Abstract
Evidence on the association between cardiovascular health (CVH) and frailty remains limited, and the role of serum 25-hydroxyvitamin D (25(OH)D) in this relationship is not fully understood. This study aimed to examine the association between frailty and CVH and to investigate the independent association of serum 25(OH)D and its potential interaction with frailty. This cross-sectional study used data from the National Health and Nutrition Examination Survey 2007 to 2018, including 8015 adults aged ≥20 years. Frailty was assessed using a frailty index, and CVH was evaluated based on the American Heart Association’s Life’s Essential 8 metrics. Multinomial logistic regression models accounting for the complex survey design were applied to assess associations. Frailty was highly prevalent in the study population. After multivariable adjustment, higher frailty levels were significantly associated with an increased risk of poor CVH in a graded manner. Serum 25(OH)D levels were independently and inversely associated with poor CVH. Stratified analyses showed that the association between frailty and CVH varied by sex and 25(OH)D status. However, no statistically significant multiplicative interaction between frailty and serum 25(OH)D levels was observed. Frailty was significantly associated with poorer CVH in US adults, demonstrating a clear dose–response relationship. Serum 25(OH)D levels were independently associated with CVH, whereas no statistically significant multiplicative interaction between frailty and serum 25(OH)D was observed. These findings highlight the importance of addressing multiple modifiable risk factors to improve CVH.
Keywords: 25-hydroxyvitamin D, cardiovascular health, cross-sectional study, frailty, NHANES
1. Introduction
With the ongoing acceleration of global population aging, frailty has emerged as a major public health concern. Frailty is characterized by a decline in multiple physiological systems and increased vulnerability to stressors, resulting in a reduced capacity to maintain homeostasis.[1,2] A substantial body of evidence has demonstrated that frailty is closely associated with a wide range of adverse outcomes, including falls, disability, hospitalization, institutionalization, and mortality.[3] Globally, the prevalence of frailty among older adults ranges from approximately 10% to 25%, increasing markedly with age and becoming particularly prominent among individuals aged 80 years and older.[4] Despite its high prevalence and serious health consequences, effective strategies for the prevention and management of frailty remain limited, underscoring the urgent need to identify modifiable risk factors and early intervention targets.
Cardiovascular health (CVH) is increasingly recognized as a comprehensive indicator reflecting overall health status. In 2022, the American Heart Association updated its CVH framework to Life’s Essential 8 (LE8), which comprises 8 metrics across 2 domains: health behaviors (diet, physical activity, nicotine exposure, and sleep) and health factors (body mass index [BMI], blood lipids, blood glucose, and blood pressure).[5,6] Previous studies have shown that ideal CVH is strongly associated with lower risks of cardiovascular disease, mortality, and functional decline. Moreover, cardiovascular disease is influenced by multiple metabolic, lifestyle, and sociodemographic factors – such as hypertension, diabetes, physical inactivity, and educational attainment – highlighting the importance of evaluating CVH as a multidimensional construct.[7] In recent years, emerging studies have begun to explore the relationship between CVH and frailty, suggesting that poorer CVH may be associated with a higher risk of frailty. However, current evidence remains limited, largely derived from cross-sectional studies, and predominantly focused on older populations, with inconsistent findings across studies. Given that frailty and poor CVH share several underlying mechanisms – including chronic inflammation, metabolic dysregulation, and reduced physiological reserve – further investigation into their relationship is essential for developing integrated prevention strategies.
Vitamin D, particularly serum 25(OH)D, plays an important role in musculoskeletal health, cardiovascular function, and all-cause mortality.[8] Existing evidence suggests that higher levels of 25(OH)D are associated with more favorable CVH profiles.[9] In addition, numerous studies have demonstrated an inverse association between 25(OH)D levels and frailty, indicating that vitamin D deficiency is linked to increased frailty severity and a higher risk of adverse health outcomes.[10–12] For instance, a large epidemiological study reported that vitamin D deficiency was significantly associated with increased cardiovascular risk and poorer overall health status in community-dwelling populations.[13] However, most previous studies have examined these relationships in isolation – either between vitamin D and frailty or between vitamin D and CVH – while comprehensive analyses integrating all 3 components remain scarce. From a biological perspective, vitamin D may influence both frailty and CVH through pathways involving inflammation regulation, maintenance of muscle function, and cardiovascular metabolic processes.
Despite these advances, several important gaps remain in the current literature. First, most prior studies have focused on pairwise associations (e.g., 25(OH)D and frailty or CVH), with limited attention to their combined effects. Second, research on the relationship between frailty and CVH is still relatively sparse, particularly in the general adult population. Third, it remains unclear whether vitamin D acts independently or interacts with frailty in influencing CVH. Therefore, there is a clear need to systematically evaluate the interrelationships among frailty, vitamin D status, and CVH within a unified analytical framework.
Accordingly, the objectives of this study were to assess the association between frailty and CVH among adults aged ≥20 years and to examine the independent effects of serum 25(OH)D levels and frailty on CVH, as well as to test for potential interaction between them. This study contributes to the existing literature in 3 key aspects. First, to our knowledge, this is the first study to simultaneously examine the relationships among frailty, CVH, and vitamin D status within a unified analytical framework. Second, we evaluated the independent effects of vitamin D and frailty on CVH and further investigated their potential interaction, providing new insights into underlying biological mechanisms. Third, by applying the updated LE8 metrics, this study provides contemporary population-based evidence aligned with current standards for CVH assessment. These findings may enhance our understanding of the interplay between nutritional status, physiological vulnerability, and CVH, and provide a theoretical basis for developing comprehensive interventions targeting modifiable risk factors.
