Five-year trajectories of gait speed and hand grip strength in older adults with cardiometabolic multimorbidity: a national retrospective cohort study

Abstract

Background

Cardiometabolic multimorbidity, defined as the coexistence of diabetes and heart disease, is increasingly common in older adults and is associated with adverse health outcomes. While individual conditions such as diabetes or heart disease have been linked to declines in physical function, little is known about how their coexistence affects objective physical performance measures over time. This study aimed to compare changes in gait speed and hand grip strength over five years among older adults with and without cardiometabolic multimorbidity and to identify factors associated with these declines.

Methods

We conducted a retrospective secondary analysis of the National Health and Aging Trends Study linked to Medicare administrative data from 2015 to 2019. Community-dwelling participants aged 66 years and older enrolled in Medicare fee-for-service were included. The participants were categorized into four groups: diabetes only, heart disease only, both conditions, or neither. Gait speed and hand grip strength were assessed annually over five years. Generalized estimating equation models were used to estimate changes in physical function, adjusting for demographics, clinical characteristics, and socioeconomic factors.

Results

The analytic sample included 4,351 participants. At baseline, older adults with both diabetes and heart disease presented significantly lower gait speed and hand grip strength than those with only one or neither condition. Over five years, the cardiometabolic multimorbidity group experienced the most pronounced declines. In fully adjusted models, cardiometabolic multimorbidity was associated with a decline in gait speed (β = -0.034, SE = 0.010) and hand grip strength (β = -0.048, SE = 0.015). Additionally, female sex (β = -0.049 for gait speed; β = -0.460 for hand grip strength), poorer self-rated health (β = -0.122 for gait speed; β = -0.061 for hand grip strength), and non-White race (β: -0.098 for African American and β : -0.055 for Others for gait speed;) were independently associated with steeper declines in physical function.

Conclusion

Older adults with coexisting diabetes and heart disease experience accelerated declines in gait speed and hand grip strength compared with those with either condition alone or neither. These findings highlight the need for targeted functional monitoring and preventive interventions in this high-risk population, with particular attention given to sex, perceived health, and race-related disparities in physical aging.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12877-025-06390-x.

Background

Approximately 65% of older adults in the United States live with multimorbidity, and a rapidly increasing proportion exhibit cardiometabolic multimorbidity (CMM), i.e., the coexistence of two or more cardiometabolic conditions [1–7]. The increasingly frequent coexistence of cardiometabolic conditions likely stems from shared risk factors and organ system interactions [7–9]. For example, diabetes-induced vascular impairments contribute to the development of ischemic heart disease [10–13], while the presence of diabetes in individuals with heart disease accelerates the risk of worse clinical outcomes. [14, 15] Consequently, CMM substantially increases the risk of hospitalization, and at least one-third of these admissions are considered preventable. [4–6, 16, 17] Beyond these clinical outcomes, the combined pathophysiological burden of CMM may also accelerate declines in physical function, thereby compounding disability risk in older adults. Previous research has shown a direct association between physical function and risk of hospital admission in older adults. [18] However, the impact of CMM on objective measures of physical function in older adults has not yet been examined. Specifically, there is limited evidence on how the coexistence of diabetes and heart disease influences gait speed and hand grip strength, which are established physical biomarkers of health and function in older adults.

Gait speed and hand grip strength independently predict mobility, quality of life, cognitive function, frailty, and mortality in older adults [19–22]. While age-related declines in these physical biomarkers have been described in specific medical conditions such as diabetes [23], osteoarthritis [24, 25], and stroke [26], these trajectories have not been examined in CMM. Despite recognition that both gait speed and hand grip strength decline with age, the extent to which CMM, particularly the coexistence of diabetes and heart disease, impacts these declines remains unclear. Specifically, it is unknown whether the age-related declines in gait speed and hand grip strength are similar for those with CMM compared with those without. Therefore, this study aimed to make this comparison over five years among older adults across four groups: (1) diabetes only, (2) heart disease only, (3) CMM including both diabetes and heart disease, and (4) a control group with no heart disease or diabetes. A secondary aim was to determine the factors associated with a decline in gait speed and hand grip strength in older adults with CMM [27]. We hypothesize that the coexistence of diabetes and heart disease in CMM is associated with faster declines in gait speed and hand grip strength compared to what is typical of aging or in the presence of either condition alone.

