 ]. In the United States, the prevalence of CMM is 14.4% among individuals aged 45 years and older and exceeds 30% in those aged 65 years and above [ 4 ]. In China, the 7‐year cumulative incidence of CMM among middle‐aged and older adults is as high as 22.01%, with the coexistence of hypertension and diabetes mellitus being the most common combination [ 5 ]. Notably, the health hazards of CMM far exceed those of a single CVD. Compared with older adults with one CVD, patients with CMM have a 1.89‐fold higher risk of all‐cause mortality and a 12‐year shorter life expectancy [ 6 , 7 ]. Owing to the high medical expenditures and impaired quality of life associated with it, CMM has become one of the key drivers aggravating the global burden of disease.
Among fragility fractures, hip fracture has emerged as one of the most concerning skeletal health crises in the context of global aging due to its high rates of disability and mortality, as well as its devastating impact on older adults' health [ 8 ]. Hip fracture refers specifically to non‐traumatic fractures of the proximal femur, including the femoral neck, intertrochanteric, and subtrochanteric regions [ 9 ]. According to authoritative estimates from the Global Burden of Disease Study (GBD), approximately 1.83 million new hip fracture cases were reported worldwide in 2019, and this number is projected to reach around 6 million by 2050, with individuals aged 65 years and older accounting for more than 80% of all cases [ 10 ]. In China, the annual incidence of hip fracture among people aged 60 years and older is 1.2‰, and the incidence is significantly higher in females than in males [ 11 ]. Even more concerning, the prognosis of hip fracture is poor: approximately 20%–30% of patients die from complications such as pulmonary infection and deep vein thrombosis within 1 year after the fracture [ 12 ]. Unfortunately, fewer than 50% of patients can regain their pre‐fracture functional status, while 30%–40% face prolonged bed confinement and loss of independent living ability [ 12 ].
Previous studies have indicated a close association between hypertension and orthopedic disorders. A study by Du et al. [ 13 ] demonstrated that a history of hypertension is linked to an increased risk of future fractures. In addition, the prevalence of osteoporosis in patients with hypertension is significantly higher than that in non‐hypertensive individuals [ 14 ]. Multiple studies have shown a positive correlation between diabetes mellitus and fractures. A meta‐analysis involving 11,358 patients revealed that type 2 diabetes mellitus increases the risk of hip fractures by 34% [ 15 ]. With respect to CVD, a cross‐sectional study indicated that bone fractures may be associated with an elevated risk of myocardial infarction during long‐term follow‐up among older adults, with this association being more pronounced in individuals aged over 50 years [ 16 ]. Stroke is also a risk factor for fractures; a prospective cohort study showed that, following a stroke, trabecular bone thickness and density decrease, along with reduced bone strength [ 17 ].
Although the above evidence has clarified the association between individual CMDs and fracture risk, these studies are limited to single‐disease analyses and cannot fully capture real‐world clinical scenarios. In routine clinical practice, middle‐aged and older adults frequently experience multiple coexisting CMDs. Notably, multimorbidity is not merely a simple accumulation of individual diseases; rather, it exerts a distinct synergistic amplification effect, and its overall adverse health consequences far exceed the sum of the risks attributable to each single condition [ 18 ]. At present, few domestic or international studies have treated CMM as an independent exposure factor. Systematic evidence on the comprehensive effects of multimorbidity on hip fracture remains scarce, and high‐quality prospective cohort data investigating the underlying synergistic mechanisms are still lacking. Examining only a single chronic disease fails to explain the additional fracture risk conferred by multiple coexisting conditions and cannot meet clinical needs for stratified prevention and management of high‐risk populations. Therefore, based on a nationally representative prospective cohort, this study focuses on hip fracture, a highly fatal and disabling outcome, to systematically examine the association between CMM and hip fracture risk. This study seeks to address the limitations of previous research and provide reliable evidence to support comprehensive clinical prevention and intervention strategies for hip fracture.
