Muscle weakness is a prognostic indicator of disability and chronic disease multimorbidity.

Abstract

Background:

The objective of this study was to use nationally-representative data on Americans greater than 50 years of age to determine the association between grip strength and inflammation as independent predictors of incident disability, chronic multimorbidity and dementia.

Methods:

Middle age and older adults (n=12,618) from the 2006–2008 waves of the Health and Retirement Study with 8-years of follow-up were included. Longitudinal modeling was performed to examine the association between baseline grip strength (normalized to body mass: NGS) and high sensitivity C-reactive protein (hs-CRP) (≥3.0 mg/L) with incident physical disabilities (i.e., ≥2 limitations to activities of daily living), chronic multimorbidity (≥2 of chronic conditions), and dementia.

Results:

The odds of incident disability were 1.25 (95% CI: 1.20–1.30) and 1.31 (95% CI: 1.26–1.36) for men and women respectively, for each 0.05-unit lower NGS. The odds of incident chronic multimorbidity were 1.14 (95% CI: 1.08–1.20) and 1.14 (95% CI: 1.07–1.21) for men and women respectively for each 0.05-unit lower NGS. The odds of incident dementia were 1.10 for men (95% CI: 1.02–1.20) for each 0.05-unit lower NGS, but there was no significant association for women. Elevated hs-CRP was only associated with chronic multimorbidity among men (OR=1.29; 95% CI: 1.00–1.73) and women (OR=1.60; 95% CI: 1.26–2.02).

Conclusions:

Our findings indicate a robust inverse association between NGS and disability and chronic, multimorbidity in older men and women, and dementia in men. Elevated hs-CRP was only associated with chronic multimorbidity in men and women. Healthcare providers should implement measures of grip strength in routine health assessments and discuss the potential dangers of weakness as well as interventions to improve strength with their patients.

Introduction

The expansion of the aging population has led to a diversification of noncommunicable disease morbidity, including increased prevalence of aging-related mobility impairments and a substantial reduction in the number of healthy, nondisabled years (1–3). As of 2020, there were over 43 million individuals in the United States with disabilities-representing approximately 13.2% of the civilian population (4). The healthcare burden associated with disability is nearly twice that of individuals without disabilities, and is likely owing to a higher risk for chronic disease in these individuals (5). Early screening and continued health promotion efforts are vital to reduce the escalating burden associated with multimorbidity and disability in the growing aging population. Of relevance to both, there is a large body of evidence linking muscular weakness, as determined by low grip strength, to a host of negative aging-related health outcomes including diabetes (6, 7), disability (8–11), cognitive decline (including Alzheimer’s disease (12, 13)) (14, 15), and early all-cause mortality (8, 16–20). Given these links, grip strength has been labeled a “biomarker of aging” (21); and yet, this metric provides no biological plausibility linking weakness with chronic disease, physical and cognitive impairments, or specific causes of mortality.

There is also a robust consensus that the low-grade, chronic inflammation in aging (“inflammageing”) is a significant risk factor for both morbidity and mortality in elderly people (22), and may represent a strong mechanism linking age-related increases in adiposity and metabolic dysregulation with sarcopenia and muscle weakness (23, 24). Indeed, inflammageing has been linked to cardiovascular diseases, type 2 diabetes mellitus, chronic kidney disease, cancer, depression, dementia, osteoporosis, disability, sarcopenia, and anemia (25, 26). The main driver of chronic inflammation is likely from the intraabdominal depot, as visceral adiposity produces pro-inflammatory (e.g., high-sensitivity C-reactive protein (hsCRP) and chemotactic compounds causing macrophage infiltration and a senescence phenotype (27). It seems highly plausible that the association between weakness and age-related disease may be mediated through inflammatory pathways; and yet, what remains to be determined is the longitudinal association between muscular weakness, inflammageing, and altered ageing trajectories leading to disease and disability in a population-representative sample. The purpose of this study was to examine the association between grip strength and inflammation as independent predictors of incident disability, chronic multimorbidity and dementia. We also sought to examine the extent to which the association between grip strength and disability, chronic multimorbidity and dementia was mediated by high hs-CRP. We hypothesized that grip strength would be significantly and inversely associated with incident physical disability, chronic disease multimorbidity, and dementia. Additionally, we hypothesized that chronic inflammation, as demonstrated by elevated serum levels of hsCRP, would also be independently associated with each outcome, and explain some of the proportion of the association attributed to the exposure of grip strength.