Methods

Study design

This was a retrospective secondary longitudinal analysis of data from the National Health and Aging Trends Study (NHATS) linked to the Centers for Medicare and Medicaid Services (CMS) data from 2015 to 2019. Specifically, we linked the 2015 Master Beneficiary Summary File -base (MBSF-(base)) and the Master Beneficiary Summary File -Chronic Conditions file (MBSF-(CC)) from CMS to the NHATS data. The NHATS is an ongoing longitudinal study initiated by the National Institute on Aging to guide efforts to reduce disability, maximize functioning, and enhance the quality of life of older individuals. The NHATS conducts annual in-person interviews and physical examinations of a nationally representative sample of Medicare beneficiaries aged 65 and older in the United States (U.S).

Inclusion criteria

Participants who were 66 years and older and identified as community-dwelling during the 2015 NHATS interview were included in the analysis. We included participants with diabetes, heart disease, or both using ICD-10 codes from the MBSF-(CC) files. The MBSF-(CC) file uses ICD-10 diagnosis codes to determine diagnoses for participants enrolled only in the Medicare fee-for-service insurance plan. Therefore, we included only those enrolled in Medicare Fee-For-Service to ensure the accuracy of disease group classification. We excluded participants enrolled in Medicare Advantage Plans. Heart disease was defined as a history of one of the following conditions: ischemic heart disease, acute myocardial infarction, or heart failure. We created four distinct groups based on presence or absence of CMM: (1) diabetes only, (2) heart disease only, (3) CMM including both diabetes and heart disease, and (4) a control group with no heart disease or diabetes. The MBSF-(CC) file was also used to identify additional chronic conditions. The MBSF-(CC) records 27 chronic conditions via a predefined, validated algorithm based on specific diagnoses and/or procedure codes found on the institutional and non-institutional Medicare fee-for-service claims within the specified reference period [28]. Based on these conditions, we excluded anyone with a known neurological condition, such as stroke, Parkinson’s disease, Alzheimer’s disease, or dementia or other neurological conditions that could directly impact gait speed. Finally, we excluded anyone who did not have complete information on gait speed or hand grip strength in 2015.

Analytical sample

Out of the 8,334 participants who responded to the NHATS round 5 (2015) surveys, 7,070 were community-dwelling at the time of the interview, and 6,766 were enrolled in the Medicare fee-for-service plan. Additionally, 841 individuals with neurological conditions and 1574 with incomplete or missing survey data at baseline were excluded. The final baseline cohort with complete information from 2015 surveys had 4,351 individuals who were included in the primary analysis. Of these, participants who retained complete data at all time points from 2015 to 2019 (n = 2,450 for gait speed, and n = 2,284 for hand grip strength) constituted the complete cohort and were used in secondary stability analysis (Fig. 1).

Fig. 1.

Cohort selection

Measures

Gait speed and hand grip strength were the primary outcome variables for this study. In NHATS, participants completed two trials of a 3-meter course walk at their usual pace from a standing start. The time taken to walk the course was recorded. For analysis, gait speed in meters per second (m/s) was derived from the distance walked in meters and time in seconds, and an average of two trials was used. For hand grip strength, participants were asked to squeeze with their dominant hand on a Jamar Plus dynamometer at their maximal effort while seated with their arms at their sides and elbows bent at 90 degrees. The reading on the dynamometer was recorded over two trials. The average of two trials was normalized by body mass index (BMI) for analysis [29].

Covariates

Demographic data included age (in years), race (Caucasian, African American, and other), sex (male or female), and marital status (married or not married). Socioeconomic data included home ownership status (own or rent) and the number of people in the social network [30–34]; clinical variables included BMI (in kilograms per meter squared), smoking status (yes or no), self-rated health (excellent, good, or fair/poor), and functional comorbidity index (FCI). The FCI score was calculated by summing all chronic conditions identified in the index after excluding diabetes and heart disease [35]. Additionally, osteoporosis and arthritis (rheumatoid and osteoarthritis) were included as independent covariates rather than as part of the FCI index, as these conditions have previously been shown to have a direct impact on mobility and function [36–39].

