Abstract
Background
Circulating levels of procollagen type III N-terminal peptide (P3NP) may reflect increased fibrosis of skeletal muscle and other tissues with aging. Herein, we tested if P3NP was associated with baseline and 7-year change in physical function.
Method
Participants (n = 400) were from the Long Life Family Study, a study of exceptional familial longevity. Plasma P3NP concentration was measured using a sandwich enzyme-linked immunosorbent assay (inter-assay coefficient of variation <5.5%). At baseline and 7-year follow-up visits, physical function was measured using the Short Physical Performance Battery (SPPB score 0–12), which consists of gait speed, balance, and chair-rise tests. Grip strength was measured using a handheld dynamometer. The association between log-transformed P3NP and physical function was examined using generalized estimating equations adjusted for familial relatedness, age, sex, height, weight, lifestyle characteristics, liver function, kidney function, lung function, and chronic disease prevalence.
Results
Participants were aged 73.1 ± 15.2 years (range: 39–104), 54% female, had body mass index of 26.6 ± 4.3 kg/m2, and gait speeds of 1.0 ± 0.3 m/s. One standard deviation higher log-transformed P3NP was related to worse baseline SPPB score (β = −0.9points), gait speed (β = −0.05m/s), chair-rises per-second (β = −0.46chair-rises/10 seconds), and grip strength (β = −2.0kg; all p < .001). Higher P3NP was also associated with greater declines in gait speed (β = −1.41, p < .001) and transitioning to being unable to perform chair-rises (β = 0.41, p < .001) after 7 years.
Conclusion
Plasma P3NP may be a strong, novel biomarker of current and future physical function. Future research is needed to extend our findings to other cohorts and determine mechanisms underlying these associations.
Keywords: Biomarker, Disability and epidemiology, Fibrosis
Impaired physical function with aging is an enormous clinical and public health problem that leads to higher risks for disability (1,2), institutionalization (1), health care costs (3,4), and mortality (1,5). Impairments in physical function, especially mobility impairments, affect ~50% of the 27.8 million Americans aged 70 and older (6–8). Effective strategies to prevent or improve physical function in older adults are currently limited to lifestyle interventions, such as weight loss and physical activity. However, lifestyle changes are very difficult to adopt, especially for those at the highest risk for functional decline and disability (9,10). To extend the benefits of these interventions, it is imperative that we gain a better understanding of the biological mechanisms underlying age-related functional decline (8,11).
Procollagen type III N-terminal peptide (P3NP) is released into circulation as a result of collagen biosynthesis (8,12), and is a biomarker of active fibrosis in several tissues such as the heart (13,14), lungs (15–17), and liver (18–20), which may contribute to declining physical function with aging. P3NP is also increased during periods of skeletal muscle repair/growth (21) and has been shown to increase following initiation of structured physical activity (22) and with growth hormone and testosterone treatment (23). However, most community-dwelling adults are not taking anabolic agents or performing structured physical activity (7). Instead, muscle mass decreases with age (24) and higher levels of P3NP may reflect replacement of muscle with more fibrotic tissue (25). Thus, P3NP has also been proposed as a potential biomarker for sarcopenia (25). Indeed, the Framingham Offspring Study showed that higher levels of P3NP are associated with lower lean mass in older women, but associations with physical function have not been examined (26).
To study the potential utility of P3NP as a biomarker of physical function, we conducted an ancillary study of circulating P3NP levels in a subset of 400 men and women (age range: 39–104) from the Long Life Family Study (LLFS). We examined the association between plasma P3NP levels with grip strength, gait speed, time to complete 5 repeated chair-rises, and the Short Physical Performance Battery (SPPB) score in these individuals cross-sectionally at study initiation and with individual changes during a 7-year follow-up period. We also determined if this relationship differed by generation, and in those aged 65 and older compared with those younger than 65 years.