Methods
Study Population
The China Health and Retirement Longitudinal Study (CHARLS) is a nationally representative cohort study in China, designed to collect comprehensive health‐related data from residents aged 45 years and older in mainland China. During the survey, CHARLS staff members extensively collected information on participants' demographic characteristics, socioeconomic status, health‐related behaviors, and a range of acute and chronic conditions. First launched in 2011, CHARLS used a stratified, multistage sampling approach to survey populations across 28 provinces in China, covering 150 counties and including 17,708 individuals from more than 10,000 households [ 19 ].
The CHARLS study was approved by the Biomedical Ethics Review Committee of Peking University (approval number IRB00001052‐11015), and the entire research process was conducted in strict accordance with the Declaration of Helsinki. All research procedures involving human participants were implemented in compliance with the 2013 version of the Declaration of Helsinki and the relevant national and institutional ethical guidelines, ensuring that the study met the highest ethical standards. In addition, written informed consent was obtained from all participants in this study, guaranteeing that they voluntarily participated after fully understanding the research content. Meanwhile, the preparation of this study report strictly followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [ 20 ].
Participants from the 2011 baseline survey were enrolled in this study. The following groups were sequentially excluded: (1) 407 participants younger than 45 years; (2) 412 participants with undiagnosable CMM at baseline; (3) 1528 participants with a history of hip fracture at baseline; (4) 5678 participants lost to follow‐up; and (5) 3369 participants with missing data on confounding variables (Figure 1 ).
Selection process of research subjects. CMM, Cardiometabolic Multimorbidity.
Diagnosis of Hip Fracture
During subsequent follow‐up interviews, researchers asked participants the question, “Have you experienced a hip fracture since the last interview?” using a self‐report approach. To ensure the accuracy of the reported information, interviewers provided participants with a detailed explanation of the specific location of the hip joint, and hip fracture status was ultimately recorded as a binary outcome of “yes” or “no”. To further improve data reliability, participants were required to complete two self‐assessments of their health status during the interview: the first at the beginning of this section and the second at the end of this section, with the questions in both self‐assessments presented in a random order [ 21 ].
Definition of CMM
CMM refers to the concurrent presence of two or more diseases related to the cardiovascular and metabolic domains, specifically including hypertension, diabetes mellitus, heart disease, and stroke [ 22 ]. To determine whether an individual had any of the aforementioned conditions, researchers used the following questioning methods: “Has a doctor ever told you that you have hypertension?” “Have you been diagnosed with diabetes mellitus, or do you have elevated blood glucose levels (including impaired glucose tolerance and elevated fasting blood glucose)?” “Do you have heart disease (e.g., myocardial infarction, coronary heart disease, angina pectoris, congestive heart failure, or other types of heart disease)?” and “Have you ever been diagnosed with stroke (including cerebral infarction and intracerebral hemorrhage) by a doctor?”
Covariates
A structured questionnaire was used to collect baseline sociodemographic characteristics and health‐related information on participants, with the aim of thoroughly characterizing their sociodemographic profiles and health status. Sociodemographic variables included sex, age, education level, marital status, and residence. Lifestyle factors comprised smoking and alcohol consumption. Health‐related variables included body mass index (BMI), kidney disease, dyslipidemia, emotional disorders, history of falls, triglyceride (TG) levels, total cholesterol (TC) levels, and C‐reactive protein (CRP) levels.
Statistical Analysis
Participants were categorized into two groups according to whether they developed a new‐onset hip fracture. For comparisons of continuous variables, the t ‐test was used, with results reported as mean and standard deviation (SD). For categorical variables, the chi‐square test was applied, with results presented as number of cases (n) and percentages (%). Kaplan–Meier survival analysis and the log‐rank test were used to compare cumulative risks between groups. CMM was entered as a binary variable in a multivariable Cox regression model to estimate the hazard ratio (HR) and 95% confidence intervals (CIs) for the association between CMM and hip fracture. Covariates were added in four sequential steps in the Cox model to examine the independent association between CMM and incident hip fracture. Model 0 was the unadjusted crude model. Model 1 adjusted for age, gender, educational level, place of residence, and marital status. Model 2 further adjusted for smoking, alcohol consumption, and BMI based on Model 1. Model 3 additionally adjusted for kidney disease, dyslipidemia, emotional problems, falls, TG, TC, and CRP. Subgroup analyses and interaction tests were conducted to assess potential effect modification by covariates. Specifically, multiplicative interaction terms between each covariate and CMM were constructed and included in the model, and the likelihood ratio test was used to compare models with interaction terms against the original model.
