Elevated IL-6 and CRP Levels Are Associated With Incident Self-Reported Major Mobility Disability: A Pooled Analysis of Older Adults With Slow Gait Speed

Abstract

Background

Elevated interleukine-6 (IL-6) and C-reactive protein (CRP) are associated with aging-related reductions in physical function, but little is known about their independent and combined relationships with major mobility disability (MMD), defined as the self-reported inability to walk a quarter mile.

Methods

We estimated the absolute and relative effect of elevated baseline IL-6, CRP, and their combination on self-reported MMD risk among older adults (≥68 years; 59% female) with slow gait speed (<1.0 m/s). Participants were MMD-free at baseline. IL-6 and CRP were assessed using a central laboratory. The study combined a cohort of community-dwelling high-functioning older adults (Health ABC) with 2 trials of low-functioning adults at risk of MMD (LIFE-P, LIFE). Analyses utilized Poisson regression for absolute MMD incidence and proportional hazards models for relative risk.

Results

We found higher MMD risk per unit increase in log IL-6 (hazard ratio [HR] = 1.26; 95% confidence interval [95% CI] 1.13–1.41). IL-6 meeting predetermined threshold considered to be high (>2.5 pg/mL) was similarly associated with higher risk of MMD (HR = 1.31; 95% CI 1.12–1.54). Elevated CRP (CRP >3.0 mg/L) was also associated with increased MMD risk (HR = 1.38; 95% CI 1.10–1.74). The CRP effect was more pronounced among participants with elevated IL-6 (HR = 1.62; 95% CI 1.12–2.33) compared to lower IL-6 levels (HR = 1.19; 95% CI 0.85–1.66).

Conclusions

High baseline IL-6 and CRP were associated with an increased risk of MMD among older adults with slow gait speed. A combined biomarker model suggests CRP was associated with MMD when IL-6 was elevated.

Persistent, low-grade inflammation is implicated in the etiology and progression of several chronic health conditions (1,2), including age-associated disease and disability (3,4). In particular, elevations in interleukin-6 (IL-6) and C-reactive protein (CRP) have been identified in several longitudinal studies as robust, independent predictors of disability, impaired mobility, and slow walking speed (5–8). These observations point to the inflammatory pathway as a potential intervention target to reduce aging-related disability. Indeed, recent trials have been specifically designed to assess the effect of inflammation-lowering interventions (ie, diet, exercise, losartan, and fish oil supplementation) on physical function (9,10), with emerging data suggestive of potential benefit. For example, in the Intensive Diet and Exercise for Arthritis randomized clinical trial, weight loss-associated reductions in IL-6 were inversely related to clinically meaningful improvement in gait speed among older adults with knee osteoarthritis (9). However, whether this translates into the reduced onset of a clinically relevant endpoint in older adults (eg, major mobility disability [MMD]) has yet to be determined.

MMD (11,12), defined as the inability to walk ¼ mile or 400 m, is of major public health significance. The ability to walk ¼ mile is measured in the US census and in most epidemiologic surveys (13,14). Public health agencies use the ability to walk ¼ mile or 400 m to define the need and policy impact of interventions. Finally, people reporting the inability to walk 400 m incur higher health care costs of $4000 per person per year, compared with those not reporting inability to walk 400 m (14). The loss in mobility is caused by a myriad of factors many of which are biologically oriented. In that regard, systemic inflammation is a leading contender as a major causal and potentially modifiable factor explaining some of the mobility loss. Therefore, understanding the relationship between baseline IL-6 and CRP concentrations and MMD onset is a first step for designing future trials focused on reducing inflammation for the prevention of MMD (10).

Standardization of physical function assessments has resulted in numerous studies incorporating similar subjective and objective disability outcome measures—along with traditional aging-related biomarkers—which allows for the possibility of data pooling across studies. Combining studies with heterogeneous inclusion criteria offers the distinct advantage of providing estimates that are more robust and inclusive of aging subpopulations than any one individual study, while also providing higher power due to increased sample size (15,16). Additionally, the inclusion of studies with measurements of both biomarkers creates opportunities to explore the association of each biomarker separately and in combination with regard to MMD risk.