2. Methods
2.1. Study population
This cross-sectional study utilized data from the National Health and Nutrition Examination Survey (NHANES) cycles from 2007 to 2018. NHANES is a nationally representative, cross-sectional survey designed to assess the health and nutritional status of the noninstitutionalized US population using a complex, multistage probability sampling design that incorporates stratification, clustering, and oversampling of specific subgroups.[14]
The survey is conducted biennially by the National Center for Health Statistics under the Centers for Disease Control and Prevention. Data collection includes household interviews, standardized physical examinations conducted in Mobile Examination Centers, and laboratory testing.[15–20]
A total of 25,072 adults aged ≥20 years were initially included. Participants with missing data on CVH, frailty index (FI), serum 25(OH)D concentrations, or poverty income ratio were excluded. Ultimately, 8015 individuals were included in the final analysis (Fig. 1).
Figure 1.
Flowchart for study population. 25(OH)D = 25-hydroxyvitamin D, CI = confidence interval, CVH = cardiovascular health, NHANES = National Health and Nutrition Examination Survey.
2.2. Ethics approval and consent to participate
The NHANES protocol was approved by the National Center for Health Statistics Research Ethics Review Board, and all participants provided written informed consent. All procedures were conducted in accordance with the Declaration of Helsinki.
2.3. Assessment of the physical frailty phenotype
Frailty was assessed using a 36-item FI, adapted from previously validated NHANES-based studies (Table 1). The FI was calculated as the ratio of the number of health deficits present to the total number of deficits considered, with values ranging from 0 to 1. Participants were categorized into 4 groups: FI ≤ 0.1 (fit), 0.1 < FI ≤ 0.2 (pre-frail), 0.2 < FI ≤ 0.3 (mild frailty), and FI > 0.3 (moderate-to-severe frailty).[21–23]
Table 1.
Thirty-six-item frailty index.
| Self-reported items | ||
|---|---|---|
| 1. Angina/angina pectoris | 10. Overnight hospital stays | 19. Difficulty walking between rooms on same floor |
| 2. Heart attack | 11. Medications | 20. Difficulty standing up from armless chair |
| 3. Coronary heart disease | 12. Weak/failing kidneys | 21. Difficulty getting in and out of bed |
| 4. Stroke | 13. Confusion or inability to remember things | 22. Difficulty dressing yourself |
| 5. Thyroid condition | 14. Difficulty managing money | 23. Difficulty grasping/holding small objects |
| 6. Cancer | 15. Difficulty stooping, crouching, kneeling | 24. Difficulty attending social event |
| 7. Arthritis | 16. Difficulty lifting or carrying | 25. Self-reported health |
| 8. High blood pressure | 17. Health compared to 1 yr ago | |
| 9. Diabetes mellitus | 18. Frequency of healthcare use | |
| Laboratory items | ||
|---|---|---|
| 26. Platelet count SI (150–450 unit 1000 cells/µL) | 30. Red cell distribution width (≤14.6%) | 34. Uric acid (M: 240–510, F: 160–430 µmol/L) |
| 27. Blood urea nitrogen (3–20 mg/dL) | 31. Lactate dehydrogenase (≤190 U/L) | 35. Total calcium (2.0–2.5 mmol/L) |
| 28. Bicarbonate (≤28 mmol/L) | 32. Alkaline phosphatase (≤115 U/L) | 36. Pulse rate (60–99 bpm) |
| 29. Systolic blood pressure (90–140 mm Hg) | 33. Pulse pressure (30–60 mm Hg) | |
SI = International System of Units.
2.4. Determining the level of cardiovascular health
CVH was evaluated using the LE8 score recommended by the American Heart Association. This metric includes 4 health behaviors (diet, physical activity, smoking status, and sleep health) and 4 health factors (BMI, blood lipids, blood glucose, and blood pressure). The total CVH score (range: 0–100) was calculated as the average of the 8 components and categorized as low (0–49), intermediate (50–79), or high (80–100)[6] (Table 2).
Table 2.
Metrics for measurement and quantitative assessment of CVH.