To evaluate the robustness of the findings, sensitivity analyses were conducted. First, hip fracture cases occurring within 2 years of recruitment were excluded. Second, multiple imputation was applied to missing covariates. Third, propensity score matching was performed at a 1:3 ratio based on the presence or absence of hip fracture. Nearest‐neighbor matching was used, with a caliper set to 0.02.
A two‐sided p ‐value of less than 0.05 was considered statistically significant. All statistical analyses were performed using R version 4.4.4 ( http://www.r‐project.org ).
Results
Population Characteristics
After applying inclusion and exclusion criteria, 6314 participants aged 45 years and older were enrolled. Over a 9‐year follow‐up period, 287 new hip fracture cases (4.55%) were identified. The incidence of hip fracture was significantly higher among participants with CMM than among those without CMM (13% vs. 8%, p = 0.015). Baseline prevalence of CMM was significantly higher in the new‐onset hip fracture group than in the control group ( p = 0.02); however, there was no significant difference in the baseline CMM subgroup combinations between the new‐onset hip fracture group and the control group (all p > 0.05). In addition, participants in the new‐onset hip fracture group were more likely to be older and illiterate, single and residing in rural areas, to be underweight with kidney disease, and to have a history of falls (all p < 0.05). By contrast, there were no significant differences between the two groups in terms of sex, smoking and alcohol consumption history, blood lipid levels, emotional disorders, or CRP levels (all p > 0.05) (Table 1 ).
TABLE 1 Characteristics stratified by hip fracture in this study. Variables Total ( n = 6314) Control ( n = 6027) Hip fracture ( n = 287)
p
CMM, n (%) 0.02 No 5770 (91) 5519 (92) 251 (87) Yes 544 (9) 508 (8) 36 (13) Heart disease + stroke, n (%) 0.401 No 6282 (99) 5995 (99) 287 (100) Yes 32 (1) 32 (1) 0 (0) Heart disease + hypertension, n (%) 0.317 No 5968 (95) 5701 (95) 267 (93) Yes 346 (5) 326 (5) 20 (7) Heart disease + diabetes mellitus, n (%) 0.164 No 6237 (99) 5956 (99) 281 (98) Yes 77 (1) 71 (1) 6 (2) Stroke + hypertension, n (%) 1 No 6238 (99) 5954 (99) 284 (99) Yes 76 (1) 73 (1) 3 (1) Stroke + diabetes mellitus, n (%) 0.479 No 6300 (100) 6014 (100) 286 (100) Yes 14 (0) 13 (0) 1 (0) Hypertension + diabetes mellitus, n (%) 0.472 No 6147 (97) 5870 (97) 277 (97) Yes 167 (3) 157 (3) 10 (3) Heart disease + stroke + hypertension, n (%) 0.625 No 6290 (100) 6003 (100) 287 (100) Yes 24 (0) 24 (0) 0 (0) Heart disease + stroke + diabetes mellitus, n (%) 1 No 6309 (100) 6022 (100) 287 (100) Yes 5 (0) 5 (0) 0 (0) Stroke + hypertension + diabetes mellitus, n (%) 1 No 6304 (100) 6017 (100) 287 (100) Yes 10 (0) 10 (0) 0 (0) Heart disease + hypertension + diabetes mellitus, n (%) 0.293 No 6263 (99) 5980 (99) 283 (99) Yes 51 (1) 47 (1) 4 (1) Heart disease + stroke + hypertension + diabetes mellitus, n (%) 1 No 6310 (100) 6023 (100) 287 (100) Yes 4 (0) 4 (0) 0 (0) Sex, n (%) 0.436 Female 3522 (56) 3355 (56) 167 (58) Male 2792 (44) 2672 (44) 120 (42) Age (years), (mean (SD)) 58.29 (8.47) 58.13 (8.38) 61.51 (9.77) < 