Translate this page into:
Relationship between social determinants of health and hip fracture in the American population: a cross-sectional NHANES study
⁎Corresponding author: Bo Chen. m13581498454@163.com
-
Received: ,
Accepted: ,
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.
Abstract
Abstract
This article aims to investigate the link between social determinants of health (SDoH) and the occurrence of hip fractures, and to analyze the potential underlying mechanisms of influence.
This paper conducted a deep analysis among the general sample population from the National Health and Nutrition Examination Survey (NHANES) 1999–2018. Within each survey period, self-reported data covering eight domains of SDoH were collected, including employment status, poverty-to-income ratio (PIR), food security, educational level, health insurance coverage, type of health insurance, homeownership, and marital status. Concurrently, a history of hip fractures among participants was learned through self-reporting. Correlation analysis, Lasso regression analysis, and the Boruta algorithm were employed to explore the link between SDoH and hip fractures.
This study comprehensively analyzed 11,254 adult participants. The findings indicated a notable link between eight sub-items of SDoH and the risk of hip fracture. Specifically, the cumulative number of adverse SDoH was positively linked with the risk of hip fracture. Especially as 4-7 adverse factors were accumulated, the incidence of hip fractures climbed greatly.
This paper suggests a marked link between adverse SDoH and the occurrence of hip fractures, particularly when these adverse factors accumulate to a certain number (4–7). A profound probe into the interactions among these SDoH and their mechanisms of influence on fracture risk may offer crucial evidence for the prevention and intervention of hip fractures.
Keywords
National cross-sectional study
Hip fracture
Social determinants of health
Lasso regression analysis
Boruta algorithm
1 Background
Fractures have emerged as a global public health challenge, heavily burdening the global economy.1,2 Statistics indicate that in 2019, the global incidence of new fractures reached 178 million, with a 33.4 % increase since 1990. Furthermore, an increase in years lived with disability (YLDs) due to fractures has been reported in 25.8 million people, up 65.3 % from 1990.3 Fractures not only result in absences, decreased productivity, disability, diminished quality of life (QoL), and exorbitant healthcare costs, but also cause substantial burdens on individuals, families, societies, and healthcare systems 4–6. Among various types of fractures, hip fractures, due to their severity and the potential risk of secondary fractures, are considered a known factor contributing to the rise in morbidity and mortality.7,8 The treatment process for hip fractures typically involves extended hospital stays, complex treatment plans, and vast healthcare resources. A systematic review estimates that the total cost of hospitalization for hip fractures is 10,075 US dollars. Additionally, within the first 12 months following a hip fracture, the total costs for healthcare and social care are projected to be 43,669.7 US dollars.9 Numerous factors contribute to fractures, with falls being the primary cause,3 accounting for over 90 % of hip fractures.10 In the analysis of fall causes, close links between social determinants of health (SDoH) and falls have been evidenced by some studies.11
SDoH serve as indicators of health equity and are linked to health outcomes. SDoH encompass non-medical risk factors like income, education, employment, housing, food security, and access to affordable healthcare services.12 The Healthy People 2030 initiative aims to achieve health equity and boost well-being by addressing SDoH issues. SDoH are considered primary drivers of health inequalities, having outstanding implications for the occurrence, progression, and prognosis of diseases.13
Currently, research on links between adverse SDoH and the efficacy of treatment and complications following distal radius fracture surgery has made some progress. Some research also reports the associations of SDoH with the efficacy and complications following total knee replacement and hemiarthroplasty 14–17. However, the links between SDOH and the occurrence of hip fractures remain elusive. Therefore, this paper utilizes available open data from the National Health and Nutrition Examination Survey (NHANES) 1999–2018 to investigate the possible impact of cumulative SDoH on hip fractures within a representative population of the United States. This will back the development of targeted preventive strategies, thereby reducing the incidence of hip fractures and related health inequalities and bringing new perspectives for formulating global fracture prevention and public health policies.
2 Methods
2.1 Study design and population
NHANES is a national, ongoing cross-sectional study administered by the Centers for Disease Control and Prevention (CDC), specialized in collecting health, nutritional, and sociological data from the civilian population of the United States. The study employed a complex, stratified, and multistage probability sampling design to ensure that the sample represented the civilian population of the United States.18 Data were collected through various methods, comprising interviews, physical examinations, dietary surveys, and laboratory tests, with data released biennially. The study design and methods of NHANES are detailed on its official website (https://wwwn.cdc.gov/nchs/nhanes/Default.aspx). Furthermore, the data collection process of NHANES was approved by the Institutional Review Board at the National Center for Health Statistics. All participants have signed written informed consent forms during the recruitment period.
In the initial phase of this study, data from ten periods of the NHANES 1999–2018 were collected and integrated, totaling 101,316 general participants. During the data processing, participants under the age of 20 (n = 46,235) and those with missing hip fracture data (n = 14,041) were excluded. Additionally, individuals with incomplete covariate data (n = 29,786), encompassing age, gender, race, body mass index (BMI), smoking status, drinking behavior, femoral BMD, and spinal BMD, were also excluded. After a series of stringent exclusions, 11,254 participants were included in this paper (Supplementary Fig. S1).