| Domain | CVH metric | Method of measurement | Quantification of CVH metric: adults (≥20 yr of age) | |
|---|---|---|---|---|
| Health behaviors | 1.Diet | Measurement: Self-reported daily intake of a DASH-style eating pattern. |
Quantiles of DASH-style diet adherence or HEI-2015 (population) | |
| Scoring (population): | ||||
| Points | Quantile | |||
| 100 | ≥95th percentile (top/ideal diet) | |||
| 80 | 75th–94th percentile | |||
| 50 | 50th–74th percentile | |||
| 25 | 25th–49th percentile | |||
| 0 | 1st–24th percentile (bottom/least ideal quartile) | |||
| 2.Physical activity | Measurement: Self-reported minutes of moderate or vigorous PA per week |
Metric: Minutes of moderate- (or greater) intensity activity per week | ||
| Scoring: | ||||
| Points | Minutes | |||
| 100 | ≥150 | |||
| 90 | 120–149 | |||
| 80 | 90–119 | |||
| 60 | 60–89 | |||
| 40 | 30–59 | |||
| 20 | 1–29 | |||
| 0 | 0 | |||
| 3.Nicotine exposure | Measurement: Self-reported use of cigarettes |
Metric: Combustible tobacco use; or secondhand smoke exposure | ||
| Scoring: | ||||
| Points | Status | |||
| 100 | Never smoker | |||
| 75 | Former smoker, quit ≥5 yr | |||
| 50 | Former smoker, quit 1–<5 yr | |||
| 25 | Former smoker, quit <1 yr | |||
| 0 | Current smoker | |||
| Subtract 20 points (unless score is 0) for living with active indoor smoker in home | ||||
| 4.Sleep health | Measurement: Self-reported average hours of sleep per night |
Metric: Average hours of sleep per night | ||
| Scoring: | ||||
| Points | Level | |||
| 100 | 7–<9 | |||
| 90 | 9–<10 | |||
| 70 | 6–<7 | |||
| 40 | 5–<6 or ≥10 | |||
| 20 | 4–<5 | |||
| 0 | <4 | |||
| Health factors | 5.BMI | Measurement: Body weight (kilograms) divided by height squared (meters squared) |
Metric: BMI (kg/m2) | |
| Scoring: | ||||
| Points | Level | |||
| 100 | <25 | |||
| 70 | 25.0–29.9 | |||
| 30 | 30.0–34.9 | |||
| 15 | 35.0–39.9 | |||
| 0 | ≥40.0 | |||
| 6.Blood lipids | Measurement: Plasma total and HDL cholesterol with calculation of non-HDL cholesterol |
Metric: Non-HDL cholesterol (mg/dL) | ||
| Scoring: | ||||
| Points | Level | |||
| 100 | <130 | |||
| 60 | 130–159 | |||
| 40 | 160–189 | |||
| 20 | 190–219 | |||
| 0 | ≥220 | |||
| If drug-treated level, subtract 20 points | ||||
| 7.Blood glucose | Measurement: FBG or casual HbA1c |
Metric: FBG (mg/dL) or HbA1c (%) | ||
| Scoring: | ||||
| Points | Level | |||
| 100 | No history of diabetes and FBG <100 (or HbA1c <5.7) | |||
| 60 | No diabetes and FBG 100–125 (or HbA1c 5.7–6.4) (prediabetes) | |||
| 40 | Diabetes with HbA1c <7.0 | |||
| 30 | Diabetes with HbA1c 7.0–7.9 | |||
| 20 | Diabetes with HbA1c 8.0–8.9 | |||
| 10 | Diabetes with HbA1c 9.0–9.9 | |||
| 0 | Diabetes with HbA1c ≥10.0 | |||
| 8.Blood pressure | Measurement: Appropriately measured systolic and diastolic BPs |
Metric: Systolic and diastolic BPs (mm Hg) | ||
| Scoring: | ||||
| Points | Level | |||
| 100 | <120/<80 (optimal) | |||
| 75 | 120–129/<80 (elevated) | |||
| 50 | 130–139 or 80–89 (stage 1 hypertension) | |||
| 25 | 140–159 or 90–99 | |||
| 0 | ≥160 or ≥100 | |||
| Subtract 20 points if treated level | ||||
BMI = body mass index, BP = blood pressure, CVH = cardiovascular health, DASH = dietary approaches to stop hypertension, FBG = fasting blood glucose, HbA1c = hemoglobin A1c, HDL = high-density lipoprotein.
2.5. Measurement of serum 25(OH)D
Serum 25(OH)D, including both 25(OH)D2 and 25(OH)D3, was obtained from NHANES laboratory data.[15] Based on previous studies, concentrations were categorized into 2 groups: <75 nmol/L and 75 to 250 nmol/L.[24]
2.6. Covariates
Potential confounders considered in this study included demographic variables (age, sex, and race/ethnicity), socioeconomic factors (education level, marital status, and poverty income ratio), and season of examination. Because CVH was assessed using the LE8 score, which already incorporates diet, physical activity, smoking exposure, sleep health, BMI, blood lipids, blood pressure, and blood glucose, these individual LE8 components were described as participant characteristics but were not included as covariates in the final multivariable regression models to avoid potential overadjustment.
2.7. Statistical analysis
The NHANES provides sample weights for each participant. All analyses incorporated these sampling weights, along with strata and primary sampling units, to account for the complex survey design and to obtain nationally representative estimates.
Categorical variables were presented as weighted percentages, and group comparisons were performed using the Rao–Scott χ2 test. Continuous variables with non-normal distributions were summarized as weighted medians and interquartile ranges, and differences between groups were assessed using weighted rank-sum tests.
The outcome variable, CVH, was categorized into 3 ordered groups (low, intermediate, and high). An ordinal logistic regression model was initially considered; however, the proportional odds assumption was formally tested and found to be violated. Therefore, multinomial logistic regression was applied because it does not require the proportional odds assumption and allows category-specific estimation of associations across CVH levels. The final multivariable models were adjusted for age, sex, race/ethnicity, education level, marital status, poverty income ratio, season of examination, and serum 25(OH)D levels. To evaluate potential effect modification, multiplicative interaction terms between FI and serum 25(OH)D were tested, and prespecified subgroup analyses were subsequently performed.
Key assumptions of multinomial logistic regression were evaluated, including independence of observations, absence of multicollinearity (assessed using variance inflation factors), and correct model specification. The independence of irrelevant alternatives, an inherent assumption of multinomial logistic regression, was considered when interpreting the results.
All analyses were performed using SAS version 9.4 (SAS Institute Inc.) with survey-specific procedures. A two-sided P < .05 was considered statistically significant.
3. Results
Weighted baseline characteristics of the study participants are presented in Table 3. In the weighted analysis, the prevalence of frailty and pre-frailty was 85.5%, corresponding to approximately 30.7 million US adults. The mean age of frail participants was 63 ± 14 years, and 53% were female. Overall, CVH levels among frail individuals were predominantly low or intermediate. Across individual components, glucose, blood pressure, and lipid scores showed unfavorable distributions. In addition, obesity, 25(OH)D deficiency, and adverse lifestyle factors were common among frail participants.
Table 3.