0.001 Education, n (%) < 0.001 High school or above 616 (10) 595 (10) 21 (7) Illiterate 1792 (28) 1676 (28) 116 (40) Junior high school or below 3906 (62) 3756 (62) 150 (52) Marital status, n (%) < 0.001 Married 5712 (90) 5475 (91) 237 (83) Single 602 (10) 552 (9) 50 (17) Residence place, n (%) 0.001 Rural 4302 (68) 4081 (68) 221 (77) Urban 2012 (32) 1946 (32) 66 (23) Smoke, n (%) 0.31 No 4469 (71) 4274 (71) 195 (68) Yes 1845 (29) 1753 (29) 92 (32) Drink, n (%) 0.519 No 3778 (60) 3612 (60) 166 (58) Yes 2536 (40) 2415 (40) 121 (42) BMI (kg/m 2 ), n (%) 0.001 18.5~24 3295 (52) 3131 (52) 164 (57) < 18.5 371 (6) 343 (6) 28 (10) 24~28 1881 (30) 1818 (30) 63 (22) > 28 767 (12) 735 (12) 32 (11) Kidney disease, n (%) 0.039 No 5986 (95) 5722 (95) 264 (92) Yes 328 (5) 305 (5) 23 (8) Dyslipidemia, n (%) 0.161 No 5738 (91) 5470 (91) 268 (93) Yes 576 (9) 557 (9) 19 (7) Emotional problems, n (%) 0.351 No 6254 (99) 5971 (99) 283 (99) Yes 60 (1) 56 (1) 4 (1) Falls, n (%) < 0.001 No 5305 (84) 5090 (84) 215 (75) Yes 1009 (16) 937 (16) 72 (25) TG (mg/dL), (mean (SD)) 133.73 (111.89) 134.27 (113.46) 122.50 (70.67) 0.082 TC (mmol/L), (mean (SD)) 5.03 (1.00) 5.03 (1.00) 4.97 (0.98) 0.361 CRP (mg/L), (mean (SD)) 2.43 (6.58) 2.43 (6.62) 2.58 (5.74) 0.69 Abbreviations: BMI, body mass index; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive Protein; SD, standard deviation; TC, Total Cholesterol; TG, Triglyceride.
Characteristics stratified by hip fracture in this study.
Abbreviations: BMI, body mass index; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive Protein; SD, standard deviation; TC, Total Cholesterol; TG, Triglyceride.
Longitudinal Association Between CMM and New‐Onset Hip
During the 9‐year follow‐up period, 287 new hip fracture cases (4.55%) were identified. Among them, 36 participants were diagnosed with hip fracture who had CMM at baseline, and 251 had no CMM at baseline. Kaplan–Meier survival analysis showed that the cumulative incidence of hip fracture was significantly higher in the CMM group than in the control group (log‐rank p = 0.015) (Figure 2 ).
The Kaplan–Meier curves of cumulative risk for hip fractures between participants With CMM and Without CMM at baseline. CMM, Cardiometabolic Multimorbidity.
Table 2 presents the Cox regression results assessing the association between CMM and new‐onset hip fracture. During the 9‐year follow‐up period, in the unadjusted Model 0, compared with the non‐CMM group, CMM was associated with a 54% higher incidence of hip fracture (HR = 1.54, 95% CI: 1.09–2.18; p = 0.02). In Model 1, adjusted for demographic factors, the incidence of hip fracture was 48% higher in the CMM group (HR = 1.48, 95% CI: 1.04–2.11; p = 0.03). In Model 2, which further adjusted for smoking, alcohol consumption, and BMI, CMM was associated with a 60% higher incidence of hip fracture (HR = 1.60, 95% CI: 1.12–2.29; p = 0.01). After full adjustment for demographic factors, lifestyle factors, and health‐related factors, CMM remained an independent risk factor, with a 70% increased risk of hip fracture (HR = 1.70, 95% CI: 1.18–2.47; p = 0.005).
TABLE 2 The results of Cox regression on the association between CMM and hip fracture.