Methods

Cohort

We used a longitudinal panel study (Health and Retirement Study [HRS]) to determine the longitudinal association between muscle strength, inflammation, and incident disability, cognitive deficit, and chronic disease multimorbidity in middle age and older adults. The HRS is a prospective epidemiological resource gathering extensive questionnaires, physical and cognitive measures, and biological samples in a cohort of more than 43,000 participants. HRS is a multistage area probability survey of non-institutionalized, community dwelling Americans aged 51 years and older. Study details have been previously described (28), and all questionnaires are available on the HRS website (http://hrsonline.isr.umich.edu). Briefly, the HRS is the longest running longitudinal study of older adults in the United States, with consistent response rates of ~85%. The HRS follows respondents longitudinally until death, and new cohorts have been enrolled since the 1992 baseline interviews, in order to maintain the population representativeness of the study sample (29). Starting in 2006, a random one-half sample of HRS participants was selected for an enhanced face-to-face interview that included physical and biological measurements (e.g., hand grip strength, blood, and saliva collection), and the other half completed only the core interview. The other random one-half completed the enhanced face-to-face interview that included physical and biological measurements in 2008. These half-sample alternate waves are done so that longitudinal information from the enhanced face-to-face interview is available every 4 years at the individual level, and the expanded content is available every wave on a nationally representative half-sample of the longitudinal panel. This rotating design has continued in all biennial HRS waves since 2006, resulting in follow-up data on the physical, biomarker, and cognitive health measures every 4 years for the subsamples. Our study sample of n=12,618 included all HRS participants aged 51 or older (66.8 ± 8.9 years) interviewed in 2006/2008, when physical measure and biomarker collection began in the HRS, and with an 8-year follow-up assessment in 2014/2016. The age, sex and racial breakdown of our sample is consistent with that of Americans aged 50 years and older in the United States, based on U.S. census data (30).

Exposures:

Grip strength:

In HRS, hand grip strength was assessed using a Smedley spring-type hand dynamometer (Scandidact, Odder, Denmark). Participants were instructed to squeeze the device with maximal effort and then let go. Grip strength assessments were administered while participants were standing with their arm at their side, and with the elbow flexed at a 90-degree angle. After one practice trial, two measurements were taken with each hand, alternating hands. The maximum measurement from the four trials were considered maximal grip strength capacity. Normalized grip strength (NGS) as grip strength per body mass (i.e., $\left(\frac{\mathit{\text{Grip Strength }}\left(kg\right)}{\mathit{\text{Body Mass }}\left(kg\right)}\right)$) was examined as a continuous predictor.

Chronic Inflammation:

Peripheral blood samples were obtained in HRS from participants by cleaning a finger with an alcohol prep pad and pricking the finger with a lancet. Samples were collected on a dry blood spot card, mailed to a lab, and assayed for high sensitivity C-reactive protein (hsCRP) using a standard enzyme-linked immunosorbent assay (ELISA). The hsCRP assay had a lower limit of detection of 0.04 mg/L, with a within-assay coefficient of variability (CV) of 8% and an inter-assay CV of 11%. To exclude acute inflammation, calculated whole-blood equivalent hsCRP values were used to exclude participants with hsCRP values greater than 10 mg/L, as previously described in detail (31). Briefly, to calculate whole-blood equivalent hsCRP values, the blood spot hsCRP values from HRS were regressed onto the whole-blood equivalent hsCRP value from The National Health and Nutrition Examination Survey according to HRS study protocol (31). Elevated hsCRP was defined as a value greater than 3.0 mg/L (32).