Weighted baseline characteristics of participants across the spectrum of frailty (stratified by frailty index, FI), NHANES 2007 to 2018.
| Characteristic | Frailty index (FI) | P-value | |
|---|---|---|---|
| FI ≤ 0.1 (fit) | FI > 0.1 (frail) | ||
| Number of participants, unweighted (n) | 1086 (13.40) | 7019 (86.60) | |
| Population estimate, weighted (N) | 5,211,554 | 30,669,576 | |
| Gender | <.001*** | ||
| Male | 2,928,908 (56%) | 14,413,855 (47%) | |
| Female | 2,282,646 (44%) | 16,255,722 (53%) | |
| Age (≥20 yr) | 55 ± 16 | 63 ± 14 | <.001*** |
| Ethnicity | <.001*** | ||
| Mexican American | 1,273,019 (24%) | 6,133,369 (20.2%) | |
| Non-Hispanic White | 2,438,346 (47%) | 16,561,357 (54%) | |
| Non-Hispanic Black | 925,658 (18%) | 6,321,516 (21%) | |
| Other/multiracial | 574,530 (11%) | 1,653,333 (5.4%) | |
| Education level | <.001*** | ||
| Less than high school | 1,169,516 (22.5%) | 8,878,948 (29%) | |
| High school/GED/equivalent | 1,191,559 (23%) | 7,555,351 (25%) | |
| College/AA degree | 1,479,924 (28%) | 8,703,067 (28%) | |
| College graduate or above | 1,370,554 (26%) | 5,507,891 (18%) | |
| Marital status | <.001*** | ||
| Living with partner | 3,308,357 (63.3%) | 17,258,545 (56%) | |
| Living without partner | 1,903,198 (36.7%) | 13,392,283 (44%) | |
| Family poverty income ratio | <.001*** | ||
| <1.30 | 1,553,458 (30%) | 10,934,803 (36%) | |
| 1.30–3.50 | 1,869,847 (36%) | 11,719,889 (38%) | |
| >3.50 | 1,788,249 (34%) | 8,014,883 (26%) | |
| Season | .10 | ||
| Winter or spring | 2,435,000 (47%) | 13,105,666 (43%) | |
| Summer or autumn | 2,776,554 (53%) | 17,563,911 (57%) | |
| Body mass index score | <.001*** | ||
| ≤30 | 1,433,059 (27%) | 12,989,695 (42%) | |
| 31–70 | 2,074,199 (40%) | 10,218,302 (33%) | |
| 71–100 | 1,704,296 (33%) | 7,461,580 (24%) | |
| Blood lipids score | <.001*** | ||
| ≤40 | 2,099,778 (40%) | 11,888,089 (39%) | |
| 41–60 | 1,465,273 (28%) | 5,649,543 (18%) | |
| 61–80 | 248,633 (4.8%) | 6,068,728 (20%) | |
| 81–100 | 1,397,869 (27%) | 7,063,217 (23%) | |
| Blood sugar score | <.001*** | ||
| ≤60 | 2,579,611 (49%) | 20,135,072 (66%) | |
| 61–100 | 2,631,943 (51%) | 10,534,504 (34%) | |
| Blood pressure score | <.001*** | ||
| ≤25 | 597,490 (11%) | 9,300,330 (30%) | |
| 26–50 | 1,270,205 (24%) | 6,866,582 (22%) | |
| 51–80 | 1,038,871 (20%) | 9,099,221 (30%) | |
| 81–100 | 2,304,988 (44%) | 5,403,444 (18%) | |
| Sleep score | <.006** | ||
| ≤90 | 2,296,724 (44%) | 15,546,378 (51%) | |
| 91–100 | 2,914,830 (56%) | 15,123,199 (49%) | |
| Smoking score | <.001*** | ||
| ≤55 | 1,297,628 (25%) | 7,833,920 (26%) | |
| 56–75 | 968,521 (19%) | 8,456,706 (28%) | |
| 76–100 | 2,945,404 (57%) | 14,378,950 (47%) | |
| Physical activity score | <.001*** | ||
| ≤15 | 2,538,678 (49%) | 19,134,308 (62%) | |
| 16–90 | 792,384 (15%) | 4,354,815 (14%) | |
| 91–100 | 1,880,491 (36%) | 7,180,453 (23%) | |
| Diet score | .093 | ||
| ≤20 | 2,310,671 (44%) | 15,113,384 (49%) | |
| 21–100 | 2,900,883 (56%) | 15,556,193 (51%) | |
| 25(OH)D (nmol/L) | <.001*** | ||
| <75 | 3,547,172 (68%) | 18,329,893 (60%) | |
| 75–250 | 1,664,382 (32%) | 12,339,683 (40%) | |
| Life’s Essential 8 score | <.001*** | ||
| Poor (<50) | 817,367 (16%) | 11,972,876 (39%) | |
| Intermediate (50–79) | 4,059,685 (78%) | 17,988,742 (59%) | |
| Ideal (≥80) | 334,501 (6.4%) | 707,958 (2.3%) | |
Data are presented as weighted frequencies (percentages) for categorical variables and weighted mean ± SD for continuous variables. Group comparisons: design-based Kruskal–Wallis test for continuous variables; Rao–Scott adjusted χ2 test for categorical variables. All analyses incorporated NHANES examination sample weights.
Frailty was defined by a frailty index (FI) > 0.1, which captures the full continuum of vulnerability, including both pre-frail (0.1 < FI ≤ 0.2) and frail (FI > 0.2) stages. Therefore, the “Frail” group in this study is not limited to moderate-to-severe frailty.