2.2 SDoH assessment
The definition of SDoH and their subitems was elaborated in detail under the Healthy People 2030 framework.13 This framework proposed a prospective approach to promote health and prevent disease. Based on the five domains delineated by Healthy People 2030, this paper established eight subitems of SDoH (Supplementary Table S1). The data for these subitems were sourced from the standard NHANES questionnaire 1999–2018 and were categorized into advantage and disadvantage levels. Specifically, these subitems covered economic stability (comprising employment status, poverty-to-income ratio (PIR), and food security), educational opportunities and quality (education level), health care opportunities and quality (insurance coverage and type of insurance), neighborhood and built environment (home ownership), and social and community environment (marital status).19 These components were categorized into advantage and disadvantage categories, like economic stability: unemployment (disadvantageous) versus high income (advantageous); education: low education level (disadvantageous) versus high education level (advantageous); health care: no insurance (disadvantageous) versus insurance coverage (advantageous). Each disadvantageous factor within a subitem was assigned a score of 1, while advantageous factors were scored as 0. The total score was computed by summing the scores of all subitems (0–8 points). 0 points indicated that all SDoH indicators were in an advantageous state (the individual was in a socially advantageous position with the lowest health risks). 1–8 points denoted the presence of disadvantageous SDOH indicators, with higher scores indicating a greater cumulative number of disadvantageous factors and a gradual increase in health risks.20
2.3 Hip fracture assessment
During the NHANES interview, a self-reported history of hip fracture was provided. Among study participants aged 20–85 years, those who responded affirmatively (yes) to the following question were categorized as hip fracture cases21: “Has a doctor ever told you that you had a hip fracture or a fracture?"
2.4 Covariates
By referring to previous literature 22–24 and the recommendations of clinical experts, this paper screened potential confounders that might influence links between SDoH and hip fractures. Collected data encompassed demographic characteristics, involving age (categorized into five groups: <18, 20–39, 40–59, 60–79, and ≥80 years), gender (male or female), race/nationality (Mexican American, Non-Hispanic White, Non-Hispanic Black, and Other), BMI, PIR, calcium, phosphorus, femoral BMD, and spinal BMD).
2.5 Statistical analysis
This study was based on the complex sampling survey data from NHANES, adhering to the NHANES analysis and reporting guidelines.25 The weighted analytical methods were employed to investigate links between SDoH scores (0–8) and hip fractures. Initially, descriptive statistics were presented to characterize the variables. Continuous variables were represented as weighted means ± standard deviations, and categorical variables were represented as weighted frequencies (percentages). Intergroup comparisons were made utilizing the chi-square test. Subsequently, three hierarchical logistic regression models were built. The raw model only analyzed the crude association of SDoH scores with hip fractures; the partially modified model incorporated demographic and socioeconomic factors (e.g., age, gender, and race); and the fully modified model further included lifestyle and clinical indicators (like smoking, alcohol consumption, and blood lipids). Additionally, Spearman's rank correlation analysis was employed to appraise linear relationships, Lasso regression to select variables, and the Brouta algorithm to appraise the importance of variables. All analyses were executed utilizing R4.1.0, with statistical importance defined as a two-tailed p < 0.05. Association strength between variables was visualized with chord diagrams.
2.6 Ethical compliance
This study utilized publicly available data from the National Health and Nutrition Examination Survey (NHANES) 1999–2018. The NHANES protocol was reviewed and approved by the Institutional Review Board (IRB) of the National Center for Health Statistics (NCHS), ensuring compliance with ethical standards for human subjects research. All participants provided written informed consent prior to data collection. As this study involved secondary analysis of deidentified, anonymized data, no additional ethical approval was required. Data handling adhered to NHANES confidentiality guidelines, and no personal identifiers were accessed or disclosed during the research process. The authors declare no conflicts of interest related to this work.
3 Results
3.1 Basic characteristics of study participants
This study enrolled 11,254 subjects, with an average age of 46.73±(15.21). The findings revealed a higher incidence of hip fractures among individuals who were female (55 %), non-Hispanic white (79 %), economically disadvantaged, and had a lower educational level. Moreover, smokers demonstrated a drastically elevated risk of hip fractures (p < 0.001). A prominent link between the eight dimensions of SDoH and the risk of hip fractures was reported. An evident difference was noticed between the hip-fracture group and the non-hip-fracture group in the distribution of SDoH scores (p < 0.001), suggesting that SDoH possibly played a role in the risk of hip fractures (Table 1).