Hip fracture~CMM
Crude model Model 1 Model 2 Model 3 HR (95% CI)
p
HR (95% CI)
p
HR (95% CI)
p
HR (95% CI)
p
Without CMM ref ref ref ref With CMM 1.54 (1.09, 2.18) 0.02 1.48 (1.04, 2.11) 0.03 1.60 (1.12, 2.29) 0.01 1.70 (1.18, 2.47) 0.005
Note: Model 0: No covariate was adjusted. Model 1: Adjusted for sex, age, education attainment, Marital status, and residence place. Model 2: Further adjusted for smoking status, drinking status, and BMI based on Model 1. Model 3: Further adjusted for kidney disease, dyslipidemia, emotional problems, falls, TG, TC, and CRP on Model 2. Abbreviations: BMI, body mass index; CI, confidence interval; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive protein; HR, hazard ratio; ref., reference; TC, total cholesterol; TG, triglyceride.
The results of Cox regression on the association between CMM and hip fracture.
Hip fracture~CMM
Note: Model 0: No covariate was adjusted.
Model 1: Adjusted for sex, age, education attainment, Marital status, and residence place.
Model 2: Further adjusted for smoking status, drinking status, and BMI based on Model 1.
Model 3: Further adjusted for kidney disease, dyslipidemia, emotional problems, falls, TG, TC, and CRP on Model 2.
Abbreviations: BMI, body mass index; CI, confidence interval; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive protein; HR, hazard ratio; ref., reference; TC, total cholesterol; TG, triglyceride.
Subgroup Analysis
Figure 3 shows the results of the subgroup analyses. Our findings suggested that age and history of falls were significant effect modifiers. Among participants aged under 60 years, CMM was associated with a 162.6% increased incidence of hip fracture (HR = 2.626, 95% CI: 1.539–4.483; p < 0.001). In contrast, among participants aged 60 years and older, no significant association was observed between CMM and hip fracture ( p = 0.327), and the difference between subgroups was statistically significant ( p for interaction = 0.048). In the subgroup with a history of falls, CMM was associated with a 229.2% higher incidence of hip fracture (HR = 3.292, 95% CI: 1.807–5.997; p < 0.0001). However, among participants without a history of falls, no significant association was found between CMM and hip fracture ( p = 0.374), and the between‐subgroup difference was statistically significant ( p for interaction = 0.014).
The subgroup analysis of association between CMM and hip fracture. CMM, Cardiometabolic Multimorbidity.
Sensitivity Analysis
To enhance the credibility of the findings, three sensitivity analyses were performed. As shown in Table 3 , first, we excluded new‐onset fracture cases occurring within 2 years to reduce the likelihood of reverse causation; the results indicated that in the fully adjusted Model 3, the association between CMM and hip fracture remained statistically significant (HR = 1.65, 95% CI: 1.10–2.48; p = 0.01). Second, to reduce potential bias arising from missing covariate data, multiple imputation was performed; after full adjustment, the association between CMM and hip fracture remained robust (HR = 1.52, 95% CI: 1.11–2.07; p = 0.01). Third, after 1:3 propensity score matching between participants with hip fracture and those without hip fracture, baseline characteristics showed no significant differences (Table S1 ). Cox regression results further confirmed a significant positive association between CMM and hip fracture (HR = 1.53, 95% CI: 1.08–2.18; p = 0.02).
TABLE 3 Sensitivity analysis. Hip fracture~CMM Crude model Model 1 Model 2 Model 3 HR (95% CI)
p
HR (95% CI)
p
HR (95% CI)
p
HR (95% CI)
p
Sensitivity 1 Without CMM ref ref ref ref With CMM 1.48 (1.01, 2.17) 0.04 1.44 (0.98, 2.11) 0.06 1.57 (1.06, 2.32) 0.02 1.65 (1.10, 2.48) 0.01 Sensitivity 2 Without CMM ref ref ref ref With CMM 1.41 (1.05, 1.89) 0.02 1.39 (1.03, 1.87) 0.03 1.47 (1.09, 1.99) 0.01 1.52 (1.11, 2.07) 0.01 Sensitivity 3 Without CMM ref With CMM 1.53 (1.08, 2.18) 0.02
Note: Sensitivity 1: New‐onset hip fractures within 2 years were excluded. Sensitivity 2: Multiple imputation for missing variables. Sensitivity 3: A 1:3 propensity score matching was conducted between participants with hip fracture and without hip fracture. Model 0: No covariate was adjusted. Model 1: Adjusted for sex, age, education attainment, Marital status, and residence place. Model 2: Further adjusted for smoking status, drinking status, and BMI based on Model 1. Model 3: Further adjusted for kidney disease, dyslipidemia, emotional problems, falls, TG, TC, and CRP on Model 2. Abbreviations: BMI, body mass index; CI, confidence interval; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive protein; HR, hazard ratio; ref., reference; TC, total cholesterol; TG, triglyceride.