Method
Study Population
Participants were selected from the LLFS: an ongoing multi-center study that recruited and enrolled families from 3 U.S. (Boston, New York, and Pittsburgh) and 1 Denmark (Odense) field center. Detailed recruitment/enrollment criteria have been described elsewhere (27). Briefly, family eligibility criteria were: a long-lived individual (proband) aged 90 years and older in Denmark and 80 and older in U.S. field centers, at least one enrolled sibling of the proband, at least one enrolled offspring of either the proband or their sibling(s), and exceptional familial longevity (Family Longevity Selection Score ≥7) (28). Upon identifying an eligible family, an attempt was made to recruit all members of the proband and offspring generations. Baseline, phenotyping, and blood collection were performed at in-person visits from 2006 to 2009. From 2014 to 2017, participants were invited to participate in a follow-up in-person visit (Visit 2). The protocol was approved by the Human Research Protection Office of the coordinating center at Washington University, the Regional Scientific Ethical Committees for Southern Denmark, and the institutional review boards at the University of Pittsburgh, Boston Medical Center, and Columbia University. All participants provided written informed consent. The cohort is primarily white (>99%) and initially recruited 2 generations: “proband” (median birth year: 1917) and “offspring” (median birth year: 1947).
Previously, a subset of LLFS participants from both the Proband and Offspring generations had gene expression measured (n = 456). An unrelated aim of this pilot study was to examine the association between gene expression and circulating P3NP levels. Thus, we measured plasma P3NP levels in 400 LLFS participants who had gene expression data. Specifically, sample selection for this pilot study was as follows: first, the 254 individuals, regardless of age, with valid baseline and follow-up physical function data were selected; and second, the remaining 146 individuals were randomly selected from the 167 individuals who were aged 65 and older. Thus, the final sample for this pilot study included 254 individuals with valid longitudinal data, and 146 individuals aged 65 and older with baseline data only, for a total for n = 400. Descriptive characteristics for the total sample and by generation can be found in Table 1. A table comparing baseline characteristics between those with and without follow-up data can be found in Supplementary Table 2.
Table 1.
General Characteristics of Participants
| Overall (N = 400) | Proband Generation (n = 155) | Offspring Generation (n = 245) | |
|---|---|---|---|
| Sex (female) | 216 (54.0%)a | 85 (54.8%) | 131 (53.5%) |
| Age (y) | 73.1 ± 15.2 | 89.6 ± 7.3 | 62.6 ± 7.9 |
| P3NP (µg/L) | 6.8 [5.4, 9.3] | 9.3 [7.5, 11.3] | 5.9 [5.0, 7.0] |
| Log-transformed P3NP | 1.98 ± 0.42 | 2.27 ± 0.40 | 1.80 ± 0.32 |
| Height (cm) | 165.4 ± 9.9 | 160.2 ± 10.0 | 168.6 ± 8.4 |
| Weight (kg) | 73.2 ± 15.0 | 66.8 ± 13.9 | 77.2 ± 14.2 |
| BMI (kg/m2) | 26.6 ± 4.3 | 25.9 ± 3.9 | 27.1 ± 4.5 |
| Completed high school (%) | 315 (78.8%) | 104 (67.1%) | 211 (86.1%) |
| Married (%) | 386 (96.5%) | 149 (96.1%) | 237 (96.7%) |