| Characteristic | Na | Overall N = 48,405,483b | NO hipfracture N = 47,838,409b | hipfracture N = 567,074b | p-valuec |
| age | 11254 | 46.73± (15.21) | 46.61± (15.17) | 56.14± (15.58) | <0.001∗∗∗ |
| sex | 11254 | 0.344 | |||
| Female | 5572 (50 %) | 5500 (50 %) | 72 (55 %) | ||
| Male | 5682 (50 %) | 5612 (50 %) | 70 (45 %) | ||
| race | 11254 | 0.154 | |||
| Mexican American | 2035 (8.2 %) | 2019 (8.2 %) | 16 (5.1 %) | ||
| Non-Hispanic Black | 2110 (9.8 %) | 2085 (9.8 %) | 25 (9.3 %) | ||
| Non-Hispanic White | 5317 (71 %) | 5231 (71 %) | 86 (79 %) | ||
| Other | 1792 (11 %) | 1777 (11 %) | 15 (6.6 %) | ||
| marital | 11254 | 0.142 | |||
| Married | 6617 (62 %) | 6549 (62 %) | 68 (54 %) | ||
| Non-Married | 4637 (38 %) | 4563 (38 %) | 74 (46 %) | ||
| poverty | 11254 | 3.16± (1.63) | 3.16± (1.63) | 2.71± (1.69) | 0.033∗ |
| edu | 11254 | 0.014∗ | |||
| High school | 2605 (24 %) | 2567 (24 %) | 38 (32 %) | ||
| Less than high school | 2790 (16 %) | 2742 (16 %) | 48 (23 %) | ||
| More than high school | 5859 (60 %) | 5803 (60 %) | 56 (45 %) | ||
| bilirubin | 11254 | 12.54± (5.53) | 12.55± (5.54) | 11.71± (4.74) | 0.140 |
| alkaline_phosphatase | 11254 | 67.36± (22.38) | 67.28± (22.38) | 73.99± (21.60) | <0.001∗∗∗ |
| albumin | 11254 | 42.76± (3.14) | 42.77± (3.14) | 42.10± (2.94) | 0.018∗ |
| tc | 11254 | 5.12± (1.06) | 5.12± (1.06) | 4.97± (0.93) | 0.428 |
| ALBI | 11254 | −2.75± (0.27) | −2.75± (0.27) | −2.73± (0.25) | 0.150 |
| BMI | 11254 | 27.87± (5.62) | 27.87± (5.60) | 27.88± (6.88) | 0.712 |
| calcium | 11254 | 9.45± (0.36) | 9.45± (0.35) | 9.41± (0.38) | 0.361 |
| femur | 11254 | 0.97± (0.15) | 0.97± (0.15) | 0.86± (0.18) | <0.001∗∗∗ |
| phosphorus | 11254 | 3.76± (0.56) | 3.76± (0.56) | 3.81± (0.58) | 0.387 |
| smoke | 11254 | 0.004∗∗ | |||
| former | 2654 (24 %) | 2603 (23 %) | 51 (30 %) | ||
| never | 6079 (54 %) | 6028 (54 %) | 51 (37 %) | ||
| now | 2521 (22 %) | 2481 (22 %) | 40 (33 %) | ||
| alcohol.user | 11254 | 0.864 | |||
| former | 1883 (14 %) | 1857 (14 %) | 26 (12 %) | ||
| heavy | 2393 (22 %) | 2359 (22 %) | 34 (24 %) | ||
| mild | 3775 (37 %) | 3728 (37 %) | 47 (38 %) | ||
| moderate | 1776 (17 %) | 1760 (17 %) | 16 (15 %) | ||
| never | 1427 (9.9 %) | 1408 (9.9 %) | 19 (11 %) | ||
| spine | 11254 | 1.03± (0.15) | 1.04± (0.14) | 0.97± (0.15) | <0.001∗∗∗ |
| SDoH | 11254 | <0.001∗∗∗ | |||
| 0 | 1910 (25 %) | 1894 (25 %) | 16 (18 %) | ||
| 1 | 1994 (22 %) | 1969 (22 %) | 25 (20 %) | ||
| 2 | 1762 (16 %) | 1747 (16 %) | 15 (12 %) | ||
| 3 | 1627 (13 %) | 1606 (13 %) | 21 (14 %) | ||
| 4 | 1498 (10 %) | 1475 (10.0 %) | 23 (15 %) | ||
| 5 | 1314 (8.0 %) | 1295 (8.0 %) | 19 (8.2 %) | ||
| 6 | 795 (4.3 %) | 781 (4.3 %) | 14 (8.0 %) | ||
| 7 | 313 (1.7 %) | 305 (1.6 %) | 8 (3.6 %) | ||
| 8 | 41 (0.2 %) | 40 (0.2 %) | 1 (0.6 %) |
3.2 Association of SDoH with hip fractures
Table 2 elucidates the links between different levels of SDoH and the risk of hip fractures. This article employed a multivariate logistic regression model to deeply analyze the eight subitems of SDoH and their total sum. Results disclosed that individuals with a higher cumulative number of SDoH were exposed to an increased risk of hip fractures, particularly in the fully modified model. The analysis demonstrated a positive link between the increase in the cumulative number of SDoH and the rise in the risk of hip fractures. Concretely, when SDoH = 4, the odds ratio (OR) value in the fully modified model was 2.34 (95 % confidence interval (CI): 1.21–4.63, p = 0.012). When SDOH = 6, the OR value escalated to 3.34 (95 % CI: 1.55–7.13, p = 0.002). When SDOH = 7, the highest risk was observed, with an OR of 5.20 (95 % CI: 2.02–12.4, p < 0.001). However, when the SDOH = 8, the result was not statistically significant (p > 0.05), which might be attributed to the small sample size. This unraveled that as the cumulative number of adverse SDoH increased, especially 4–7 factors, the incidence of hip fractures drastically increased.