Sensitivity analysis.
Note: Sensitivity 1: New‐onset hip fractures within 2 years were excluded.
Sensitivity 2: Multiple imputation for missing variables.
Sensitivity 3: A 1:3 propensity score matching was conducted between participants with hip fracture and without hip fracture.
Model 0: No covariate was adjusted.
Model 1: Adjusted for sex, age, education attainment, Marital status, and residence place.
Model 2: Further adjusted for smoking status, drinking status, and BMI based on Model 1.
Model 3: Further adjusted for kidney disease, dyslipidemia, emotional problems, falls, TG, TC, and CRP on Model 2.
Abbreviations: BMI, body mass index; CI, confidence interval; CMM, Cardiometabolic Multimorbidity; CRP, C‐reactive protein; HR, hazard ratio; ref., reference; TC, total cholesterol; TG, triglyceride.
Discussion
Main Findings of This Study
This study investigated the association between hip fracture and CMM among 6314 middle‐aged and older adults aged 45 years and older. The results showed that CMM was an independent risk factor for hip fracture, with participants with CMM experiencing a 70% higher incidence of hip fracture. Moreover, the association between CMM and hip fracture was stronger among relatively younger individuals and among those with a history of falls. These findings suggest that controlling the number of chronic cardiovascular and metabolic diseases may reduce the synergistic effect of multimorbidity on the risk of hip fracture.
Comparison With Previous Studies
This study offers compelling evidence that CMM increases the risk of hip fracture. Although research on the association between CMM and hip fracture has remained limited to date, multiple studies examining the links between individual CMM components and fractures have indirectly corroborated the robustness of the conclusions drawn here. In a case–control study, Gerber et al. [ 23 ] showed that ischemic heart disease was significantly associated with a higher risk of hip fracture, with this relationship being especially pronounced in elderly women. Wu et al. [ 24 ] further broadened the scope of the association between CVD and hip fracture. Their findings indicated that heart failure increased the risk of hip fracture by 2.89‐fold, while hip fracture also increased the risk of heart failure by 2.74‐fold, suggesting a bidirectional relationship between heart disease and fracture. As a core component of CMM, diabetes mellitus is likewise strongly supported by extensive evidence linking it to hip fracture. The 2025 clinical practice guideline issued by the American Diabetes Association explicitly states that patients with type 2 diabetes mellitus have a 1.79‐fold higher risk of hip fracture even when bone mineral density is normal, and that their lifetime fracture risk is 40%–70% greater than that of individuals without diabetes [ 25 ]. This clearly indicates that the adverse effects of diabetes mellitus on skeletal health are not mediated solely by reduced bone mineral density; they may also involve deterioration in bone quality, such as disruption of bone microstructure and abnormal bone mineralization. Using large‐scale UK Biobank data, Curtis et al. [ 26 ] found a positive association between the duration of diabetes mellitus and fracture risk, with patients who had type 2 diabetes for more than 10 years experiencing a substantially higher risk of fracture. Stroke, another major CVD event related to CMM, has also been confirmed to be associated with an increased hip fracture risk. A Korean cohort study reported that stroke patients had an overall fracture risk 40% higher than the general population, with a 2.42‐fold increase in hip fracture risk [ 27 ]. However, these studies have largely focused on single disease components. In contrast, this study is the first to directly demonstrate that the co‐occurrence of these conditions in the form of multimorbidity produces a synergistic and amplifying effect, resulting in a substantial increase in hip fracture risk. These findings therefore provide more comprehensive and direct evidence to inform the prevention and control of fracture risk in the context of multimorbidity.