In this article, we contribute to the growing body of knowledge on factors that affect the risk of MMD by modeling the absolute and relative risks of incident MMD by baseline inflammation based on both IL-6 and CRP across 3 data sets focused on aging, inflammation, and physical function. Furthermore, we measure both the study-specific and pooled log-linear risk of MMD according to IL-6 and CRP values as well as the absolute and relative risks according to previously identified cut-points for elevated inflammation (ie, IL-6 >2.5 pg/mL and CRP >3.0 mg/L) (3,17) in participants from multiple studies at risk of MMD. Finally, we explore the combined effects of elevated levels of IL-6 and CRP including the interaction of elevated IL-6 and CRP. The information gained will aid in understanding the interaction between these 2 common inflammatory markers and help refine entry criteria for trials of inflammation and mobility.

Method

Study Design and Sample Size

We conducted a pooled analysis using individual participant-level data from 3 National Institute on Aging’s Claude D. Pepper Older American Independence Center-affiliated studies reporting baseline IL-6 and CRP as well as the onset of MMD. Participant entry criteria (determined to exclude those participants at low risk of aging-related disease and functional decline) included baseline age at least 60 years and gait speed less than 1.0 m/s. Additionally, participants with baseline IL-6 more than 30 pg/mL were excluded due to this being evidence of an active acute inflammatory process at the time of measurement, and CRP analyses excluded baseline CRP more than 100 mg/L. All participants reported no MMD at baseline. A total of 6 studies (9,12,18–21) were initially evaluated for the presence of a sufficient number (n ≥50) of qualifying participants, with the Health, Aging, and Body Composition (Health ABC) study (21), the Lifestyle Interventions and Independence for Elders Pilot (LIFE-P) study (20), and the LIFE trial (12), ultimately included in the present pooled analysis (n = 1732). Briefly, the Health ABC study recruited 3075 well-functioning men and women older (70–79) adults who reported no difficulty walking ¼ mile or climbing 10 steps. We used the first 5 years of follow-up time of Health ABC, with MMD assessed annually, and baseline gait speed was acquired based on data availability of 400-m walk completion time (n = 263), 6-minute walk distance (n = 131), or 2-minute walk distance (n = 68). The LIFE-P and LIFE studies were 2 trials of 424 and 1635, respectively, older adults (70–89) with poor function (SPPB ≤9) who were assigned to physical activity or health education interventions to prevent mobility disability defined as the inability to walk 400 m, assessed every 6 months. The LIFE-P study followed up participants for 1 year, while LIFE accrued MMD events up to 52 months postbaseline. Both LIFE and LIFE-P used baseline 400-m walk time for baseline gait speed.

Exposure, Outcome, and Covariate Assessment Methodology

Serum IL-6 and CRP were assessed from baseline blood draws in the morning after a 12-hour fast using published methodology by the same central laboratory (6,22), with IL-6 (primary exposure variable) considered elevated at levels above 2.5 pg/mL (3) and CRP (secondary exposure variable) considered elevated above 3.0 mg/L (17). Serum IL-6 was available for all studies, while CRP was assessed in Health ABC and LIFE-P. The main LIFE trial did not measure CRP for budgetary reasons. MMD was defined as either self-reported great/a lot of difficulty or the inability to walk one-quarter mile (Health ABC) (21) or several blocks (LIFE-P, LIFE) (23). Demographic information (ie, age, sex, and race) was collected via self-report at baseline. Body mass index (BMI; kg/m²) was measured by trained research staff with participants dressed in light clothing and without shoes.