| No-adjusted | Partially adjusted | Fully adjusted | |||||||
| ORa | 95 % CIa | p-value | ORa | 95 % CIa | p-value | ORa | 95 % CIa | p-value | |
| SDoH | |||||||||
| 0 | 1 | – | 1 | – | 1 | – | |||
| 1 | 1.50 | 0.81, 2.88 | 0.2 | 1.40 | 0.75, 2.70 | 0.3 | 1.39 | 0.74, 2.68 | 0.3 |
| 2 | 1.02 | 0.50, 2.07 | >0.9 | 0.97 | 0.47, 1.98 | >0.9 | 0.95 | 0.46, 1.95 | 0.9 |
| 3 | 1.55 | 0.81, 3.02 | 0.2 | 1.73 | 0.89, 3.39 | 0.11 | 1.66 | 0.85, 3.28 | 0.14 |
| 4 | 1.85 | 0.98, 3.57 | 0.061 | 2.41 | 1.26, 4.70 | 0.008 | 2.34 | 1.21, 4.63 | 0.012 |
| 5 | 1.74 | 0.89, 3.43 | 0.11 | 2.60 | 1.31, 5.21 | 0.006 | 2.48 | 1.23, 5.03 | 0.011 |
| 6 | 2.12 | 1.02, 4.38 | 0.041 | 3.49 | 1.65, 7.33 | <0.001 | 3.34 | 1.55, 7.13 | 0.002 |
| 7 | 3.10 | 1.25, 7.12 | 0.010 | 5.55 | 2.19, 13.0 | <0.001 | 5.20 | 2.02, 12.4 | <0.001 |
| 8 | 2.96 | 0.16, 15.1 | 0.3 | 4.86 | 0.26, 25.2 | 0.13 | 3.74 | 0.20, 20.1 | 0.2 |
3.3 The association of SDoH with hip fracture
Firstly, this paper employed Spearman's rank correlation coefficient to analyze the link between SDoH score and the incidence of hip fractures, intending to reveal their linear correlation. Chord diagrams visually displayed the association strength between variables, thereby laying the foundation for subsequent in-depth analysis. Fig. 1 depicts an apparent link between SDoH and hip fractures. A strong positive link was observed between SDoH and poverty. Educational level was also closely linked to SDoH, further consolidating the links between low educational level and social disadvantage. Smoking behavior was strongly linked with both SDoH and hip fractures, suggesting its potential mediating effect on fracture risk. Drinking behavior was weakly linked with SDoH and bone health markers. Femoral and spinal BMI were negatively linked with hip fractures (Fig. 1).
3.4 Lasso regression analysis of SDoH and hip fracture
Following the establishment of links among variables, this article employed the Lasso regression approach for variable selection and normalization. Lasso regression effectively mitigated the issue of multicollinearity in high-dimensional data by reducing the coefficients of non-essential variables to zero, thereby optimizing the model structure. Fig. 2 illustrates the process of selecting predictors for hip fracture. SDoH as a notable predictor was incorporated into the model across multiple parameter values, indicating its substantial predictive power for hip fractures. Age was confirmed as the most reliable predictor. Besides, the femur and spinal BMI were retained, further reinforcing their central roles in bone strength and fracture susceptibility. For socioeconomic status and lifestyle factors, education level and poverty index were chosen as predictors, suggesting a link between lower socioeconomic status and a higher risk of fractures. The model retained the variable of smoking, confirming the adverse effects of smoking on bone health. The impact of drinking behavior was relatively weaker and was excluded at higher parameter values. For biochemical markers, alkaline phosphatase (ALP) and albumin (Alb) were identified as valuable predictors, indicating that metabolic processes contributed substantially to bone health. Calcium and phosphorus were excluded at higher absorption values, revealing a weaker link with fracture risk. For gender and race, these demographic factors were initially enrolled in the model but excluded at higher parameter values, suggesting that their influence may be indirectly mediated through other variables like SDoH and BMI (Fig. 2).
3.5 Analysis of the importance of SDoH-related factors utilizing the Boruta algorithm
This paper further employed the Brouta algorithm to quantitatively appraise the importance of various variables. Through iterative model training and variable exclusion strategies, this algorithm progressively established the importance ranking of each variable. With the aid of this robust assessment method, this paper could delineate the relative impact of different SDoH on the incidence rate of hip fractures, and analyze their predictive power on outcome variables. Age was the primary predictor for hip fracture risk, and a significantly increased risk was found in older individuals (Fig. 3). Among bone health indicators, femoral and spinal BMDs ranked relatively high, consistent with previous research findings on the risk of osteoporosis-related fractures. SDoH were confirmed as key predictors, highlighting the importance of socioeconomic factors in hip fracture risk. For biochemical markers, Alb and ALP were identified as influential factors, suggesting that metabolic processes might contribute to the susceptibility to fractures. For poverty and education level, higher poverty rates and lower education levels were tendentially linked with elevated fracture risks. For lifestyle factors, smoking was verified as a risk factor, while alcohol consumption, though tendentially linked with fracture risks, was insufficiently substantiated. Predictors confirmed by variable classification (green area) comprised age, SDoH, femoral and spinal BMI, Alb, and ALP. Suspicious factors (yellow area) included poverty, education, smoking, calcium, and phosphorus. Some demographic and biochemical markers displayed minimal influence and were categorized as uncertain factors (red area). (Fig. 3).