Analysis

Adjustments for all complex sample design variables, including clustering, stratification, and weighting, were performed based on the NHATS weighting guidelines to ensure the generalizability of the results to the U.S. Medicare population [40]. To describe baseline characteristics across the four identified groups (diabetes only, heart disease only, CMM, and control), we reported means and standard errors for continuous variables and percentages and standard errors for categorical variables. Between-group comparisons were performed via analysis of variance (ANOVA) for continuous variables and Rao-Scott chi-square tests for categorical variables. The level of significance was set at 0.05. The Bonferroni adjustment was applied in the post-hoc analysis to compare each disease group with the control and CMM groups at a significance level of 0.01. The baseline cohort, which included all participants with complete information on gait speed and hand grip strength at baseline (n = 4,351), was used to plot changes in gait speed and hand grip strength over time across the five time points from 2015 to 2019, stratified by the four groups. Secondary stability analysis was also performed on the complete cohort, including all participants with complete information on gait speed (n = 2,450) and hand grip strength (n = 2,284) at all time points. Separate generalized estimation equation (GEE) models were estimated with gait speed and grip strength as the time-dependent outcome variables and the four disease groups as the primary predictors. To account for rate of change, time was modeled as a continuous variable in all GEE models to estimate linear trends in physical function.

Considering the potential confounding effects, we included demographics (age, sex, race/ethnicity), socioeconomic and behavioral status (marital status, social network, house ownership, smoking), and health and clinical status (BMI, overall health, FCI, osteoporosis, arthritis) as covariates in the GEE model to account for the dependency of repeated measurements. A three-step modeling process was conducted separately for gait speed and hand grip strength, including (1) an unadjusted model, (2) a reduced model with adjustments for demographics and socioeconomic and behavioral status, and (3) a full model with adjustments for all confounding factors.

Results

In the weighted sample, the control group was significantly younger (mean age 72.34 ± 0.14) than the disease groups. The CMM group had the greatest number of comorbidities compared to the control groups (4.76 ± 0.11 vs. 1.36 ± 0.07). A significantly greater proportion of individuals rated their health as fair or poor in the disease groups, specifically in the CMM groups (30.59%), than in the control group (11.14%). No baseline differences were noted in marital or home ownership status across the control and disease groups (Table 1). Figure 2 shows the average gait speed and hand grip strength over five years for all groups in both the baseline and complete cohorts. At baseline (in 2015), the CMM group exhibited the lowest gait speed (baseline cohort: 0.732 ± 0.011 m/s; complete cohort: 0.799 ± 0.017 m/s) and hand grip strength (baseline cohort: 0.904 ± 0.021 kg/BMI; complete cohort: and 0.935 ± 0.028 kg/BMI) compared with the control group (baseline cohort:0.872 ± 0.007 m/s for gait speed and 1.034 ± 0.010 kg/BMI for hand grip strength). This difference persisted throughout the five-year observation period from 2015 to 2019 for both cohorts (Fig. 2). In the complete cohort (Fig. 2, right panel), the CMM group exhibited the steepest declining trend in gait speed and hand grip strength (gait speed slope: −0.016, p < 0.001; hand grip strength slope: −0.015; p < 0.001) compared with that of the control group (Fig. 2, right panel).

Table 1.