Outcomes:

Physical Disabilities:

The HRS assesses physical disability status using self-reported difficulty with six self-care activities of daily living (ADLs): eating, bathing, dressing, transferring, toileting and walking across a room. For each activity, difficulty was recorded as present (i.e., difficulty with activity or cannot/does not do) or absent (no difficulty). We used an ADL summary score for each individual based on the sum of all reported difficulties across all 6 activities. Presence of 2 or more limitations to any ADL (i.e., ADL summary score ≥2) was considered a physical disability. Three ADL summary scores were calculated for each individual—one at baseline (2006/2008), and then one at each follow-up (every 2 years, 2010–2016).

Medical Conditions and Chronic Disease Multimorbidity:

The HRS assesses the presence of medical conditions using participants’ responses to questions about whether a doctor has “…ever told you that you had [condition]?” In addition to the self-reported medical conditions, we defined existing conditions on the basis of direct physical measurement or from abnormal biological measures. Self-reported or direct measures of the following chronic medical diseases or chronic conditions included arthritis, cancer, chronic kidney disease (measured by cystatin C [cystatin C > 1.55 mg/L]), depression, diabetes heart disease, hypertension (based on direct measurement or self-reported use of prescription medications), lung disease, and stroke. Affirmative responses were summed to create a variable that represented the total number of medical conditions (0 to 10). Multimorbidity was defined as the presence of at least 2 chronic conditions among the list of the aforementioned self-reported chronic diseases or examination/laboratory abnormalities. These conditions have been selected in accordance with guidance from the literature pertaining to multimorbidity in older adults and adults with disabilities (33–35), and availability and reliability of the diagnoses within the HRS data.

Cognitive Function and Category Definitions:

The HRS assesses cognitive function in self-respondents with a range of tests adapted from the Telephone Interview for Cognitive Status (TICS). Based on previous work with HRS (36, 37), we used a 27-point cognitive scale that includes an immediate and delayed 10-noun free recall test, a serial 7 subtraction test, and a backward count from 20 test (see: http://hrsonline.isr.umich.edu/sitedocs/userg/dr-006.pdf). Cutpoints for normal, cognitive impairment—no dementia (CIND), and dementia were validated against the prevalence of CIND and dementia in the Aging, Demographics, and Memory Study (ADAMS), an HRS substudy of Alzheimer’s disease and dementia that uses a 3-to 4-hour in-home neuropsychological and clinical assessment as well as expert clinician adjudication to obtain a gold-standard diagnosis of CIND or dementia (36, 38). Respondents who scored from 0 to 6 on the 27-point scale were classified as having dementia, 7 to 11 as having CIND, and 12 to 27 as normal. We used cognitive assessment data from the HRS Cross-Wave Imputation of Cognitive Functioning Measures (1992–2016, Final V6.0) to correct for item-level non-response among self-respondents. On average, 4.5% of respondents had at least one score missing and thus imputed data from each wave across the years included in these analyses. For respondents represented by a proxy, an 11-point scale were calculated using the proxy’s assessment of the respondent’s memory ranging from excellent to poor (score, 0–4), the proxy’s assessment of whether the respondent had limitations in 5 instrumental activities of daily living (IADLs) (managing money, taking medication, preparing hot meals, using phones, and shopping for groceries; score, 0–5), and the survey interviewer’s assessment of whether the respondent had difficulty completing the interview because of a cognitive limitation (a score of 0–2 indicating, none, some, and prevents completion). Using this information, respondents with high scores (6–11) were classified as having dementia, and those with mid-range scores (3–5) as having CIND (36). Inclusion of proxy respondents, where possible, was critical to represent the most complete range of cognitive function in the population. Conversion to proxy is related to cognitive impairment; those who are cognitively impaired are less likely to participate in these studies (39).