4 Discussion
This paper systematically discussed the cumulative impact of SDoH on the risk of hip fracture. It proved that the increase of adverse SDoH factors was significantly positively linked with the risk of hip fracture. After validation in a fully modified multivariate logistic regression model, we discovered that when the cumulative number of SDoH adverse factors reached 4–7, not only did it exert an independent and significant predictive effect on the risk of hip fracture, but the risk also rapidly rose. Further analysis indicated that the impact of SDoH was multidimensional. Firstly, as for socioeconomic factors, economic status, food insecurity, and educational level are closely associated with the risk of hip fracture. Additionally, SDoH indirectly influences fracture risks through lifestyle factors (e.g., smoking) and bone health indicators (e.g., femoral and spinal BMI). The study also discovered that SDoH was associated with biomarkers (like ALP and Alb levels), suggesting its potential role in bone metabolism. Key social determinants encompass economic status, food insecurity, educational level, healthcare services, and marital status, which reveal potential pathways of health inequity. Therefore, further investigation of the adverse factors of SDoH is vital for reducing the occurrence of hip fractures. Furthermore, the study uncovered that lower BMI values were associated with higher SDoH scores, implying that the bones of socially vulnerable populations may be more fragile, thus elevating fracture risks. In regard to biochemical markers, ALP is positively associated with hip fracture, suggesting its potential value as a biomarker of bone turnover. Alb levels are negatively linked with hip fracture, revealing that poor nutritional status may increase the risk of fractures. With respect to demographic factors, age is a primary factor closely linked with hip fracture. Gender and race exhibited a moderate correlation with hip fracture, with potential differences in fracture risk among different population subgroups.
5 Economic status and hip fracture risk
Economic status exhibited a marked influence on hip fracture in this article. The PIR is obviously lower in individuals with hip fractures than those without fractures, reflecting that poor economic conditions may lead to unfavorable conditions in terms of nutritional intake, living environment, and health management in individuals, thereby increasing the risk of fractures. Past research has proven that regions with lower socioeconomic status, flawed social welfare systems, higher unemployment rates, a larger proportion of low-income groups, and a higher proportion of single-parent families experience dominantly higher risks of hip fractures.26 In the United Kingdom, regional analyses based on the Townsend deprivation score indicate that regions with a higher proportion of socially impoverished populations have a higher risk of hip fractures, a trend closely linked with the decline in economic status.26 Similar research findings have been verified in Oslo, Norway. Kaastad et al.27 discovered that urban areas with worse socioeconomic conditions not only saw a higher incidence of hip fractures but also a dramatically increased mortality rate. A study in the United States disclosed that the risk of hip fractures displayed a distinct linear decreasing trend with the increase in family income.28 This phenomenon can be partly explained by the prevalent unhealthy lifestyles in socially impoverished areas. For instance, regions with a higher level of social poverty have a higher proportion of individuals engaging in unhealthy lifestyles like smoking, drinking, malnutrition, and insufficient physical activity, all of which are pivotal risk factors for hip fractures 27–29. Furthermore, the variances in regional economic status may also indirectly influence the risk of hip fractures through other pathways. For example, regions with high unemployment rates are often linked to a higher degree of industrialization, which may lead to an increased risk of hip fractures among workers due to increased occupational exposure. What's more, regions with high unemployment rates have a higher percentage of young individuals consuming excessive alcohol, further magnifying the risk of hip fractures related to traffic accidents. In low- and middle-income countries, the incidence of obesity and diabetes is surging, particularly in the older population, while both diseases are closely related to the risk of hip fractures. Sarcopenic obesity and diabetes strongly increase the risk of falls 30–36. With rapid aging in low- and middle-income countries, the incidence of cognitive impairment is also rising, further upping the risk of falls and hip fractures.37 Notably, prominent differences in the risk of hip fractures are recorded among different age spectrums. Hip fractures in young individuals, particularly males, are mainly caused by high-energy traumatic events like traffic accidents, occupational accidents, or sports injuries. In the older population, hip fractures are more commonly linked with osteoporosis, falls, and relatively minor trauma. In summary, the risk of hip fractures is closely related to socioeconomic status and exhibits diverse characteristics across different regions and populations. Regions with lower economic status not only face a higher risk of fractures but also have to deal with complex challenges posed by poverty, unemployment, lifestyle, and a lack of medical resources. Therefore, formulating targeted regional intervention measures and reinforcing interdisciplinary collaboration, particularly in low- and middle-income countries, are vital for reducing the incidence of hip fractures and the burden on healthcare institutions.