25(OH)D = 25-hydroxyvitamin D, FI = frailty index, GED = General Educational Development, NHANES = National Health and Nutrition Examination Survey.
P < .01.
P < .001.
Significant differences were observed between frail and non-frail individuals in age, sex, race/ethnicity, education level, marital status, poverty income ratio, BMI, glucose score, blood pressure score, lipid score, sleep score, smoking score, physical activity score, CVH score, and 25(OH)D levels (all P < .001). In contrast, no significant differences were found in dietary score or season (P = .093 and P = .10, respectively).
As shown in Table 4, in the multinomial logistic regression analysis with ideal CVH as the reference category, 25(OH)D levels were significantly inversely associated with poor CVH, whereas the FI was significantly positively associated with poor CVH (both P < .001). No statistically significant multiplicative interaction between FI and serum 25(OH)D was identified in the multinomial logistic regression models (P for interaction > .05). For intermediate CVH, only FI remained significantly associated with CVH (P < .01), whereas neither serum 25(OH)D nor the FI × 25(OH)D interaction term reached statistical significance.
Table 4.
Interaction effects of 25(OH)D and frailty index on cardiovascular health.
| Variable | Outcome variable | OR (95% CI) | P |
|---|---|---|---|
| 25(OH)D | Poor | 0.17 (0.07–0.38) | <.001*** |
| 25(OH)D | Intermediate | 0.68 (0.32–1.47) | .334 |
| FI | Poor | 4.13 (2.17–7.86) | <.001*** |
| FI | Intermediate | 3.06 (1.48–6.35) | <.01** |
| 25(OH)D × FI | Poor | 1.07 (0.73–1.58) | .717 |
| 25(OH)D × FI | Intermediate | 0.78 (0.52–1.17) | .242 |
Ideal CVH is the reference group. All models are weighted using NHANES examination sample weights and adjusted for age, gender, marital status, ethnicity, education, poverty income ratio, and season.
25(OH)D = 25-hydroxyvitamin D, CI = confidence interval, CVH = cardiovascular health, FI = frailty index, NHANES = National Health and Nutrition Examination Survey, OR = odds ratio.
P < .01.
P < .001.
As shown in Table 5, in the multivariable-adjusted model, the FI was significantly positively associated with the risk of poor CVH, with a graded increase in risk across higher FI levels. Compared with participants with FI ≤ 0.1, the risk of poor CVH increased progressively across higher FI categories.
Table 5.
Adjusted RRs and 95% CIs for FI with poor CVH, overall and by subgroup.
| Groups | FI ≤ 0.1 | 0.1 < FI ≤ 0.2 | 0.2 < FI ≤ 0.3 | FI > 0.3 | P for interaction |
|---|---|---|---|---|---|
| Case subjects/N | 179/1086 | 790/3105 | 683/1785 | 1211/2129 | – |
| Overall† | 1.0 | 2.11 (1.61–2.76) | 3.83 (2.81–5.22) | 6.95 (5.20–9.28) | |
| Gender‡ | <.001*** | ||||
| Male | 1.0 | 2.10 (1.52–2.89) | 3.86 (2.56–5.83) | 5.44 (3.75–7.88) | |
| Female | 1.0 | 2.22 (1.54–3.20) | 4.12 (2.78–6.09) | 8.93 (6.07–13.12) | |
| 25(OH)D (nmol/L)‡ | <.001*** | ||||
| <75 | 1.0 | 1.91 (1.40–2.61) | 3.32 (2.35–4.70) | 6.26 (4.51–8.69) | |
| 75–250 | 1.0 | 3.44 (1.94–6.10) | 6.96 (3.71–13.08) | 12.13 (6.71–21.94) | |
All estimates were weighted using NHANES examination sample weights.
25(OH)D = 25-hydroxyvitamin D, CI = confidence interval, CVH = cardiovascular health, FI = frailty index, NHANES = National Health and Nutrition Examination Survey, RR = relative risk.
Adjusted for age, gender, marital status, ethnicity, education, poverty income ratio, season, and 25(OH)D.
Adjusted for all variables in † except the stratification variable.
P < .001.
In prespecified subgroup analyses, significant heterogeneity in the associations was observed across sex and serum 25(OH)D strata (both P for heterogeneity <.001). In analyses stratified by serum 25(OH)D levels, a graded positive association between FI and the risk of poor CVH was observed in both groups, with a stronger association among participants with sufficient serum 25(OH)D concentrations.
In the continuous variable analysis, 25(OH)D levels were significantly inversely associated with the risk of poor CVH (OR = 0.990, 95% confidence interval: 0.986–0.994, P < .001) in Table 6, and the direction of the association between FI and poor CVH was consistent with that observed in the primary analysis.
Table 6.
Association of FI and 25(OH)D with poor CVH.
| Variable | OR (95% CI) | P |
|---|---|---|
| 0.1 < FI ≤ 0.2 | 2.108 (1.611–2.759) | <.001*** |
| 0.2 < FI ≤ 0.3 | 3.829 (2.809–5.221) | <.001*** |
| FI > 0.3 | 6.946 (5.198–9.280) | <.001*** |
| 25(OH)D (continuous) | 0.990 (0.986–0.994) | <.001*** |
Ideal CVH was the reference group for FI categories. All models are weighted using NHANES examination sample weights and adjusted for age, gender, marital status, ethnicity, education, poverty income ratio, and season.
25(OH)D = 25-hydroxyvitamin D, CVH = cardiovascular health, FI = frailty index, NHANES = National Health and Nutrition Examination Survey, OR = odds ratio.
P < .001.