| Ever smoker (%) | 170 (42.5%) | 49 (31.6%) | 121 (49.4%) |
| Current smoker (%) | 29 (7.3%) | 1 (0.7%) | 28 (11.4%) |
| Drink >1/wk (%) | 89 (22.3%) | 14 (9.0%) | 75 (30.6%) |
| COPD (%) | 10 (2.5%) | 4 (2.6%) | 6 (2.5%) |
| Arthritis (%) | 124 (31.2%) | 59 (38.6%) | 65 (26.6%) |
| Kidney disease (%) | 5 (1.3%) | 1 (0.7%) | 4 (1.6%) |
| Depression (%) | 49 (12.3%) | 15 (9.7%) | 34 (14.9%) |
| Stroke (%) | 27 (6.8%) | 19 (12.3%) | 8 (3.3%) |
| Lung disease (%) | 50 (12.5%) | 17 (11.0%) | 33 (13.5%) |
| Peripheral artery disease (%) | 17 (4.3%) | 14 (9.0%) | 3 (1.2%) |
| Cancer (%) | 95 (23.8%) | 50 (32.3%) | 45 (18.4%) |
| Hypertension (%) | 219 (54.8%) | 105 (67.7%) | 114 (46.5%) |
| Diabetes (%) | 23 (5.8%) | 15 (9.7%) | 8 (3.3%) |
| CVD (%) | 35 (8.8%) | 24 (15.5%) | 11 (4.5%) |
| CES-D score | 3.0 [1.0, 6.0] | 4.0 [2.0, 7.0] | 2.0 [1.0, 5.0] |
| FEV, mL | 2405.1 ± 845.1 | 1736.1 ± 587.2 | 2783.1 ± 726.4 |
| eGFR | 62.4 ± 15.1 | 54.8 ± 15.4 | 67.2 ± 12.8 |
| IL-6 (pg/mL) | 2.5 ± 7.0 | 3.4 ± 9.2 | 2.0 ± 5.1 |
| hsCRP (mg/L) | 3.1 ± 4.6 | 3.8 ± 4.9 | 2.6 ± 4.2 |
| Albumin (g/dL) | 3.97 ± 0.30 | 3.88 ± 0.30 | 4.02 ± 0.29 |
| Grip strength (kg) | 28.0 ± 11.6 | 19.9 ± 8.3 | 33.1 ± 10.5 |
| Gait speed (m/s) | 1.0 ± 0.3 | 0.8 ± 0.3 | 1.2 ± 0.2 |
| Chair-rise time (s) | 11.4 ± 4.0 | 14.2 ± 5.2 | 10.0 ± 2.3 |
| Chair rises/10 s | 4.24 ± 2.02 | 2.83 ± 2.00 | 5.13 ± 1.44 |
| SPPB score | 11.0 [9.0, 12.0] | 8.0 [5.0, 10.0] | 12.0 [11.0, 12.0] |
Notes: BMI = body mass index; CES-D = Center for Epidemiologic Studies Depression Scale; COPD = chronic obstructive pulmonary disease; CVD = cardiovascular disease; eGFR = estimated glomerular filtration rate; FEV = forced expiratory volume; hsCRP = high-sensitivity C-reactive protein; IL = interleukin; P3NP = procollagen type III N-terminal peptide; SPPB = Short Physical Performance Battery.
aData are presented as mean ± SD for normal continuous, median [interquartile range] for non-normal continuous, and frequency (%) for dichotomous.
Grip Strength and Physical Function
At both baseline and Visit 2, grip strength was measured using a JAMAR handheld dynamometer (Sammos Preston Rolyan, Bolingbrook, IL). Physical function was measured with the SPPB, which includes 3 tests: 4-m gait speed, a balance battery, and 5 repeated chair-rises; each test is scored from 0 to 4, with 4 being best, for a total SPPB score of 0–12 (29). Gait speed (m/s) and time to complete the chair-rises (s) were also treated as continuous variables. In order to include those unable to complete the chair-rise test due to a physical problem, in sensitivity analyses, we calculated the number of chair-rises/10 s ([5/s to complete]*10), assigning 0 to those unable to complete the task (30).
Procollagen Type III N-Terminal Peptide
At in-person visits, fasting blood samples were drawn according to a standardized protocol (31). Unprocessed ethylenediaminetetraacetic acid (EDTA) whole blood tubes were shipped for next-day delivery to the LLFS central laboratory, the Advanced Research and Diagnostic Laboratory (ARDL), at the University of Minnesota. All samples were stored at −70 °C until immediately prior to performing the assay.