Statistical Analyses

Participant characteristics were summarized overall and by study using descriptive measures. Study-specific event rates of reported MMD across low and high levels of inflammation for each inflammatory biomarker were fit separately using a Poisson regression modeling approach. Models included the main effects of inflammation level (low/high) and study as well as study by inflammation level interaction. Overall reported MMD event rates across studies were summarized using Poisson regression stratified by study and used elevated inflammation levels as the primary independent variable. The study-specific linear association between inflammatory biomarker levels and MMD risk used Cox proportional hazards regression models fit with log-transformed inflammatory values, study identifier, and inflammation by study interaction. The overall linear association between inflammation and MMD risk used study as stratification variables. Finally, the study-specific risk of MMD by elevated inflammation levels (>2.5 pg/mL for IL-6 and >3.0 mg/L for CRP) used Cox proportional hazards models with inflammation level, study, and study by inflammation level interactions, and the overall model used the inflammatory level as a predictor and study as a stratification variable. A final model estimated the event rates (Poisson regression model) and relative risk (Cox proportional hazards model) of elevated IL-6, CRP, and their interaction, stratified by study. Supplemental analyses compared study-specific and overall objective MMD using similar modeling approaches for IL-6 exposure, except substituting the self-reported outcome with failure to complete a 400-m walk in 15 minutes as the outcome of interest. Additionally, to account for the potential effect of covariates, the stratified Cox proportional hazards models were rerun adjusting for baseline BMI, age, and sex. Sensitivity and specificities for predicting the 5-year risk of MMD for elevated biomarkers were estimated separately and in combination. All analyses were conducted using SAS v9.4 (SAS Institute, Cary, NC) specifying a nominal Type I error rate of 0.05.

Results

Baseline Characteristics

The number of participants who met entry criteria is presented overall and by study in Table 1. The highest yields came from the 2 studies (LIFE, LIFE-P) that specifically included participants at high risk of MMD based on physical performance. Relative to Health ABC, which recruited a high-functioning sample, participants in the LIFE and LIFE-P trials were 3–5 years older at baseline, more likely to be female and White, and presented with slower gait speed. All studies were comparable with respect to baseline IL-6 concentration (mean [SD]: 4.0 [3.4]; range 3.1–4.3 pg/mL). The proportion of participants self-reporting MMD events varied from a low of 14.7% (LIFE-P) to 61.5% (Health ABC), which is partially a function of the duration of follow-up time.

Table 1.

Baseline Characteristics for Participating Studies

Variable	Overall	Health ABC*	LIFE	LIFE-P
Study N (% of original)	1732 (100.0)	478 (15.5)	1016 (62.1)	238 (56.1)
Age (years)	77.5 ± 5.1	74.2 ± 3.0	79.0 ± 5.3	77.8 ± 4.1
Female (%)	1024 (59.1)	147 (30.8)	705 (69.4)	172 (72.3)
White race (%)	1099 (63.5)	154 (32.2)	766 (75.4)	179 (75.2)
BMI (kg/m2)	30.15 ± 6.11	29.25 ± 6.24	30.38 ± 5.97	30.96 ± 6.27
Gait speed (m/s)	0.82 ± 0.13	0.88 ± 0.11	0.79 ± 0.13	0.80 ± 0.13
IL-6 (pg/mL)	3.97 ± 3.37	3.69 ± 3.27	4.30 ± 3.55	3.12 ± 2.47
IL-6 >2.5 pg/mL (%)	1082 (62.5)	254 (53.1)	710 (69.9)	118 (49.6)
CRP (mg/L)	9.46 ± 28.39	11.92 ± 34.20	—	4.53 ± 6.44
CRP >3.0 mg/L (%)	370 (53.7)	270 (59.9)	—	100 (42.0)
MMD events (%)	670 (38.7)	294 (61.5)	341 (33.6)	35 (14.7)
FU Time (years)	2.61 ± 1.45	2.86 ± 1.80	2.87 ± 1.14	0.94 ± 0.16
Median FU time (years)	3.0	2.5	3.0	1.0

Notes: Health ABC = Health, Aging, and Body Composition; LIFE = Lifestyle Interventions For Elders; LIFE-P = Lifestyle Interventions for Elders Pilot; BMI = body mass index; IL-6 = interleukine-6; CRP = C-reactive protein; MMD = major mobility disability; FU = follow-up. Mean ± SD or N (%) unless otherwise noted.