Baseline characteristics in 2015 (Weighted N = 25,164,206)

Variables	Control	Diabetes Only	Heart Disease Only	CMM
Unweighted N	2,299	476	894	682
Weighted N	14,654,137	2,685,017	4,487,226	3,337,826
Age, years(a)*&¥†	72.34 (0.14)	72.90 (0.27)	75.88 (0.30)	75.42 (0.26)
Race/Ethnicity(b)*#&¥‡
White	93.02 (1.23)	75.92 (2.49)	86.84 (1.13)	77.99 (1.98)
African American	6.50 (0.53)	11.79 (1.38)	6.14 (0.65)	9.77 (1.09)
Other	10.48 (1.07)	12.29 (2.35)	7.02 (0.99)	12.24 (1.81)
Female(b)*&¥†	56.31 (1.22)	55.59 (2.86)	48.44 (1.67)	42.38 (2.30)
Married(b)	64.66 (1.39)	59.06 (2.45)	63.11 (1.72)	62.32 (2.24)
BMI, kg/m2 (a)*#¥‡	27.39 (0.13)	30.37 (0.38)	27.78 (0.19)	29.94 (0.26)
House ownership(b)
Own	82.27 (1.24)	78.67 (2.39)	82.99 (1.23)	77.68 (2.02)
Rent	17.73 (1.24)	21.33 (2.39)	17.01 (1.23)	22.32 (2.02)
Smoking(b)*¥	49.76 (1.45)	48.68 (2.90)	52.54 (1.39)	57.34 (2.23)
Social Network(a)*¥‡	2.29 (0.05)	2.23 (0.07)	2.29 (0.05)	2.08 (0.06)
Self-rated health(b)*#&¥†‡
Excellent/Good	88.86 (0.87)	81.79 (2.42)	81.47 (1.34)	69.41 (2.45)
Fair/Poor	11.14 (0.87)	18.21 (2.42)	18.53 (1.34)	30.59 (2.45)
FCI score(a)*#&¥†‡	1.36 (0.07)	3.02 (0.13)	3.78 (0.08)	4.76 (0.11)
Osteoporosis(b)*&¥	8.61 (0.79)	11.12 (1.67)	19.04 (1.25)	16.93 (1.45)
Arthritis(b)*#&¥†	20.63 (1.37)	48.33 (2.40)	56.01 (2.22)	61.45 (2.26)
Gait Speed, m/s*#&¥	0.87 (0.01)	0.79 (0.01)	0.80 (0.01)	0.73 (0.01)
Hand Grip Strength, kg/BMI*#&¥‡	1.03 (0.01)	0.91 (0.02)	0.98 (0.02)	0.90 (0.02)

The distribution of baseline characteristics was compared among four groups using the Analysis of Variance (ANOVA) for continuous variables, and Rao-Scott Chi-square tests for categorical variables

BMI Body Mass Index, FCI Functional Comorbidity Index score 

^(a) Continuous variables were presented with Mean (SE)

^(b) Categorical variables were presented with percent (SE)

^*p < 0.05 for the comparison among four groups

^#p < 0.01 for the comparison between diabetes only and control groups

^&p < 0.01 for the comparison between heart disease only and control groups

^¥p < 0.01 for the comparison between diabetes and heart disease and control groups

^†p < 0.01 for the comparison between diabetes only and diabetes and heart disease groups

^‡p < 0.01 for the comparison between heart disease only and diabetes and heart disease groups

Fig. 2.

Average of Gait Speed and Hand Grip Strength During 2015 and 2019 Stratified by Disease Group

GEE model for gait speed

Compared to the control group, all disease groups exhibited significant declines in gait speed in the unadjusted model (Model 1 in Table 2), with the CMM group demonstrating approximately twice the decline observed in the diabetes only and heart disease only groups (CMM: β [SE]: −0.145 [0.011]; diabetes: −0.072 [0.012]; heart disease: −0.077 [0.009], all p < 0.01). After adjusting for demographic and clinical characteristics, only the CMM group had a significant gait speed decline (β [SE]: −0.034 [0.010], p < 0.01). Both demographic and clinical characteristics were significant predictors of gait speed decline, with pronounced reductions observed among participants reporting poor self-rated health (β [SE]: −0.122 [0.009], p < 0.01), those identifying as non-White (β [SE]: −0.098 [0.009] for African American; β [SE]: −0.055 [0.014] for Others), females (β [SE]: −0.049 [0.007]) (Supplementary Table S1). No interaction effect between race and sex with disease groups was found.

Table 2.