Covariates:

The following sociodemographic measures were included in the primary analytic approach as independent covariates: age, self-reported race/ethnicity (White, Black, Hispanic, other), sex, marital status, education (<high school graduate, high school graduate, some college, college graduate, and ≥college graduate), net worth (quartiles), history of smoking (never vs. ever smoked), and body-mass index (BMI) (derived from height and weight). Standard categories were applied to determine if each participant was obese (≥ 30 kg/m²).

Data Analysis:

Descriptive statistics were used to explore the distribution, central tendency, and variation of each measurement, with an emphasis on graphical methods such as histograms, scatterplots, and boxplots. Bivariate analyses were conducted using Chi-Square tests or Fisher’s Exact test in the case of small sample size for categorical variables or ANOVA or nonparametric tests for continuous variables. Incident physical disability, chronic disease multimorbidity, and dementia were each independently assessed after removing individuals with prevalent physical disability, chronic disease multimorbidity, and dementia at baseline. Thereafter, three separate logistic regression models were used to determine the independent odds for incident physical disability, and chronic disease multimorbidity, and dementia with baseline strength and elevated hsCRP (>3.0 mg/L) as the exposure variables, adjusted for age, sex, and sociodemographic variables. As a high strength capacity is indicative of decreased risk, the inverse of absolute and NGS (modeled per 0.05 units for interpretation of clinical significance) was used in the models.

In addition, because inflammation is a strong risk factor for age-related weakness and chronic diseases, we assessed if and the extent to which the association between NGS and the three dependent variables (physical disability, chronic disease multimorbidity, and dementia) was mediated by high hs-CRP using a counterfactual approach. We estimated (1) the direct association of NGS on all three outcomes independent of high hs-CRP (controlled direct effect [CDE]), (2) the indirect association of NGS on all three outcomes mediated through hs-CRP (natural indirect effect [NIE]), and (3) the proportion of the total NGS association on all three outcomes mediated through hs-CRP, using Valeri and VanderWeele’s formulas (40), and the proc causalmed procedure. Each model differed in sample size due to the removal of prevalent (i.e., at baseline) physical disability, chronic disease multimorbidity, and dementia. All analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC), and for all analyses, p < 0.05 was considered statistically significant.

Results

Baseline (wave 1) descriptive and health data are presented as means, standard errors, and percentages for men and women in Table 1. At baseline, women had a higher prevalence of obesity (28.4% vs 26.8%), high hsCRP (41.9% vs. 29.9%), and functional disability (6.0% vs. 3.6%) than men, respectively (all p<0.001). Incidence of chronic disease multimorbidity (58.1% vs. 54.0%) were higher in men than women (p<0.05). There was no significant difference between men and women in incidence of dementia (2.2% vs. 2.6%). At both baseline and 8-year follow-up, men were stronger than women, in terms of both absolute and NGS capacity. For both men and women, grip strength showed a significant decline from baseline (men: 0.49 ± 0.11; women: 0.36 ± 0.10) to wave 3 (men: 0.43 ± 0.11; women: 0.32 ± 0.01) (all p<0.001).

Table 1.

Demographic and health characteristics of the study population by sex at wave 1 (2006/2008).

	Men	Women
	n=4,966	n=7,652
Age, years	66.1	65.8
Body Mass Index (BMI), kg/m2	28.4	28.4
Obesity (BMI >30), %	26.8	28.5
Level of Education
Less than H.S., %	20.5	21.2
High School Graduate, %	29.9	35.3
Some College, %	20.8	23.2
College Graduate, %	13.6	10.1
Post-College Graduate, %	15.2	10.3
Married, %	78.2	57.5
Race
White, %	81.6	78.9
Black, %	12.5	15.4
Unknown/Other, %	5.9	5.7
Ethnicity
Hispanic, %	10.1	10.2
Net Worth
Quartile 1, %	15.5	20.8
Quartile 2, %	22.7	24.6
Quartile 3, %	28.1	27.1
Quartile 4, %	33.6	27.5
Grip Strength, kg	41.8	25.6
NGS (relative to body mass)	0.49	0.36
hsCRP, mg/L	3.6	4.6
High hsCRP, %	29.9	41.9
Dementia, %	2.2	2.6
Chronic Multimorbidity, %	54.0	58.1
Functional Disability, %	3.6	6.0