Stored baseline EDTA plasma samples were used for this pilot study, and P3NP was measured using the Cisbio P3NP ELISA kit at the ARDL. This assay is a 1-step sandwich enzyme-linked immunosorbent assay (ELISA), with inter-assay coefficient of variations (CVs) less than 5.5%. In addition, 3 samples were run 10 times each on the same assay plate, and these samples had intra-assay CVs of 4.0%, 2.6%, and 3.9%. The samples were run as 1 batch/d, and controls included with each batch did not show any batch effects. Plasma P3NP concentration was expressed as micrograms per liter (µg/L).
Other Measures
Participants self-reported information on sex, race, smoking, and education. Baseline disease history was based on self-reported physician diagnosis, unless otherwise noted. Cardiovascular disease (CVD) was defined as prior myocardial infarction, or coronary artery bypass, or congestive heart failure; stroke included stroke or transient ischemic attack; and hypertension included presenting with a systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, or history of hypertension confirmed by use of antihypertensive medication. Spirometry was used to determine forced expiratory volume in 1 second (FEV1) (31). Estimated glomerular filtration rate was calculated from urinary creatinine using validated race- and sex-specific equations and albumin, serum interleukin-6 (IL-6), and high-sensitivity C-reactive protein (hsCRP) were measured using validated assays as described previously (31).
Statistical Methods
Baseline characteristics were summarized using mean and standard deviation for normally distributed continuous measures, median and interquartile range for non-normally distributed continuous measures, and frequency and percent for categorical measures. P3NP levels were assessed for normality, and log transformed for analyses. Data are presented for the entire analytic sample (N = 400), and by generation (proband n = 155 and offspring n = 245). In Supplementary Table 1, we also provided a comparison of descriptive characteristics for those aged 65 years and older (n = 259) with those younger than 65 years (n = 141).
The cross-sectional associations between P3NP levels and both grip strength and physical function measures (SPPB score, gait speed, and chair-rise time) were determined using separate generalized estimating equations (GEEs) incorporating an exchangeable correlation structure to account for familial relationships of participants. Betas were calculated per standard deviation higher log-transformed P3NP and age, sex, height, and weight were forced into final models. To determine other potential covariates, the age- and sex-adjusted relationship between continuous variables known to affect physical function and P3NP were assessed using Pearson correlations; for binary traits, to be conservative, unadjusted P3NP levels were compared between those with and without a condition using the Wilcoxon rank-sum test (Tables 2 and 3). Covariates were assessed in context of the base GEE model (field center, sex, age, height, weight, and family relatedness) by groups as follows: first demographic characteristics were entered into GEE models, and those that remained p <.10 were retained. Next, prevalent chronic diseases, albumin, eGFR, and FEV1 were entered and those that remained p <.10 were retained (demographic characteristics no longer p <.10 were removed)—this was considered final Model 1. Finally, final Model 2 additionally adjusted for levels of the inflammatory cytokines IL-6 and hsCRP if they remained p <.10 in the context of final Model 1. As a sensitivity analysis, we examined relationships stratified by generation, sex, and age 65 and above using the same covariates as in overall final models.
Table 2.