4. Discussion
In this nationally representative sample of US adults, we systematically evaluated the relationships among the FI, CVH (based on the LE8 metrics), and serum 25(OH)D levels. Our findings demonstrated a significant and graded positive association between FI and poor CVH, with a clear dose–response pattern indicating progressively higher risk of poor CVH with increasing frailty severity. In addition, serum 25(OH)D levels were independently and inversely associated with the risk of poor CVH. Stratified analyses further suggested heterogeneity in the associations according to sex and serum 25(OH)D status, with stronger associations observed among women and individuals with sufficient 25(OH)D levels. However, no statistically significant multiplicative interaction between FI and serum 25(OH)D was identified.
Compared with prior studies that primarily focused on single factors, this study integrates frailty, CVH, and 25(OH)D status within a unified analytical framework, thereby providing more comprehensive population-based evidence. While CVH has been widely recognized as a robust composite indicator for predicting cardiovascular events and all-cause mortality,[6,25] evidence regarding its relationship with frailty remains relatively limited. Moreover, although previous studies have independently examined the associations of 25(OH)D with frailty or CVH,[26–28] few have simultaneously investigated these 3 components. By leveraging a nationally representative dataset, our study confirms the robustness of the association between frailty and CVH and further highlights the differential roles of sex and 25(OH)D status, thus extending the current body of literature.
Consistent with previous research, we observed a strong positive association between FI and the risk of poor CVH, with a stepwise increase in risk across higher FI levels. Frail individuals are more likely to present with multiple cardiometabolic abnormalities, including hypertension, dysglycemia, and dyslipidemia, which collectively contribute to impaired CVH.[29] In addition, prospective studies have demonstrated that frailty is associated with increased risks of cardiovascular events and mortality.[30] The observed dose–response relationship in our study further reinforces the importance of frailty as an integrative marker of overall health status in cardiovascular risk assessment.
From a mechanistic perspective, frailty is considered a multidimensional syndrome characterized by cumulative declines across multiple physiological systems. Its development has been closely linked to chronic inflammation, oxidative stress, endothelial dysfunction, and impaired energy metabolism. Notably, these processes are also central to the pathogenesis of atherosclerosis and cardiovascular disease.[31] Therefore, the observed association between FI and CVH likely reflects shared underlying pathophysiological pathways.
In sex-stratified analyses, the association between FI and poor CVH was more pronounced in women, which is consistent with previous findings.[32] Several factors may explain this difference. Women generally have a longer life expectancy and are therefore more likely to reach ages at which frailty becomes prevalent. In addition, postmenopausal declines in estrogen levels may exacerbate inflammatory responses and vascular dysfunction, thereby increasing cardiovascular susceptibility. Differences in health behaviors, social factors, and chronic disease burden may also contribute to the amplified impact of frailty on CVH among women. These findings underscore the importance of considering sex-specific strategies in cardiovascular prevention.
Regarding vitamin D, our results showed that serum 25(OH)D levels were significantly inversely associated with poor CVH, in line with existing literature. Epidemiological studies have consistently reported that vitamin D deficiency is associated with major cardiovascular risk factors, including hypertension, diabetes, and dyslipidemia,[33] as well as lower CVH scores.[7] Some interventional studies further suggest that vitamin D supplementation may improve certain cardiometabolic parameters.[34] Potential mechanisms include modulation of the renin–angiotensin system, improvement of endothelial function, and anti-inflammatory and antioxidative effects.[35]
Interestingly, subgroup analyses suggested that the association between frailty and poor CVH was stronger among individuals with sufficient serum 25(OH)D concentrations. Although this observation may appear inconsistent with the absence of a statistically significant multiplicative interaction, these analyses address different statistical questions. As discussed in methodological studies, subgroup analyses and multiplicative interaction tests evaluate different aspects of effect modification.[36] The interaction term evaluates whether the joint effect of frailty and serum 25(OH)D deviates from multiplicativity, whereas subgroup analyses assess whether the strength of the association differs across predefined strata. Therefore, our findings suggest potential effect heterogeneity rather than evidence of a synergistic interaction between frailty and vitamin D status. The stronger association observed among participants with sufficient serum 25(OH)D concentrations may reflect differences in baseline health status, residual confounding, or variations in population characteristics and categorization methods. Given the observational nature of this study, these subgroup findings should be interpreted cautiously and warrant confirmation in future prospective studies.
The continuous variable analysis further supported the robustness of our findings. When 25(OH)D was modeled as a continuous variable, its inverse association with poor CVH remained significant, and the direction of the association between FI and CVH was unchanged. This consistency indicates that our results are not driven by arbitrary categorization and enhances the methodological reliability of the study.
Several limitations should be acknowledged. First, due to the cross-sectional design, causal relationships cannot be established. Second, although multiple confounders were adjusted for, residual confounding cannot be completely excluded. Third, the FI was constructed based on self-reported measures, which may introduce information bias. Fourth, a single measurement of serum 25(OH)D may not accurately reflect long-term vitamin D status. Additionally, 25(OH)D levels may be influenced by unmeasured factors such as supplement use and sunlight exposure. Furthermore, a substantial number of participants were excluded because of missing data on frailty, serum 25(OH)D, CVH, or covariates. Although complete-case analysis ensured consistency across all multivariable models, these exclusions may have introduced selection bias and limited the generalizability of the findings. Future studies based on prospective cohorts and randomized controlled trials are warranted to validate these findings and explore underlying mechanisms.