NGS, normalized grip strength

From a total possible cohort of n=12,618 there were complete data on (1) n=7,739 for modeling the association between baseline NGS and incident physical disability (prevalent disability at baseline was 5%); (2) n=4,926 for modeling the association between baseline NGS and incident chronic disease multimorbidity (prevalent chronic disease multimorbidity at baseline was 54.9%); and (3) n=7,606 for modeling the association between baseline NGS and incident dementia (prevalent dementia at baseline was 2.3%).

Figure 1 is a forest plot that visualize the decrease in odds of physical disability, chronic disease multimorbidity, and dementia for men and women.

Figure 1.

Forest plot shows the odds of lower NGS (i.e., for every 0.05 lower NGS) on incident physical disability, chronic disease multimorbidity, and dementia for men and women, after adjusting for all sociodemographic variables and elevated hs-CRP.

Physical Disability

There was a significant and independent association between NGS and incident physical disability in both men and women (Table 2). Specifically, the odds ratio (OR) of incident physical disability for every 0.05 lower NGS was 1.20 (95%CI: 1.20–1.30) in men and 1.31 (95%CI: 1.26–1.36) in women. There was a significant and independent association between high hsCRP (>3.0 mg/L) and incident physical disability in women (OR: 1.65; 95%CI: 1.27–2.16) only. Mediation analyses demonstrated some minimal evidence of causal mediation for both men and women. Specifically, among men, (1) the OR for the controlled direct effect was 1.25 (95%CI: 1.19–1.31), (2) the OR for the natural indirect effect was 1.01 (95%CI: 1.00–1.02), and (3) the proportion of the total NGS association on physical disabilities mediated through hs-CRP was 5.5% (p=0.02). Similarly, among women, mediation analyses demonstrated that (1) the OR for the controlled direct effect was 1.30 (95%CI: 1.24–1.36), (2) the OR for the natural indirect effect was 1.02 (95%CI: 1.01–1.03), and (3) the proportion of the total NGS association on physical disabilities mediated through hs-CRP was 5.7% (p<0.001).

Table 2.

Multiple logistic regression models for independent predictors of incident physical disability status (≥2 limitations for ADLs) in men and women.

Model Predictor(s)	Men		Women
	Odds Ratio	95% CI	Odds Ratio	95% CI
Age Category (Reference: <65)
Age 65–79	2.19	1.57, 3.16	1.91	1.48, 2.50
Age >80	5.98	3.80, 9.41	9.03	6.44, 12.67
Marital Status (Reference: Not Married)	0.74	0.54, 1.02	0.85	0.68, 1.07
Smoking Status (Reference: Non-Smoker)	1.67	1.12, 2.49	1.55	1.13, 2.12
Race/ethnicity (Ref: Non-Hispanic White)
Non-Hispanic Black	0.77	0.49, 1.19	1.76	1.32, 2.35
Hispanic or Mexican American	0.95	0.58, 1.54	1.73	1.22, 2.45
Other Race or Multi-Racial	0.98	0.50, 1.89	1.13	0.68, 1.87
Education Level (Reference: <HS Graduate)
High School Graduate	0.57	0.39, 0.82	0.71	0.54, 0.93
Some College	0.81	0.55, 1.21	0.71	0.51, 0.97
College Graduate	0.39	0.23, 0.67	0.54	0.34, 0.88
>College Graduate	0.50	0.31, 0.83	0.69	0.44, 1.09
Net Worth Quartiles (Reference: Quartile 1)
Quartile 2	0.74	0.50, 1.10	0.73	0.55, 0.97
Quartile 3	0.63	0.42, 0.96	0.52	0.38, 0.71
Quartile 4	0.53	0.34, 0.82	0.43	0.30, 0.62
Elevated hs-CRP	0.91	0.64, 1.28	1.65	1.27, 2.16
Normalized Grip Strength*	1.25	1.20, 1.30	1.31	1.26, 1.36