Untransformed P3NP Levels by Participant Characteristics
| P3NP Level (µg/L) | |||
|---|---|---|---|
| Noa | Yesa | p b | |
| Age 65 y and older | 5.5 [4.7, 6.6] | 9.1 [6.2, 10.4] | <.0001 |
| Sex (female) | 6.7 [5.4, 9.2] | 7.0 [5.5, 9.3] | .7080 |
| Ever smoker (%) | 7.3 [5.7, 10.0] | 6.3 [5.2, 8.1] | .0018 |
| Current smoker (%) | 7.0 [5.5, 9.6] | 5.5 [5.1, 6.5] | .0004 |
| Drink >1/wk (%) | 7.3 [5.7, 10.0] | 5.7 [5.1, 7.0] | <.0001 |
| Heart attack (%) | 6.7 [5.4, 9.1] | 10.1 [6.4, 11.2] | .0071 |
| Congestive heart failure (%) | 6.7 [5.4, 9.0] | 12.3 [10.6, 18.4] | <.0001 |
| COPD (%) | 6.8 [5.4, 9.3] | 6.5 [5.6, 9.7] | .9779 |
| Arthritis (%) | 6.4 [5.2, 8.6] | 7.7 [6.0, 10.6] | <.0001 |
| Kidney disease (%) | 6.8 [5.4, 9.3] | 9.6 [7.3, 13.4] | .0599 |
| Depression (%) | 6.8 [5.4, 9.3] | 6.9 [5.7, 8.2] | .8646 |
| Heart disease (%) | 6.7 [5.4, 9.0] | 10.5 [6.6, 11.9] | .0003 |
| Stroke (%) | 6.7 [5.4, 9.2] | 7.9 [6.5, 10.5] | .0334 |
| Lung disease (%) | 6.8 [5.4, 9.3] | 7.3 [5.7, 9.6] | .6030 |
| Peripheral artery disease (%) | 6.7 [5.4, 9.3] | 8.2 [6.5, 10.7] | .0229 |
| Cancer (%) | 6.7 [5.4, 9.1] | 7.5 [5.6, 10.0] | .1038 |
| Hypertension (%) | 6.2 [5.1, 7.8] | 7.6 [5.7, 10.0] | <.0001 |
| Diabetes (%) | 6.7 [5.4, 9.1] | 8.5 [5.8, 10.9] | .0528 |
| CVD (%) | 6.6 [5.3, 8.8] | 10.9 [7.6, 12.8] | <.0001 |
Notes: COPD = chronic obstructive pulmonary disease; CVD: cardiovascular disease; P3NP = procollagen type III N-terminal peptide. Data are median [interquartile range].
aYes vs No refers to having vs not having the condition in the far-left column. bp-Values were generated using the Wilcoxon rank-sum test and correspond to the difference in P3NP level between the “No” and “Yes” columns.
Table 3.
Age- and Sex-Adjusted Correlations With Log-Transformed Plasma P3NP Levels
| Pearson r | p | |
|---|---|---|
| eGFR | −0.3540 | <.0001 |
| FEV1 | −0.0393 | .4495 |
| hsCRP | 0.1352 | .0208 |
| IL-6 | 0.2210 | <.0001 |
| Albumin | −0.1927 | .0001 |
Note: eGFR = estimated glomerular filtration rate; FEV1 = forced expiratory volume in 1 s; hsCRP = high-sensitivity C-reactive protein; IL-6 = interleukin-6; P3NP = procollagen type III N-terminal peptide.
We also examined the association between baseline P3NP levels and longitudinal change in grip strength, gait speed, and chair-rise time. To limit measurement error when calculating change using 2 time points, changes in physical function and grip strength were calculated using a 2-stage growth curve model (using Proc Mixed in SAS v9.4, Cary, NC) adjusted for field center, sex, age, age2, with an unstructured covariance matrix, with the intercept and slope as random effects, and familial relatedness as random effects. This model estimates both individual change and population effects, while properly modeling error at each time point. To examine the association between P3NP and the estimated change in physical function, GEE using the same model-building steps as in cross-sectional analyses was used. Base models were also adjusted for the baseline values of the outcome, and weight change from baseline to Visit 2. Finally, to capture those who were unable to perform the chair-rise test due to a physical problem, we did 2 things. First, we calculated change in chair-rises/10 s and analyzed this in GEE models using the same model-building approach described above, except baseline chair-rises/10 s was forced into all models. Second, we examined change in chair-rise time categorically including 3 groups: the first 3 were based on distributional tertiles (1 = best trajectory and 3 = worst trajectory), and the fourth group consisted of those who transitioned from being able to perform the test at baseline to being unable at Visit 2. All analyses were performed using SAS v9.4.