*The first 5 years of follow-up were used for Health ABC.

Association Between Baseline IL-6 and Incident MMD

Event rates for each of the studies by IL-6 inflammation level (≤2.5 pg/mL and >2.5 pg/mL) are presented in Table 2. The lowest rates were observed in LIFE study participants with low baseline IL-6 (8.6 events/100 person-years [PY]), but participants with elevated IL-6 experienced rates of 13.1 events/100 PY. Meanwhile, higher event rates overall were reported in Health ABC (19.6 events/100 PY for low IL-6; 23.3 events/100 PY for high IL-6). All of the studies demonstrated higher MMD event rates for participants with elevated inflammation, which is reflected in the overall rate combined across the studies. The log-linear association between IL-6 and MMD relative risk indicated that overall higher log IL-6 values at baseline were associated with significantly higher MMD risk per unit increase in log IL-6 (hazard ratio [HR] = 1.26; 95% confidence interval [95% CI] 1.13–1.41). High IL-6 (>2.5 pg/mL) was similarly associated with a higher risk of MMD compared to lower values (HR = 1.31; 95% CI 1.12–1.54). Elevated IL-6 had 66% (443/670; 95% CI 63%–70%) sensitivity and 40% (423/1062; 95% CI 37%–43%) specificity for predicting MMD over 5 years.

Table 2.

Association Between Baseline IL-6 and Incident MMD by Study and Overall (estimates with 95% confidence intervals)

Study	MMD Event Rate/100 PY*		Log IL-6†		IL-6 >2.5 pg/mL‡
	IL-6 ≤2.5 pg/mL	IL-6 >2.5 pg/mL	Hazard Ratio	p	Hazard Ratio	p
Health ABC	19.6 (16.6–23.3)	23.3 (19.9–27.1)	1.30 (1.11–1.51)	.001	1.18 (0.94–1.48)	.162
LIFE	8.6 (6.9–10.7)	13.1 (11.6–14.8)	1.23 (1.04–1.46)	.016	1.51 (1.18–1.94)	.001
LIFE-P	14.2 (8.7–23.1)	17.1 (10.9–26.8)	1.25 (0.74–2.13)	.403	1.21 (0.62–2.35)	.576
Overall§	13.3 (11.7–15.1)	15.8 (14.4–17.4)	1.26 (1.13–1.41)	.021	1.31 (1.12–1.54)	.001

Note: Health ABC = Health, Aging, and Body Composition; LIFE = Lifestyle Interventions For Elders; LIFE-P = Lifestyle Interventions for Elders Pilot; IL-6 = interleukine-6; MMD = major mobility disability.

*Estimated from a Poisson regression model with IL-6 category, study, and IL-6 category by study interaction as predictors.

^†Estimated from Cox proportional hazards regression models with log IL-6, study, and log IL-6 by study as predictors.

^‡Estimated from Cox proportional hazards regression models with IL-6 category, study, and IL-6 category by study interaction as predictors.

^§All models in “Overall” line use Study as strata rather than covariates.

Association Between Baseline CRP and Incident MMD

The association between baseline CRP levels and MMD events is presented in Table 3. The LIFE study did not have CRP measurements available, but the 2 studies with available baseline CRP data displayed consistent associations between elevated CRP and elevated MMD risk. The study-specific log-linear CRP associations with MMD were positive, although only Health ABC achieved statistical significance (p = .019). However, the overall association when stratified by the 2 studies indicated an increasing risk of MMD as log CRP increased (p = .007). Finally, elevated CRP levels (CRP >3.0 mg/L) were associated with a 38% increased risk (HR = 1.38; 95% CI 1.10–1.74; p = .006) of MMD compared to lower baseline CRP levels. Elevated CRP had 63% (200/317; 95% CI 58%–68%) sensitivity and 54% (202/372; 95% CI 49%–59%) specificity for predicting MMD over 5 years.