^*OR and 95%CI per each 0.05 unit lower

Chronic Disease Multimorbidity

There was a significant and independent association between NGS and incident chronic disease multimorbidity in both men and women (Table 3). The OR of incident chronic disease multimorbidity was 1.14 (95%CI: 1.08–1.20) for men and 1.14 (95%CI: 1.07–1.21) for women, for every 0.05 lower NGS. There was a significant and independent association between high hsCRP (>3.0 mg/L) and incident chronic disease multimorbidity in women (OR: 1.60; 95%CI: 1.26–2.02) only. Mediation analyses demonstrated some minimal evidence of causal mediation among women only. Specifically, (1) the OR for the controlled direct effect was 1.16 (95%CI: 1.10–1.24), (2) the OR for the natural indirect effect was 1.03 (95%CI: 1.02–1.04), and (3) the proportion of the total NGS association on chronic disease multimorbidity mediated through hs-CRP was 21.3% (p=0.01).

Table 3.

Multiple logistic regression models for independent predictors of incident chronic disease multimorbidity (≥2 chronic diseases) in men and women.

Model Predictor(s)	Men		Women
	Odds Ratio	95% CL	Odds Ratio	95% CL
Age Category (Reference: <65)
Age 65–79	1.67	1.32, 2.10	1.55	1.27, 1.88
Age >80	1.30	0.78, 2.17	1.97	1.32, 2.93
Marital Status (Reference: Not Married)	1.00	0.76, 1.31	0.87	0.71, 1.07
Smoking Status (Reference: Non-Smoker)	1.04	0.75, 1.43	1.33	1.01, 1.75
Race/ethnicity (Ref: Non-Hispanic White)
Non-Hispanic Black	0.79	0.54, 1.17	1.29	0.95, 1.74
Hispanic or Mexican American	0.81	0.55, 1.21	1.01	0.72, 1.40
Other Race or Multi-Racial	1.09	0.68, 1.74	0.93	0.62, 1.41
Education Level (Reference: <HS Graduate)
High School Graduate	0.90	0.63, 1.30	1.06	0.79, 1.42
Some College	0.96	0.65, 1.43	1.09	0.80, 1.50
College Graduate	0.91	0.60, 1.38	1.06	0.73, 1.54
>College Graduate	0.87	0.57, 1.32	1.09	0.75, 1.58
Net Worth Quartiles (Reference: Quartile 1)
Quartile 2	1.10	0.75, 1.62	0.97	0.72, 1.31
Quartile 3	1.10	0.76, 1.60	0.73	0.54, 0.99
Quartile 4	0.82	0.56, 1.21	0.70	0.51, 0.96
Elevated hs-CRP	1.29	1.00, 1.73	1.60	1.26, 2.02
Normalized Grip Strength*	1.14	1.08, 1.20	1.14	1.07, 1.21

^*OR and 95%CL per each 0.05 unit lower

Dementia

There was a significant and independent association between NGS and incident dementia in men only (OR: 1.10; 95%CI: 1.02–1.20), for every 0.05 lower NGS (Table 4), and there was no evidence of causal mediation between NGS and incident dementia after accounting for the effect of high hsCRP.

Table 4.

Multiple logistic regression models for independent predictors of incident dementia in men and women.