Age and Fall History Are Potential Effect Modifiers
In the subgroup analyses, we identified age and a history of falls as significant effect modifiers. Prior epidemiological studies have consistently shown a strong relationship between age and fracture risk. A Turkish study involving more than 3.28 million fracture patients found that the overall fracture incidence among individuals aged 20–64 years was 0.94%, rising progressively with age; the average incidence peaked at 1.16% in the 55–64 age group, which was approximately 30% higher than in the 20–34 age group [ 28 ]. A retrospective study in Japan reported that the incidence of hip, distal radius, and proximal humerus fractures all increased with age, with the age‐related rise in risk being more pronounced in women [ 29 ]. In the present study, the association between CMM and hip fracture was stronger among individuals under 60 years of age. This finding likely reflects the early detrimental impact of CMM on skeletal health. Although relatively younger individuals would theoretically be in a stage of comparatively stable bone strength, the presence of CMM may trigger premature processes of bone loss and deterioration of bone microstructure. In addition, such changes are often overlooked in the absence of overt symptoms, leading to an accumulation of high hip fracture risk in this population before they reach advanced old age.
This study also found that among patients with CMM, those with a history of falls had a significantly higher risk of hip fracture. CMM increases hip fracture risk through multiple systemic pathophysiological pathways. On the one hand, conditions such as diabetes mellitus and hypertension can directly reduce bone mineral density and disrupt bone microstructure, making the hips inherently more fragile [ 25 , 30 ]. On the other hand, complications including neuropathy, cardiac insufficiency, and vascular sclerosis—induced by heart disease, stroke, and diabetes—impair balance, exercise tolerance, and postural control, thereby indirectly increasing the likelihood of falls [ 31 , 32 ]. A history of falls is a direct risk factor for hip fracture, as individuals with such a history commonly exhibit underlying vulnerabilities, including impaired balance and insufficient muscle strength [ 33 ]. When individuals with a history of falls are also diagnosed with CMM, CMM further exacerbates skeletal fragility and physical control impairment, producing a synergistic effect with the history of falls. Consequently, the association between CMM and hip fracture is stronger in the subgroup of patients with a history of falls.
Potential Mechanisms Underlying the Association Between CMM and Hip Fracture
From a pathophysiological perspective, first, sarcopenia—common among patients with CMM—reduces the buffering capacity of the lower‐limb muscles against falls and increases the risk of hip fracture [ 34 ]. Second, the association between CMM and hip fracture may be mediated by the synergistic effects of multiple pathways, including metabolic disorders, disruption of the bone microenvironment, and functional decline. Chronic hyperinsulinemia can impair bone formation via two mechanisms. On the one hand, insulin can competitively bind to insulin‐like growth factor‐1 receptors on the surface of osteoblasts, thereby suppressing osteoblast proliferation and collagen synthesis [ 35 ]. On the other hand, hyperglycemia promotes the accumulation of advanced glycation end products, which disrupt the cross‐linking structure of bone collagen fibers, reduce bone elasticity and shear resistance, and make the proximal femur more vulnerable to fracture after low‐energy trauma [ 36 ]. In addition, the chronic low‐grade inflammatory state associated with CMM further disturbs the balance between bone formation and bone resorption. Inflammatory factors can directly stimulate osteoclast proliferation and inhibit the Wnt/β‐catenin signaling pathway, thereby reducing osteoblast activity [ 37 ]. Finally, dyslipidemia—often comorbid with CMM—can impair vascular endothelial function, decrease skeletal blood supply, and further aggravate osteocyte apoptosis [ 38 ].