Table 3.

Association Between Baseline CRP Inflammation and Incident MMD (estimates with 95% confidence intervals)

Study	MMD Event Rate/100 PY*		Log CRP†		CRP >3.0 mg/L‡
	CRP ≤3.0 mg/L	CRP >3.0 mg/L	Hazard Ratio	p	Hazard Ratio	p
Health ABC	18.0 (14.8–21.9)	24.8 (21.4–28.7)	1.12 (1.02–1.23)	.019	1.35 (1.06–1.72)	.017
LIFE-P	12.3 (7.5–20.0)	20.3 (13.0–31.9)	1.23 (0.93–1.63)	.144	1.66 (0.85–3.22)	.136
Overall§	16.9 (14.1–20.3)	24.3 (21.1–27.9)	1.13 (1.03–1.24)	.007	1.38 (1.10–1.74)	.006

Note: Health ABC = Health, Aging, and Body Composition; LIFE-P = Lifestyle Interventions for Elders Pilot; CRP = C-reactive protein; MMD = major mobility disability.

*Estimated from a Poisson regression model with CRP category, study, and CRP category by study interaction as predictors.

^†Estimated from Cox proportional hazards regression models with log CRP, study, and log CRP by study as predictors.

^‡Estimated from Cox proportional hazards regression models with CRP category, study, and CRP category by study interaction as predictors.

^§All models in “Overall” line use Study as strata rather than covariates.

Association Between IL-6 and CRP and Incident MMD

Table 4 shows that the combination of elevated CRP and elevated IL-6 yield MMD rates and risk ratios were higher than any of the individual categories, after statistical testing revealed a nonsignificant interaction (p = .22). The effect of elevated CRP was more pronounced among participants with elevated IL-6 (HR = 1.62; 95% CI 1.12–2.33) compared to those with lower IL-6 levels (HR = 1.19; 95% CI 0.85–1.66). The group with both elevated IL-6 and CRP had a 37% increased risk of MMD (HR = 1.37; 95% CI 1.04–1.80; p = .026) compared to participants with combined lower IL-6 (≤2.5 pg/mL) and CRP (≤3.0 mg/L) at baseline (not presented in tables). The survival distributions of the combined inflammation subgroups are presented in Figure 1. The greatest absolute risk appears to occur among participants with elevated CRP, although importantly the subgroup with elevations in both inflammatory biomarkers had the lowest survival probabilities. The combination of elevated IL-6 and elevated CRP had 44% (138/317; 95% CI 38%–49%) sensitivity and 71% (263/372; 95% CI 66%–75%) specificity for predicting MMD over 5 years, while a predictor of elevated IL-6 or elevated CRP had 75% (237/317; 95% CI 70%–80%) sensitivity and 34% (125/372; 95% CI 29%–38%) specificity.

Table 4.

Combined IL-6 and CRP and Incident MMD (estimates with 95% confidence intervals)

	MMD Event Rate/100 PY*				Relative Risk of MMD†
					IL-6 ≤2.5 pg/mL		IL-6 >2.5 pg/mL
	CRP ≤3.0 mg/L, IL-6 ≤2.5 pg/mL	CRP >3.0 mg/L, IL-6 ≤2.5 pg/mL	CRP <3.0 mg/L, IL-6 >2.5 pg/mL	CRP >3.0 mg/L, IL-6 >2.5 pg/mL	CRP >3.0 vs. CRP ≤3.0	p	CRP >3.0 vs. CRP ≤3.0	p
					Hazard Ratio		Hazard Ratio
Overall	18.0 (14.4–22.4)	21.8 (17.0–28.0)	15.0 (10.9–20.7)	25.6 (21.6–30.2)	1.19 (0.85–1.66)	.311	1.62 (1.12–2.33)	.010

Note: IL-6 = interleukine-6; CRP = C-reactive protein; MMD = major mobility disability.