Association between cardiac metabolic morbidity and physical function

Variables	Model 1 β (SE)	Model 2 β (SE)	Model 3 β (SE)
Gait Speed, m/sec
CMM Group (ref: Control)
Diabetes Only	−0.072 (0.012)**	−0.054 (0.011)**	−0.007 (0.010)
Heart Disease Only	−0.077 (0.009)**	−0.038 (0.009)**	−0.005 (0.009)
Diabetes/Heart Disease	−0.145 (0.011)**	−0.103 (0.010)**	−0.034 (0.010)**
Hand Grip Strength, kg/BMI
CMM Group (ref: Control)
Diabetes Only	−0.126 (0.022)**	−0.118 (0.017)**	−0.019 (0.013)
Heart Disease Only	−0.055 (0.018)**	−0.058 (0.014)**	−0.015 (0.013)
Diabetes/Heart Disease	−0.134 (0.019)**	−0.163 (0.016)**	−0.048 (0.015)**

*p < 0.05, **p < 0.01; SE Standard Error, FCI Functional Comorbidity Index score

Model 1: Outcome = CMM Group + Year

Model 2: Outcome = CMM Group + Year + Age + Sex + Race/Ethnicity + Marital Status + Social Networks + House ownership + Smoking

Model 3: Outcome = CMM Group + Year + Age + Sex + Race/Ethnicity + Marital Status + Social Networks + House ownership + Smoking + BMI + Self-rated health + Chronic Conditions + Osteoporosis + Arthritis

GEE model for hand grip strength

Consistent with the findings on gait speed, we observed similar effects on hand grip strength (Table 2). In the fully adjusted model, CMM alone was significantly associated with hand grip strength decline (β [SE] = −0.048 [0.015], p < 0.01). Additionally, being female (β [SE]: −0.460 [0.010], p < 0.01) and reporting poor self-perceived health (β [SE]: −0.061 [0.013], p < 0.01) were strongly associated with reduced hand grip strength. (Supplementary Table S1). A significant interaction effect between sex and disease group was observed (p < 0.001), and sex-stratified analyses was conducted to estimate the disease effect separately. As shown in Supplementary Table S2, among males, the CMM group was associated with a significantly greater decline in hand grip strength (β [SE]: −0.070 [0.025], p < 0.05). In contrast, this association was not significant among females.

Discussion

Our findings highlight significant impairments in objective biomarkers of physical function in a nationally representative, longitudinal cohort of community-dwelling older adults with CMM. We found that older adults with CMM demonstrated significantly lower gait speeds and hand grip strengths than did those without CMM and experienced the most pronounced declines in these biomarkers over five years. These findings confirm our original hypothesis that the disease interaction in CMM may accelerate the age-related declines in gait speed and hand grip strength.

Gait speeds of > 0.8 m/s are essential for independent community ambulation [41]. The CMM group in this study exhibited a significantly lower gait speed in 2015 as compared to the control group, and remained consistently lower than all other groups over five years, with statistically significant but modest decreases (0.799 to 0.731 m/s) in the complete cohort. Within the baseline cohort, gait speeds in the CMM group were also significantly lower than those observed in the single disease groups over time. Although the complete cohort did not demonstrate statistically significant differences between CMM and disease groups, the CMM group exhibited a consistent trend toward slower gait speeds over the five-year period. Moreover, the change in gait speed over time was statistically significant and exceeded the smallest meaningful change threshold of > 0.05 m/s, indicating that these declines were clinically meaningful [42, 43]. In community-dwelling adults, a change in gait speed of 0.04–0.06 m/s is considered a small meaningful change, and a change of 0.1 m/s is considered a substantial meaningful change [42, 43].