6 Food insecurity: nutritional deficiencies amplify the risk of hip fractures
Despite no independent statistical analysis of food insecurity in this article, the negative link between socioeconomic-related indicators and Alb levels suggests that nutritional status may be a key mechanism through which SDoH influence the risk of fractures. Previous research has systematically elaborated on the close link between food insecurity and malnutrition,38 a finding further verified by a recent systematic review.39 A study states that food insecurity typically leads to a decline in dietary quality, as individuals, under conditions of food scarcity, tend to opt for more economically viable yet nutritionally inferior foods, like those high in fat, high in carbohydrates, low in vitamins, low in protein, and low in micronutrients.40 This dietary pattern is detrimental to the intake and absorption of critical nutrients (e.g., calcium, vitamin D, and protein), which play a crucial role in bone formation and repair. Furthermore, vitamin D deficiency is closely associated with an increased risk of falls.41 Research has shown that food insecurity causes a reduction in energy and nutrient intake among older people,42 while insufficient energy and nutrient intake are key drivers of frailty.43 Recently, Perez-Zepeda et al. further corroborated that severe food insecurity was directly linked to an increased risk of frailty in adults.44 In short, food insecurity may directly increase the risk of falls among older people through multiple mechanisms, thereby raising the risk of hip fractures. Specifically, malnutrition resulting from food insecurity not only affects bone health but also increases the likelihood of falls among older individuals due to frailty.45 Noticeably, recent studies have suggested that among orthopedic trauma populations, the proportion of individuals with positive food insecurity screening is as high as 37 %, significantly above the national prevalence of 14 %.46 These data further reinforce the strong link between food insecurity and the risk of hip fractures. Based on these findings, interventions targeting food insecurity are of particular importance. In addition to directly addressing food insecurity, targeted nutritional supplementation measures are also necessary in mitigating the adverse effects of food insecurity. Thus, improving nutritional status through social security programs, specific nutritional supplements, or other interventions may dominantly reduce the incidence of hip fractures among individuals at high risk of food insecurity. These interventions not only improve overall nutritional status but also effectively reduce bone health issues caused by malnutrition, particularly among the older population.
7 Educational level: a higher risk of fractures among individuals with a lower educational level
This paper demonstrated a clear negative link between educational level and the risk of hip fractures. Previous meta-analyses have disclosed that individuals with lower educational levels have a 23 % increased risk of fragility fractures relative to those with higher educational levels.47 This observation is further validated by specific incidence data: the age-standardized incidence rates of the first hip fracture within the basic education population (males: 4.3 per 1000 person-years, females: 6.2 per 1000 person-years) were distinctly higher than within the higher education population (males: 3.7 per 1000 person-years, females: 5.6 per 1000 person-years).48 As a valuable indicator reflecting human capital and non-material resources, educational level is relatively stable during adulthood and can independently predict fracture risk regardless of future health conditions.47 This association may be mediated by various modifiable lifestyle behavioral factors. First, smoking and lack of physical activity are more pervasive in the lower educational groups,49 and both have been testified to be independently linked with the risk of hip fractures.50,51 Second, heavy drinking is highly linked to both lower educational levels and high fracture risks.52 Furthermore, in developed countries, lower educational groups (especially females) exhibit higher obesity rates and BMI.53 In contrast, individuals with higher educational levels tend to adopt regular physical exercise and other healthy lifestyles.54,55 These individual behavioral differences collectively constitute a critical pathway through which educational level influences fracture risk. Education is not only linked to health cognition and health behaviors (e.g., smoking cessation and nutritional diet) but also determines an individual's ability to access health resources and comprehend medical advice. Therefore, for future public health interventions, it is recommended to focus on individuals with lower educational levels. By implementing health education and promoting changes in healthy behaviors, their risk of fractures can be reduced. Simultaneously, there is an urgent need to delve deeper into the specific causal link between educational level and fracture risk to optimize intervention strategies. Additionally, improving educational levels may become a vital strategy for enhancing long-term bone health and reducing the incidence of fractures.
8 Role of marital support and medical accessibility
Despite no statistical links between marital status and fracture risk in this paper, its role as a crucial indicator of social connections within SDoH still holds potential health implications. Individuals who are single, widowed, or divorced may not have sufficient health activities and may not be able to cope with acute risk events like falls, possibly due to the lack of social support and life care from a partner. This phenomenon is particularly noteworthy in the older population living alone. Furthermore, although the variable medical insurance was not retained in the predictive model, the synergistic effects of nutritional status, BMI, and SDoH scores suggest that the accessibility of healthcare services may still indirectly influence fracture risk. Inadequate utilization of healthcare services may result in delayed osteoporosis screening and the failure to promptly correct vitamin D deficiency, thereby increasing the likelihood of fractures. While marital status and healthcare accessibility are not direct strong predictors, they remain potential influencers on fracture risk as key dimensions of social support and health management. Public health interventions should prioritize social care support for older people living alone and enhance the accessibility of healthcare for vulnerable groups through policy optimization, like strengthening osteoporosis screening, nutritional interventions, and preventive health management, to reduce the risk of fractures.
9 Limitations
The strength of this paper lies in the utilization of a large sample size and diverse racial distribution from NHANES, which provides sufficient statistical power to test and validate the research hypotheses. However, some limitations are also exposed. Firstly, due to the cross-sectional design, it is infeasible to infer causality, and conclusions can only rely on statistical associations. Secondly, self-reported or recall-based data may introduce measurement bias and misclassification risks. Furthermore, although self-reported indicators of social determinants (e.g., homeownership and marital status) can partially reflect their impact on health, these variables have certain limitations. For instance, homeownership, though reflecting economic status, cannot comprehensively represent community safety or environmental quality. Marital status can serve as a proxy indicator for social support, but it is hard to encompass the complexity of social relationships or the level of individual social engagement. These limitations may lead to biased interpretations of research results in related fields. Hence, future research should combine longitudinal designs to infer causality and incorporate more comprehensive indicators of social determinants (e.g., community environment and quality of social relationships) to more accurately reveal their associations with health outcomes. Additionally, mixed-method research approaches can be considered to address the limitations of quantitative analysis through qualitative data, thereby deepening the understanding of complex social factors.