*Estimated from a Poisson regression model stratified by study.

^†Estimated from Cox proportional hazards regression model stratified by study with CRP category, IL-6 category, and CRP × IL-6 interaction as predictors.

Figure 1.

Kaplan–Meier plot of time to major mobility disability (MMD) for interleukine-6 (IL-6) and C-reactive protein (CRP) subgroups from LIFE-P and Health ABC studies. IL-6 units are pg/mL and CRP units are mg/L.

Association Between Baseline IL-6 and Objective 400-m Walk MMD

Because included studies (Health ABC, LIFE, and LIFE-P) had an objective assessment of MMD defined as completion of a 400-m walk course in less than 15 minutes, it was of interest to compare these results to that of subjective MMD. In Supplementary Table S1, we observe that for these 3 studies, the absolute and relative risks of MMD assessed using objective measures were quite similar to those of the subjective counterparts (Table 2). All 3 studies showed consistently elevated risk of objectively measured MMD for individuals with IL-6 more than 2.5 pg/mL, with both the LIFE study and the overall hazard ratios achieved statistical significance (both p < .001). Finally, to address the potential effects of covariates in Supplementary Table S2, we present the hazard ratios and p values for inflammation categories from Tables 2–4 and note that the effects of IL-6, CRP, and their combination remain significant after adjustment for sex, age, and BMI.

Discussion

The novel findings presented here indicate that IL-6 and CRP levels exceeding 2.5 pg/mL and 3.0 mg/L, respectively, were associated with significantly increased risk of subjectively and objectively measured incident MMD. The major difference between prior work (3,24) and the current findings is twofold. First, we focus on individuals with slow gait speed at risk for MMD. We then examine both the independent and combined relationships of these biomarkers of inflammation to self-reported MMD and confirm that the patterns in the data parallel those for objectively measured MMD. These results suggest that individuals with lower IL-6 have similar rates of MMD, regardless of their CRP levels. When both IL-6 and CRP are high, the risk of MMD is significantly elevated above that of participants with normal IL-6 and CRP; however, the effect is not truly additive. Our findings help to clarify the utility of CRP and IL-6 meta-analytic results (25) on inflammation and mobility-related outcomes from other cohort studies and show their predictive value among older adults who are already impaired in gait speed (3,5).

MMD has been linked to institutionalization morbidity, mortality, hospitalization, and poor quality of life (26–29). However, its cause is multifactorial without a clear biological target to preserve mobility into late life. Chronic inflammation has been of key interest in the biology of aging for the past 50 years (30); however, the cause and consequences remain unclear. For example, inflammation may indicate an ongoing process of age-related damage and repair resulting in a cumulative burden of disease-related damage and cellular senescence. Alternately, elevated inflammatory biomarkers might represent a consequence of senescent cells that are resistant to apoptosis are known to secrete cytokines such as IL-6 and chemokines. Their circulation may disrupt normal tissue function, promoting central adiposity, atherosclerotic plaque, sarcopenia (31), and osteoporosis (32,33). Moreover, age-related renal dysfunction (34) and atherosclerosis (35,36) may contribute to the elevation of inflammatory markers through decreased excretion and increased production of these markers. The culmination of these effects has a negative impact on tissues that support mobility that lead to increased risk of MMD.