Our results are consistent with and expand upon previous literature. Findings from the Swedish National Study on Aging and Care in Kungsholmen in older adults with cardiovascular multimorbidity, demonstrated even greater declines in walking speed of up to 0.7 m/s over a 9-year observation period compared to the 5-year observation period and relatively younger cohort in our study. [44] Similarly, a large UK Biobank longitudinal study with a median follow-up of 13 years found that sarcopenia measured by hand grip strength and gait speed significantly influenced the transition from a healthy state to single cardiometabolic disease, from single cardiometabolic disease to CMM, and ultimately to death. [45]

Several potential mechanisms may explain the accelerated gait speed decline in CMM compared with isolated conditions. Gait involves an interaction of multiple neuromotor and physiological processes, including motor control, muscular and cardiovascular endurance, muscle strength, balance, and sensory and perceptual functions [41]. Heart disease amplifies age-related cardiovascular changes, resulting in deconditioning and reduced functional mobility [46]. The presence of diabetes imposes additional musculoskeletal and neuromotor stress through insulin resistance related impairments in muscle growth and repair and sensorimotor deficits in joint proprioception [15, 46, 47]. Together, these factors accelerate functional mobility loss, including gait speed.

Compared with the control group, the CMM group also demonstrated significant decreases in hand grip strength throughout the five-year follow-up. Notably, the hand grip strength of the diabetes-only group also decreased, similar to that of the CMM group. In contrast, values in the heart disease-only group remained stable and comparable to those of the control group over the same period. These observations suggest that the presence of diabetes primarily drove the declines in hand grip strength in the CMM group. This finding is consistent with previous literature suggesting that diabetes is more strongly linked to weakness, which first manifests in distal muscle groups [48]. Distal symmetric polyneuropathy in diabetes impairs motor nerves leading to muscle atrophy and weakness [48]. While much of the literature on diabetes has focused on distal lower extremity muscle weakness, specifically in the ankle and intrinsic muscles of the foot, select research indicates that diabetes also significantly affects distal hand muscles, reducing normalized hand grip and pinch strengths [49, 50]. Additional contributors to the loss of distal muscle strength include intramuscular fat infiltration, low-grade systemic inflammation, and microvascular compromise [51, 52]. In contrast, heart disease showed no significant impact on hand grip strength compared with controls. Since small muscles involved in isometric hand grip contraction pose significantly lower demands compared to large lower extremity dynamic muscle contraction, changes in grip strength may not be apparent early on in isolated heart disease [53, 54]. However, it was noteworthy that although diabetes demonstrated consistent and comparable declines in hand grip strength with those in the CMM group, the effect of diabetes on hand grip strength was minimized in the adjusted model, whereas CMM continued to demonstrate significant associations with hand grip strength decline (Table 2). This finding suggests that diabetes likely interacted with other clinical variables in the model. We did not include an examination of different interactions, as these interactions were not planned a priori. However, our findings warrant future explorations of these interactions to identify other factors contributing to hand grip strength declines in CMM.

Our adjusted models further highlight the relative contributions of various factors that affect the decline in gait speed and hand grip strength. While CMM was a significant predictor of decline in gait speed and hand grip strength in the fully adjusted model, sex, race, and perceived poor overall health had stronger associations with these outcomes. These findings suggest that the presence of CMM may have a greater impact on declines in gait speed and hand grip strength among women in nonwhite racial groups. Future research is needed to explore the mechanisms of preferential sex- and race-specific rates of decline in gait speed and hand grip strength in individuals with CMM.

Similar findings on sex, race, and overall perceived health have been reported previously. Our observations are consistent with findings from the Australian Longitudinal Study on Women’s Health, which reported that older women with cardiovascular multimorbidity experienced the greatest decline in functional independence over time. [55] Similarly, a higher prevalence of CMM and associated comorbidities was observed among African Americans, which may also contribute to diminished physical function. [56] Additionally, older adults with higher perceived psychological and social well-being at baseline have shown to exhibit greater independence and a slower rate of physical function decline in a 12-year Swedish longitudinal study. [44] The findings of our study in light of previous literature warrant further investigation into drivers of functional decline in this population with CMM.

The results of this study carry significant implications for clinical practice and preventive care strategies. Our findings suggest that individuals with CMM are at high risk for functional loss and should be followed closely for decline via these non-invasive, simple, and cost-effective physical biomarkers that are easily integrated into clinical settings. Screening for slow gait speed and hand grip strength for individuals with CMM could enable timely implementation of preventive strategies, including structured physical activity programs, nutritional support, and optimized chronic disease management to mitigate functional decline [57–59]. Additionally, prioritizing the inclusion of self-perceived health status within health screening measures, particularly for women and nonwhite racial groups with CMM, might facilitate early identification of those at most significant risk of decline, potentially capturing the broader physical and psychological stressors that contribute to mobility loss.