10 Conclusion
To sum up, this paper offers ample evidence supporting the independent link between the cumulative disadvantages of SDoH and the incidence of hip fractures among Americans. These findings highly underscore that it is urgent to implement multifaceted policy interventions, like educational initiatives, nutritional interventions, income support, and improved access to healthcare services, to mitigate the risk of fractures. Future research should probe into the cumulative effects of SDoH over time on bone health and incorporate a broader range of social determinants, including mental health and community environments, to build a more comprehensive framework for health intervention strategies.
Guardian/patient's consent
As part of the NHANES protocol, written informed consent was obtained from all participants or their guardians.
Availability of data and materials
The datasets generated and/or analyzed during the current study are publicly available in the NHANES repository, https://wwwn.cdc.gov/nchs/nhanes/Default.aspx.
Ethical statement
This study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The NHANES protocol was approved by the Institutional Review Board of the National Center for Health Statistics.
Credit author statement
Bo Chen: Conceptualization, Methodology, Supervision. Fangfang Deng: Data curation, Writing- Original draft preparation, Software, Writing- Reviewing and Editing. Huali Guo: Visualization, Investigation. Li Song: Software, Validation. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding statement
This work was supported by the Yichang city-level project (grant No. A23-1-031).
References
- The economic burden of injury: health care and productivity costs of injuries in the Netherlands. Accid Anal Prev. 2016 Aug;93:92-100.
- [Google Scholar]
- Global, regional and National burden of bone fractures in 204 countries and territories, 1990-2019: a systematic analysis from the global burden of disease study 2019. Lancet Healthy Longev. 2021 Sep;2(9):e580-e592.
- [Google Scholar]
- Direct and indirect costs of non-vertebral fracture patients with osteoporosis in the US. Pharmacoeconomics. 2010;28(5):395-409.
- [Google Scholar]
- Fragility fractures in Europe: burden, management and opportunities. Arch Osteoporosis. 2020 Apr 19;15(1):59.
- [Google Scholar]
- The cost of osteoporosis, osteopenia, and associated fractures in Australia in 2017. J Bone Miner Res. 2019 Apr;34(4):616-625.
- [Google Scholar]
- Public health impact of osteoporosis. J Gerontol A Biol Sci Med Sci. 2013 Oct;68(10):1243-1251.
- [Google Scholar]
- Current strategies in medical management of the geriatric hip fracture patient. J Am Acad Orthop Surg. 2023 Jun 15;31(12):620-626.
- [Google Scholar]
- Costs of fragility hip fractures globally: a systematic review and meta-regression analysis. Osteoporos Int. 2017 Oct;28(10):2791-2800.
- [Google Scholar]
- Characteristics of fragility hip fracture-related falls in the older adults: a systematic review. J Nutr Health Aging. 2024 Oct;28(10)
- [Google Scholar]
- Factors associated with increased risk of falls in older adults. Am J Lifestyle Med. 2023;19(3):419-430.
- [Google Scholar]
- Social determinants of health and cardiovascular disease: Current state and future directions towards healthcare equity. Curr Atheroscler Rep. 2021 Jul 26;23(9):55.
- [Google Scholar]
- Addressing health equity and social determinants of health through healthy people 2030. J Public Health Manag Pract. 2021 Nov-Dec 01;27(Suppl 6):S249-s257.
- [Google Scholar]
- The association between social determinants of health and distal radius fracture outcomes. J Hand Surg Am. 2024 Sep;49(9):875-884.
- [Google Scholar]
- Impact of race and social determinants on operative management of distal radius fractures in medicare patients. Plast Reconstr Surg Glob Open. 2025 Feb;13(2)
- [Google Scholar]
- The impact of social determinants of health on outcomes and complications after total knee arthroplasty: an analysis of neighborhood deprivation indices. J Bone Joint Surg Am. 2024 Feb 21;106(4):288-303.
- [Google Scholar]
- Social determinants of health in patients undergoing hemiarthroplasty: are they associated with medical complications, healthcare utilization, and payments for care? Arch Orthop Trauma Surg. 2023 Dec;143(12):7073-7080.
- [Google Scholar]
- Oral health in America: a report of the surgeon general. J Calif Dent Assoc. 2000 Sep;28(9):685-695.
- [Google Scholar]
- Contribution of individual and cumulative social determinants of health underlying gender disparities in periodontitis in a representative US population: a cross-sectional NHANES study. J Clin Periodontol. 2024 May;51(5):558-570.