Model Predictor(s)	Men		Women
	Odds Ratio	95% CL	Odds Ratio	95% CL
Age Category (Reference: <65)
Age 65–79	2.69	1.69, 4.27	3.10	2.15, 4.46
Age >80	7.14	3.75, 13.60	13.38	8.24, 21.70
Marital Status (Reference: Not Married)	0.80	0.53, 1.21	0.80	0.58, 1.10
Smoking Status (Reference: Non-Smoker)	1.08	0.64, 1.83	1.04	0.66, 1.64
Race/ethnicity (Ref: Non-Hispanic White)
Non-Hispanic Black	1.62	1.02, 2.59	2.81	1.93, 4.10
Hispanic or Mexican American	0.78	0.43, 1.42	2.17	1.41, 3.37
Other Race or Multi-Racial	1.17	0.50, 2.74	2.08	1.17, 3.70
Education Level (Reference: <HS Graduate)
High School Graduate	0.31	0.20, 0.48	0.45	0.32, 0.65
Some College	0.19	0.10, 0.35	0.31	0.20, 0.50
College Graduate	0.15	0.06, 0.33	0.32	0.17, 0.61
>College Graduate	0.12	0.05, 0.28	0.17	0.07, 0.38
Net Worth Quartiles (Reference: Quartile 1)
Quartile 2	0.83	0.51, 1.36	0.51	0.34, 0.75
Quartile 3	0.61	0.36, 1.05	0.52	0.34, 0.81
Quartile 4	0.40	0.21, 0.75	0.72	0.45, 1.18
Elevated hs-CRP	0.86	0.54, 1.36	0.80	0.55, 1.15
Normalized Grip Strength*	1.10	1.02, 1.20	1.02	0.94, 1.10

^*OR and 95%CL per each 0.05 unit lower

Normalized Grip Strength and Inflammation

There was an inverse correlation between hsCRP levels and NGS (r=−0.37, p<0.01). Moreover, elevated hs-CRP (>3.0 mg/L) was associated with significantly lower NGS for men (0.27 high NGS vs. 0.32 low NGS) and women (0.24 vs. 0.20) (all p<0.001).

Discussion

The principal findings of the current study were that among middle aged and older men and women, NGS was associated with incident disability and chronic disease multimorbidity, such that for every 0.05 unit lower strength relative to body mass, there was a 25–31% increased risk of incident physical disability and 14% increased risk for incident chronic disease multimorbidity at a 8-years follow-up, even after adjustment for age, race, socioeconomic status, smoking, marital status, income, and hs-CRP. Moreover, for every 0.05 unit lower NGS, there was a 10% increased risk for incident dementia among men at 8-years follow-up. The results of this study provide further support that individuals with weaker grip strength are at increased risk of chronic disease risk and physical dysfunction compared to stronger individuals (17, 41, 42).

We also demonstrated an independent association between elevated hs-CRP (>3.0 mg/L) and incident chronic disease multimobidity in both men and women, as well as with incident physical disability in women. Moreover, there was an inverse association between hs-CRP and NGS in both men and women, which is suggestive of a biological mechanism linking age-related muscle weakness and risk for disease and disability, and/or that chronic inflammation may be involved in conditions leading to frailty and mortality in elderly individuals. Interestingly, when modeled together there was some evidence of causal mediation for the NGS exposure variable with both physical disabilities among men and women, as well as chronic disease multimorbidity among women (i.e., the proportion of the total NGS association on physical disabilities (men and women) and chronic disease multimorbidity (women) was partially mediated through the effect of hs-CRP). These findings add to the growing body of literature linking muscle weakness and chronic inflammation with disease and disability, and therefore warrant improved clinical attention to these important biomarkers of aging. It is also further confirmation that targeted, early interventions (e.g., physical activity and strengthening exercise (43)) during midlife are needed to forestall or prevent the onset of disease and disability later in life.