Despite the strengths of a large, well-characterized cohort and longitudinal design, this study has some limitations. Given the large amount of missing data (> 50%) at subsequent time points, we could not perform multiple imputations on the baseline cohort. Instead, we conducted a stability analysis using a complete cohort. Although our secondary stability analyses of the complete cohort revealed similar trends, the differences were not statistically significant. To assess potential sample bias introduced by attrition, we compared baseline characteristics between participants who remained in the study across all time points (complete cohort) and those who dropped out. Compared to those who remained, participants who dropped out were more likely to be older, non-White, single with fewer social networks and poorer self-rated health, and a greater number of comorbidities (Supplementary Table S2). This may be due to attrition bias, as healthier individuals are more likely to remain in longitudinal studies. Additionally, the observational nature of this study precluded the ability to draw causal inferences. Duration of disease may have impacted gait speed and hand grip strength differently; however, the secondary nature of this analysis limited the ability to ascertain duration of diabetes or heart disease at baseline in the cohort. We normalized hand grip strength by BMI based on data availability and established prior studies. [60–62] While this approach helps account for body size variation, it may misclassify strength in individuals with sarcopenic obesity, where BMI may reflect fat mass rather than lean mass. NHATS does not include body composition measures such as fat-free mass or appendicular lean mass, which prevented the use of alternative normalization methods. Future studies with more detailed body composition data may allow for more precise evaluation of muscle quality. All covariates, including health status and BMI, were measured at baseline and may have changed during the follow-up period. This may have led to time-varying effects of these covariates not captured adequately in our models. Although individuals with Alzheimer’s disease or diagnosed dementia were excluded from the analytic sample, we did not include cognition as a covariate in our models. Future research should explore how varying degrees of cognitive function may influence functional trajectories in older adults with CMM. Finally, we were unable to include some variables, such as income, due to large amounts of missingness in the data. While several potential factors in the pathophysiology of these two disease conditions could interplay in contributing to accelerated functional decline, the specific mechanisms underlying this heightened decline are unclear. Future prospective studies must examine the mechanisms linking CMM with a decline in physical biomarkers. Finally, while gait speed and hand grip strength cut-off values have been established to predict the risk of important clinical outcomes such as hospital admissions and mortality in older adults, future research to establish these cut-offs for CMM might be needed.

Conclusions

In conclusion, our findings provide evidence that CMM, particularly the co-occurrence of diabetes and heart disease, is significantly associated with a decline in gait speed and hand grip strength in older adults over time. The stronger associations observed with perceived health status and sex highlight the multifaceted nature of functional aging and the need for comprehensive, individualized approaches to maintaining mobility and muscle strength in this population.

Abbreviations

CMM

Cardiometabolic multimorbidity

NHATS

National Health and Aging Trends Study

CMS

Center of Medicare and Medicaid Services

MBSF-(base)

Master Beneficiary Summary File (MBSF)-base

MBSF-(CC)

MBSF-Chronic Conditions file

BMI

Body mass index

GEE

Generalized Estimation Equation

FCI

Functional Comorbidity Index

ANOVA

Analysis of variance

Authors’ contributions

SG and AK conceptualized the study, research questions, concept, and design, and supervised the project. DK created the analytical cohort and conducted all preliminary analyses. LC analyzed and interpreted the data regarding the trajectories and modeling. AK provided methodological guidance for constructing the models and interpretation of results. JK and JAT provided major contributions to the writing of the manuscript. All authors read and approved the final manuscript.

Funding

This research was supported by an NIH grant: R15AG070730-01A1.

Data availability

The data that support the findings of this study are available from National Health and Aging Trends Study (NHATS) and the Center of Medicare and Medicaid Services (CMS) but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available.

Declarations

Ethics approval and consent to participate

This study was deemed an expedited review and approved by the Mass General Brigham IRB, protocol # 2022P00303. This study adhered to the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.