- [Google Scholar]
- Association between calcium level, bone mineral density, and hip fracture: analysis based on NHANES data. Healthcare (Basel). 2025 Feb 7;13(4)
- [Google Scholar]
- Global epidemiology of hip fractures: secular trends in incidence rate, post-fracture treatment, and all-cause mortality. J Bone Miner Res. 2023 Aug;38(8):1064-1075.
- [Google Scholar]
- National health and nutrition examination survey: estimation procedures, 2011-2014. Vital Health Stat 2 (177):1-26.
- [Google Scholar]
- The effect of socioeconomic deprivation on fracture incidence in the United Kingdom. Osteoporos Int. 2004 Jul;15(7):520-524.
- [Google Scholar]
- Incidence of hip fracture in Oslo, Norway: differences within the city. Bone. 1998 Feb;22(2):175-178.
- [Google Scholar]
- Occurrence of hip fractures and socioeconomic position. J Aging Health. 2000 May;12(2):193-203.
- [Google Scholar]
- Health-related quality of life and fall risk associated with age-related body composition changes; sarcopenia, obesity and sarcopenic obesity. Intern Med J. 2018 Aug;48(8):973-981.
- [Google Scholar]
- Abdominal obesity, dynapenia and dynapenic-abdominal obesity as factors associated with falls. Braz J Phys Ther. 2019 Nov-Dec;23(6):497-505.
- [Google Scholar]
- Osteosarcopenic obesity: prevalence and relation with frailty and physical performance in middle-aged and older women. J Am Med Dir Assoc. 2017 Aug 1;18(8):733.e1-733.e5.
- [Google Scholar]
- Relationship of incident falls with balance deficits and body composition in Male and female community-dwelling elders. J Nutr Health Aging. 2019;23(1):9-13.
- [Google Scholar]
- Falls and other geriatric syndromes in Blantyre, Malawi: a community survey of older adults. Malawi Med J. 2014 Dec;26(4):105-108.
- [Google Scholar]
- Falls in older people with diabetes: identification of simple screening measures and explanatory risk factors. Prim Care Diabetes. 2020 Dec;14(6):723-728.
- [Google Scholar]
- The impact of physical, neurological and cognitive deficits on risk of falls in people with type 2 diabetes mellitus: a review. Sri Lanka Journal of Diabetes Endocrinology and Metabolism. 2020;10(1)
- [Google Scholar]
- Dementia incidence and mortality in middle-income countries, and associations with indicators of cognitive reserve: a 10/66 dementia research group population-based cohort study. Lancet. 2012 Jul 7;380(9836):50-58.
- [Google Scholar]
- Prevalence of food insecurity and malnutrition, factors related to malnutrition in the elderly: a community-based, cross-sectional study from Turkey. European Geriatric Medicine. 2013;4(4):226-230.
- [Google Scholar]
- Food insecurity and nutritional status among older adults: a systematic review. Nutr Rev. 2022 Mar 10;80(4):631-644.
- [Google Scholar]
- Does living in a food insecure household impact on the diets and body composition of young children? Findings from the Southampton women's survey. J Epidemiol Community Health. 2012 Jun;66(6):e6.
- [Google Scholar]
- Vitamin D deficiency is associated with functional decline and falls in frail elderly women despite supplementation. Osteoporos Int. 2017 Apr;28(4):1347-1353.
- [Google Scholar]
- Nutrient intakes of individuals from food-insufficient households in the United States. Am J Public Health. 1997 Dec;87(12):1956-1961.
- [Google Scholar]
- Low nutrient intake is an essential component of frailty in older persons. J Gerontol A Biol Sci Med Sci. 2006 Jun;61(6):589-593.
- [Google Scholar]
- Frailty and food insecurity in older adults. Public Health Nutr. 2016 Oct;19(15):2844-2849.
- [Google Scholar]
- Food insecurity and health outcomes. Health Aff (Millwood). 2015 Nov;34(11):1830-1839.
- [Google Scholar]
- Socio-economic inequalities in fragility fracture incidence: a systematic review and meta-analysis of 61 observational studies. Osteoporos Int. 2021 Dec;32(12):2433-2448.
- [Google Scholar]
- Trends in hip fracture incidence, recurrence, and survival by education and comorbidity: a Swedish register-based study. Epidemiology. 2021 May 1;32(3):425-433.
- [Google Scholar]
- Health behaviours as explanations for educational level differences in cardiovascular and all-cause mortality: a follow-up of 60 000 men and women over 23 years. Eur J Public Health. 2008 Feb;18(1):38-43.
- [Google Scholar]
- Smoking and fracture risk: a meta-analysis. Osteoporos Int. 2005 Feb;16(2):155-162.
- [Google Scholar]
- Anthropometry, physical activity and hip fractures in the elderly. Injury. 2011 Feb;42(2):188-193.
- [Google Scholar]
- Association between alcohol consumption and both osteoporotic fracture and bone density. Am J Med. 2008 May;121(5):406-418.
- [Google Scholar]
- Socioeconomic status and risk factors for obesity and metabolic disorders in a population-based sample of adult females. Prev Med. 2009 Aug-Sep;49(2-3):165-171.
- [Google Scholar]
- Predicting healthy lifestyle patterns among retirement age older adults in the WELL study: a latent class analysis of sex differences. Maturitas. 2014 Jan;77(1):41-46.
- [Google Scholar]