IL-6 and CRP inflammatory factors have received the most attention as potential biomarkers of physical function in late life and potential markers for use in clinical trials. The challenge facing investigators is to choose which biomarker (or their combination) appropriately selects individuals at elevated risk for geriatric-related outcomes. When examined separately, the risk ratios on MMD were similar between CRP and IL-6. Interestingly, IL-6 appeared to be more strongly associated with MMD than CRP when considered in combination. For instance, elevated CRP was not significantly associated with MMD when IL-6 was low. This could be due to differences in the biological targets of CRP and IL-6. For example, IL-6, but not CRP, activates pathways that regulate muscle protein degradation and impairs myogenesis causing muscle atrophy (37). CRP is less specific; it responds broadly to acute and chronic inflammatory conditions (bacterial, rheumatic, tissue injury, necrosis, and viral infections). Another difference between the 2 biomarkers is their stability. IL-6 is more stable and less sensitive to diurnal variations whereas CRP has high intraindividual biological variability (38,39). These biological and measurement differences help to explain the risk differences and lack of additive effects of IL-6 and CRP on MMD risk. While our results demonstrate the elevated risk of MMD in the setting of increased levels of IL-6 and CRP, the magnitude of the effect is not sufficient to support population-based screening, nor are the sensitivities and specificities particularly strong as a prognostic tool. However, they might be sufficient for identifying older adults for inflammation-reducing interventions to preserve mobility levels. Future work should further explore whether these 2 biomarkers in combination with other risk factors are more sensitive and specific for screening purposes.

Strengths of the study include the large samples of individual participant data with low gait speed from 3 heterogeneous samples. Individual-level data allow for better refinement of study samples and permit the modeling of IL-6 and CRP in the same model. Despite differences in the samples, these studies shared key design characteristics such that overlapping measures across studies were frequently performed according to the same or similar protocols. The limitations of this study include differences in the operational definitions of reported MMD, which although similar could introduce some measurement error. Likewise, biomarkers were assessed at different times, albeit using the same laboratory, unlike the procedures that would typically be used in a single study, which could induce heterogeneity with respect to study-specific sensitivity and specificity across inflammation categories. Finally, although the heterogeneity of the samples across studies is generally considered a strength, the presence of active intervention in the LIFE and LIFE-P studies could lead to MMD estimates that differ from the population as a whole.

In conclusion, using a large sample of older adults with low gait speed, we found circulating inflammation levels measured by IL-6 and CRP were associated with elevated incident mobility disability risk. Furthermore, although their interaction was not significant, our findings suggest the relationship between CRP and MMD could be augmented by elevated IL-6. Thus, baseline inflammation levels might be useful in predicting individuals at potentially higher risk of developing MMD, and future work should explore whether reducing IL-6 and CRP confers a protective effect on mobility status in at-risk older adults.

Funding

This work was funded by the National Institutes of Health (NIH) grant U01AG050499. The Health, Aging, and Body Composition study was supported by National Institute on Aging (NIA) contracts N01-AG-6-2101, N01-AG-6-2103, N01-AG-6-2106; NIA grant R01-AG028050; and NINR grant R01-NR012459. This research was funded in part by the Intramural Research Program of the NIH, NIA. The Lifestyle Interventions and Independence for Elders Pilot (LIFE-P) study was funded by a grant from the NIH/NIA (U01 AG22376) and supported in part by the Intramural Research Program, NIA, NIH. The Lifestyle Interventions and Independence for Elders study was funded by cooperative agreement U01AG22376 from the NIH and NIA; supplement 3U01AG022376-05A2S from the National Heart, Lung, and Blood Institute; and was sponsored in part by the Intramural Research Program. Portions of this work were supported, in part, by the WFU Claude D. Pepper Older Americans Independence Center (P30 AG21332), University of Pittsburgh Claude D. Pepper Older Americans Independence Center (P30 AG024827) and University of Florida Claude D. Pepper Older Americans Independence Center (P30 AG028740). R.A.F.’s contribution was partially supported by the US Department of Agriculture (USDA), under agreement No. 58-8050-9-004. T.M.G. is supported by the Yale Claude D. Pepper Older Americans Independence Center (P30AG21342) and an Academic Leadership Award (K07AG043587) from the NIA.

Conflict of Interest

None declared.