Clinical and Public Health Implications of the Present Findings
Based on the findings of this study, we propose three recommendations. First, in clinical diagnosis and management, bone mineral density testing should be included in the routine assessment of patients with CMM, and timely anti‐osteoporotic therapy should be initiated for those with osteopenia or osteoporosis. Second, regular screening and intervention should be implemented for fall‐related risk factors associated with CMM. For example, in patients with orthostatic hypotension, adjusting the dosage of antihypertensive drugs can help avoid excessive blood pressure fluctuations, and in patients with peripheral neuropathy, supplementing vitamins can improve neurological function. Third, at the public health level, health education regarding the co‐management of CMM and hip fracture should be strengthened. For CMM patients with a history of falls, home environment modifications and fall emergency training should be provided to reduce the risk of hip fracture.
Advantages and Limitations
This study has several strengths. First, for the first time, based on a large‐scale community cohort of the Chinese population, it clarifies the independent risk effect of CMM on hip fracture, thereby addressing the gap created by insufficient attention to middle‐aged and older Chinese adults. Second, the study conducted several sensitivity analyses, producing relatively robust results. Third, subgroup analyses were also performed, identifying two high‐risk subgroups—relatively younger individuals and those with a history of falls—thus offering a basis for population‐specific precision interventions. Fourth, this study comprehensively accounted for various potential confounders, and the results obtained were relatively robust.
However, we acknowledge the following limitations. First, the participants were middle‐aged and older adults aged 45 years and above; therefore, the conclusions cannot be generalized to young and middle‐aged adults. In addition, the limited representation of adults aged 85 years and older may lead to an underestimation of the effect of CMM on hip fracture risk. Second, the diagnosis of CMM and hip fracture relied on self‐reports, without corroboration from clinical imaging or laboratory confirmation, which may have introduced misclassification bias. Third, although both urban and rural samples were included, coverage of underdeveloped western regions and remote rural areas was limited due to constraints related to sampling weights and sample size. Consequently, the conclusions may be more applicable to the central and eastern regions, and caution is warranted when extrapolating them to rural areas in western China. Fourth, the number of positive cases for specific CMM disease combinations was too small to construct a stable statistical model. Meanwhile, strict stratification by the number of comorbidities (two, three, and four conditions) would yield insufficient sample sizes within each subgroup, which limited further mechanistic investigation and clinical interpretation. Fifth, only 36 hip fracture events were documented among participants with CMM. Accordingly, subgroup analyses and interaction tests stratified by age, fall history, and other factors were constrained by limited statistical power. These findings should be considered exploratory and should not be overinterpreted or broadly generalized. Further large‐scale cohort studies with longer follow‐up are needed to validate these subgroup effects and enrich the relevant clinical evidence. Sixth, owing to data availability in the public CHARLS database, potential confounders such as glucocorticoids and anti‐osteoporosis medications could not be adjusted. The use of these medications may independently affect hip fracture risk and is also associated with the presence of CMM. Inadequate adjustment for these confounders may result in residual confounding bias and potentially distort the estimated association between CMM and hip fracture risk. Finally, as a prospective cohort study, this research cannot completely rule out reverse causation.
Conclusion
This study confirms that CMM is positively associated with hip fracture risk in middle‐aged and older Chinese adults, and that this association is stronger among relatively younger middle‐aged and older adults as well as among individuals with a history of falls.
Author Contributions
Peng Tian: conceptualization, writing – original draft, writing – review and editing, formal analysis, visualization, methodology, software. Weidong Zhu: conceptualization, investigation, writing – original draft, supervision. Shaoxiong Du: conceptualization, writing – original draft, writing – review and editing, formal analysis. Pei Yang: funding acquisition, project administration. Chunsheng Wang: supervision, validation, funding acquisition. Run Tian: formal analysis, validation, methodology, conceptualization. KunZheng Wang: funding acquisition, validation, supervision, project administration. Xin Huang: funding acquisition, project administration, data curation, resources.
Funding
The authors have nothing to report.
Ethics Statement
All waves of the CHARLS were reviewed and approved by the Institutional Review Board (IRB) of Peking University. Specifically, the IRB approval number assigned to the core household survey (with anthropometric measurements included) is IRB00001052‐11015, while that for the biomarker collection component is IRB00001052‐11014. All experimental procedures were carried out in strict compliance with the principles stipulated in the Declaration of Helsinki.
Consent
Written informed consent was secured from each participant prior to their enrollment in the surveys.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: Baseline characteristics after propensity score matching.