In conclusion, in a nationally representative sample of US adults, frailty was significantly and progressively associated with poorer CVH. Serum 25(OH)D levels were independently associated with CVH, whereas no statistically significant multiplicative interaction between frailty and serum 25(OH)D was observed. From a public health perspective, these findings highlight the importance of addressing modifiable risk factors through multidimensional strategies, including promoting physical activity, maintaining a healthy weight, improving metabolic health, and optimizing serum 25(OH)D status. Such approaches may provide complementary benefits in preventing frailty progression and improving CVH. Further prospective studies are needed to clarify causal relationships and inform targeted intervention strategies.
5. Conclusion
In conclusion, in the US population, frailty was significantly associated with poorer CVH, with higher frailty levels corresponding to an increased risk of poor CVH. Serum 25(OH)D levels were independently associated with CVH. However, no statistically significant multiplicative interaction between frailty and serum 25(OH)D was observed, suggesting that frailty and serum 25(OH)D may influence CVH through different mechanisms. Further prospective studies are warranted to clarify the underlying mechanisms.
Acknowledgments
The authors gratefully thank all NHANES participants and staff for their contributions to the survey.
Author contributions
Conceptualization: Wei Yan, Yihui Kong, Yuxuan Yang, Ying Chen, Lu Liu.
Data curation: Wei Yan, Yihui Kong, Yongle Wang, Lu Liu.
Funding acquisition: Wei Yan, Yihui Kong.
Formal analysis: Yanbing Pan, Lu Liu.
Investigation: Yanbing Pan, Yanzi Zhang, Lu Liu.
Methodology: Yongle Wang, Yanzi Zhang, Lu Liu.
Project administration: Lu Liu.
Resources: Wei Yan, Yihui Kong, Qiuting Jia.
Software: Qiuting Jia, Lu Liu.
Supervision: Wei Yan, Yihui Kong.
Validation: Yuxuan Yang, Lu Liu.
Visualization: Ying Chen.
Writing – original draft: Lu Liu.
Writing – review & editing: Yihui Kong, Lu Liu.
Abbreviations:
- 25(OH)D
- 25-hydroxyvitamin D
- BMI
- body mass index
- CVH
- cardiovascular health
- FI
- frailty index
- LE8
- Life’s Essential 8
- NHANES
- National Health and Nutrition Examination Survey
This research was supported by the Wu Jieping Medical Foundation (320.6750.2020-06-78) and the Beijing Health Alliance Charitable Foundation (KH004269003).
The authors have no conflicts of interest to declare.
The datasets generated and/or analyzed during the current study are publicly available.
How to cite this article: Yan W, Kong Y, Wang Y, Pan Y, Zhang Y, Jia Q, Yang Y, Chen Y, Liu L. Association of serum 25-hydroxyvitamin D and physical frailty with cardiovascular health in US adults: A cross-sectional study. Medicine 2026;105:42(e50862).
WY and YK contributed to this article equally.
Contributor Information
Wei Yan, Email: 18846015678@163.com.
Yihui Kong, Email: kongyihui@126.com.
Yongle Wang, Email: wyl594800115@163.com.
Yanbing Pan, Email: pyb1753323507@163.com.
Yanzi Zhang, Email: zyz86489086@163.com.
Qiuting Jia, Email: jqt13579836065@163.com.
Yuxuan Yang, Email: 18545139268@163.com.
Ying Chen, Email: cyld1125@163.com.
References
- [1].Dent E, Martin FC, Bergman H, Woo J, Romero-Ortuno R, Walston JD. Management of frailty: opportunities, challenges, and future directions. Lancet. 2019;394:1376–86. [DOI] [PubMed] [Google Scholar]
- [2].Hoogendijk EO, Afilalo J, Ensrud KE, Kowal P, Onder G, Fried LP. Frailty: implications for clinical practice and public health. Lancet. 2019;394:1365–75. [DOI] [PubMed] [Google Scholar]
- [3].Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. Lancet. 2013;381:752–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].O’Caoimh R, Sezgin D, O’Donovan MR, et al. Prevalence of frailty in 62 countries across the world: a systematic review and meta-analysis of population-level studies. Age Ageing. 2021;50:96–104. [DOI] [PubMed] [Google Scholar]
- [5].Shetty NS, Parcha V, Patel N, et al. AHA Life’s Essential 8 and ideal cardiovascular health among young adults. Am J Prev Cardiol. 2022;13:100452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Lloyd-Jones DM, Allen NB, Anderson CAM, et al. Life’s Essential 8: updating and enhancing the American Heart Association’s construct of cardiovascular health: a presidential advisory from the American Heart Association. Circulation. 2022;146:e18–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Bayrakçeken E, Yarali S, Ercan U, Alkan O. Patterns among factors associated with myocardial infarction: chi-squared automatic interaction detection tree and binary logit model. BMC Public Health. 2025;25:296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Holick MF, Binkley NC, Bischoff-Ferrari HA, et al. Evaluation, treatment, and prevention of vitamin D deficiency: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2011;96:1911–30. [DOI] [PubMed] [Google Scholar]
- [9].Dziedzic EA, Smyk W, Sowińska I, Dąbrowski M, Jankowski P. Serum level of vitamin D is associated with severity of coronary atherosclerosis in postmenopausal women. Biology (Basel). 2021;10:1139. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Zhou J, Huang P, Liu P, et al. Association of vitamin D deficiency and frailty: a systematic review and meta-analysis. Maturitas. 2016;94:70–6. [DOI] [PubMed] [Google Scholar]