There was no significant difference between men and women in incidence of dementia. Interestingly, while there was a significant and independent association between lower NGS and incident dementia among men, there was no significant effect in women. Given that NGS is lower for women across older adulthood, it is possible that the effect of weakness on incident dementia could have been concealed (i.e., a floor effect). In another recent study using HRS, McGrath et al. found that every 5-kg lower grip strength was associated with 10% increased odds for any cognitive impairment, an 18% increased odds for severe cognitive impairment, and a 10% increased odds for poorer cognitive functioning (44). This study also demonstrated that relative to women, lower grip strength conferred significantly higher odds of both poor cognitive and severe cognitive function in men, and thus is supportive of our findings. Future efforts are needed to further elucidate the mechanism explaining how/why weakness may confer greater risk for diminished cognitive functioning in men as compared to women. In both genders, epidemiological studies have shown an increased risk of dementia with the age-related losses of sex steroid hormones. Given that various studies have also shown that females carry an increased risk of developing dementia and Alzheimer’s disease pathology compared to males (45), perhaps the biological mechanism linking weakness with dementia among men is explainable through low testosterone. Indeed, hypogonadism has been linked with chronic disease and multimorbidity in men across adulthood, and is also robustly and inversely associated with NGS (46).

We acknowledge that our study has several limitations. First, we cannot rule out time-varying confounding since baseline measurements of all covariates were included in our final models. Thus, whether declines in NGS “cause” an elevated risk for early-onset disability and chronic disease multimorbidity, or if incident disease processes (e.g., diabetes) themselves, are a cause of diminished muscle function (i.e., competing risks, as implied from the study by Kalyani et al. (47)), is an interesting topic. We were unable to determine if other competing risks or unmeasured confounding (i.e., other risk factors [e.g., lack of physical activity] or existing diseases [e.g., cancer]) may have influenced the observed findings. Second, as with any long-term, prospective study, loss to follow up may have resulted in an underestimation of the true association if sicker individuals were more likely to drop out. Further, given that chronic, low-grade inflammation has been implicated in the pathogenesis of cognitive decline in older adults (48), we were surprised by the lack of association between elevated hs-CRP and dementia in our cohort. Thus, an additional limitation in this study was using a single marker of inflammation. There are multiple markers of chronic inflammation, including hs-CRP, low serum albumin, TNF-α, IFNγ, IL-6, IL-12, and IL-1ß. Future studies could examine multiple markers simultaneously, as chronic inflammation is highly dependent on complex signaling of multiple cytokines. Specific T-cell populations can also be a marker for aging and inflammation (49). Lastly, survey data on chronic disease and disability lack the granular details pertaining to severity of disease and degree of physical disability.

Despite these limitations, this study has various strengths. Most notably, we used grip strength as our primary exposure, which is a cost-effective, reliable proxy for total body muscle strength that can be easily administered in the clinical or community settings, and across populations of children, adolescents, and adults (19, 50–54). Moreover, the results of this study are generalizable to community-dwelling older Americans since the sample used in this study is from the longest running, longitudinal nationally- representative cohort of older adults in the United States. Given the robust and inverse independent association between NGS and chronic disease and disability in our study, further exploration into the biological mechanism linking weakness with chronic disease, disability and early mortality is needed. For example, future efforts to explore the longitudinal association between other biological clocks (e.g., DNA methylation aging using epigenetic clocks such as the Horvath (55), Hannum (56), and Levine models (57)) would shed light on factors that indicate or predict accelerated biological aging as compared to those that facilitate healthy aging.

Conclusions

Our findings indicate a robust inverse association between NGS and disability and chronic, multimorbidity in older men and women, and dementia in men. Elevated hs-CRP was only associated with chronic multimorbidity in men and women, but when modeled with NGS, there was significant evidence of causal mediation for the NGS exposure variable with both physical disabilities among men and women, as well as chronic disease multimorbidity among women. Healthcare providers should implement measures of grip strength in routine health assessments and discuss the potential dangers of weakness as well as interventions to improve strength with their patients.