- [11].Ju SY, Lee JY, Kim DH. Low 25-hydroxyvitamin D levels and the risk of frailty syndrome: a systematic review and dose-response meta-analysis. BMC Geriatr. 2018;18:206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Marcos-Pérez D, Sánchez-Flores M, Proietti S, et al. Low vitamin D levels and frailty status in older adults: a systematic review and meta-analysis. Nutrients. 2020;12:2286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Lee JH, O’Keefe JH, Bell D, Hensrud DD, Holick MF. Vitamin D deficiency an important, common, and easily treatable cardiovascular risk factor? J Am Coll Cardiol. 2008;52:1949–56. [DOI] [PubMed] [Google Scholar]
- [14].Chen TC, Clark J, Riddles MK, Mohadjer LK, Fakhouri THI. National Health and Nutrition Examination Survey, 2015–2018: sample design and estimation procedures. Vital Health Stat 2, Data evaluation and methods research. 2020:1–35. [PubMed] [Google Scholar]
- [15].Zipf G, Chiappa M, Porter KS, Ostchega Y, Lewis BG, Dostal J. National Health and Nutrition Examination Survey: plan and operations, 1999–2010. Vital Health Stat 1. 2013:1–37. [PubMed] [Google Scholar]
- [16].Parcha V, Patel N, Kalra R, Arora G, Arora P. Prevalence, awareness, treatment, and poor control of hypertension among young American adults: race-stratified analysis of the National Health and Nutrition Examination Survey. Mayo Clin Proc. 2020;95:1390–403. [DOI] [PubMed] [Google Scholar]
- [17].Kalra R, Parcha V, Patel N, et al. Increased awareness, inadequate treatment, and poor control of cardiovascular risk factors in American young adults: 2005–2016. Eur J Prev Cardiol. 2021;28:304–12. [DOI] [PubMed] [Google Scholar]
- [18].Patel N, Bhargava A, Kalra R, et al. Trends in lipid, lipoproteins, and statin use among U.S. adults: impact of 2013 cholesterol guidelines. J Am Coll Cardiol. 2019;74:2525–8. [DOI] [PubMed] [Google Scholar]
- [19].Patel N, Kalra R, Bhargava A, Arora G, Arora P. Ideal cardiovascular health among American adults after the economic recession of 2008–2009: insights from NHANES. Am J Med. 2019;132:1182–90.e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Centers for Disease Control and Prevention, National Center for Health Statistics. About the National Health and Nutrition Examination Survey. 2019. Available at: https://www.cdc.gov/nchs/nhanes/index.htm. Accessed January 18, 2024.
- [21].Blodgett JM, Theou O, Howlett SE, Rockwood K. A frailty index from common clinical and laboratory tests predicts increased risk of death across the life course. Geroscience. 2017;39:447–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Theou O, Blodgett JM, Godin J, Rockwood K. Association between sedentary time and mortality across levels of frailty. CMAJ. 2017;189:E1056–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Jayanama K, Theou O, Blodgett JM, Cahill L, Rockwood K. Frailty, nutrition-related parameters, and mortality across the adult age spectrum. BMC Med. 2018;16:188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].de la Guía-Galipienso F, Martínez-Ferran M, Vallecillo N, Lavie CJ, Sanchis-Gomar F, Pareja-Galeano H. Vitamin D and cardiovascular health. Clin Nutr. 2021;40:2946–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Abramov D, Kobo O, Mamas MA. Association of cardiovascular health metrics and mortality among individuals with and without cancer. J Am Heart Assoc. 2024;13:e032683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Ling Y, Xu F, Xia X, et al. Vitamin D supplementation reduces the risk of fall in the vitamin D deficient elderly: an updated meta-analysis. Clin Nutr. 2021;40:5531–7. [DOI] [PubMed] [Google Scholar]
- [27].Huang F, Zhou Y, Li T, et al. Association between vitamin D and cardiovascular health in Chinese children and adolescents: basing on Life’s Essential 8. Nutr Metab Cardiovasc Dis. 2024;34:2579–88. [DOI] [PubMed] [Google Scholar]
- [28].Chen L, Li X, Lv Y, et al. Physical frailty, adherence to ideal cardiovascular health and risk of cardiovascular disease: a prospective cohort study. Age Ageing. 2023;52:afac311. [DOI] [PubMed] [Google Scholar]
- [29].Wang YJ, Geng TT, Xun K, et al. Association of frailty with mortality and incident cardiovascular disease: the role of metabolic status. Biomed Environ Sci. 2025;38:1490–501. [DOI] [PubMed] [Google Scholar]
- [30].Jiang H, Wong JJ, Tan RS, et al. Effect of frailty on cardiovascular clinical trials: a systematic review and meta-analysis. JACC Adv. 2025;4:101889. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [31].Johnson N, Qu J, Wagatsuma K, et al. Frailty and cardiovascular disease: a bidirectional relationship with clinical implications. Front Cardiovasc Med. 2025;12:1684701. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Zeidan RS, McElroy T, Rathor L, Martenson MS, Lin Y, Mankowski RT. Sex differences in frailty among older adults. Exp Gerontol. 2023;184:112333. [DOI] [PubMed] [Google Scholar]
- [33].Daniel JB, de Farias Costa PR, Pereira M, Oliveira AM. Vitamin D deficiency and cardiometabolic risk factors in adolescents: systematic review and meta-analysis. Rev Endocr Metab Disord. 2022;23:995–1010. [DOI] [PubMed] [Google Scholar]
- [34].Thompson B, Waterhouse M, English DR, et al. Vitamin D supplementation and major cardiovascular events: D-Health randomised controlled trial. BMJ. 2023;381:e075230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].da Cunha CLP. Vitamin D and the cardiovascular system. Arq Bras Cardiol. 2024;121:e20240189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [36].Iakunchykova O, Lange T, Leon DA. Multiplicative and additive interactions between risk factors for coronary heart disease. Ann Epidemiol. 2024;91:82–4. [DOI] [PubMed] [Google Scholar]