Results
Overall, participants were aged 73.1 ± 15.2 years (range: 39–104), 54% female, and had a body mass index of 26.6 ± 4.3 kg/m2 (Table 1). Participants had a wide range of grip strengths and functional capacities, as exhibited by the variation in SPPB score, chair-rise time, and gait speed (Table 1). Demographic characteristics and base chronic disease prevalence can also be found in Table 1.
P3NP concentration did not differ by sex, but higher P3NP levels were associated with older age, lower eGFR, higher hsCRP, and IL-6 levels, as well as prevalent CVD, arthritis, stroke, peripheral arterial disease, and hypertension (Tables 2 and 3.). Smokers and drinkers had lower levels of P3NP, but this association was attenuated after adjusting for age (Supplementary Table 3). The bivariate relationships between baseline P3NP concentration and baseline physical function are depicted in Figure 1; P3NP levels explained 38% of the variance in total SPPB score, 16% of the variance in grip strength, and 30% of the variance in gait speed. We also examined the bivariate association between P3NP and age. Overall, P3NP levels explained 37% of the variance in age. However, when the association between P3NP and age was examined among those younger than 65, there was no relationship (r2 = .003, p = .56), but there was a strong association between P3NP and age among those aged 65 years and older (r2 = .34, p = <.001).
Figure 1.
Bivariate association between log P3NP levels and baseline SPPB score (A), gait speed (B), and grip strength (C). The bivariate relationships between baseline P3NP concentration and baseline physical function are depicted in Figure 1. P3NP levels explained 38% of the variance in total SPPB score, 16% of the variance in grip strength, and 30% of the variance in gait speed.
The adjusted associations between P3NP concentration and grip strength and physical function are shown in Table 4. All final models were adjusted for familial relatedness, field center, age, sex, height, and weight, and—as described in detail above in the statistical methods—additional covariates were assessed separately for each model. In final models, higher log-transformed P3NP levels were associated with significantly weaker grip strength, lower SPPB score, slower gait speed, and worse chair-rise performance at baseline (all p < .01; Table 4). Specifically, 1 SD higher log-transformed P3NP was associated with ~1 point lower SPPB score, 2 kg weaker grip strength, and 0.05 m/s slower gait speed. Additional adjustment for levels of inflammation (IL-6 and/or hsCRP) did not affect these associations. The associations between P3NP and grip strength, total SPPB score, and chair-rise performance were similar when models were stratified by sex or generation; however, the association with gait speed varied by sex and generation. While the associations were all in the same direction, the relationship between P3NP and gait speed was stronger in women compared with men, and in the Proband compared with the Offspring generation (data not shown). When the associations were examined stratified by age 65 and above (using the same final Model 1 as in Table 4), the associations with baseline SPPB score (β = −1.2points,p < .001), gait speed (β = −0.04m/s, p = .016), and chair-rise time with penalty (β = 9.6seconds,p < .001) were only significant in those aged 65 and older (Supplementary Tables 5a and 5b). However, for baseline grip strength, the magnitude of the beta coefficient was stronger in those younger than 65 years (β = −2.5kg; p = .007), compared with those older than 65 (β kg = −1.76, p < .001; Supplementary Tables 5a and 5b).
Table 4.
Association Between Log-Transformed P3NP Levels and Measures of Physical Function
| Final Model 1a | Final Model 2b | |||
|---|---|---|---|---|
| β ± SEf | p (additional covariates) | β ± SEf | p (additional covariates) | |
| Total SPPB score (points 0–12) | −0.9 ± 0.1 | <.0001 (eGFR) | −0.83 ± 0.13 | <.0001 (none) |
| Grip strength (kg) | −2.0 ± 0.4 | <.0001 (smoke, drink, eGFR, HTN) | −2.01 ± 0.48 | <.0001 (HTN, hsCRP, eGFR) |
| Gait speed (m/s) | −0.05 ± 0.01 | .0007 (none) | −0.04 ± 0.01 | .0033 (IL-6) |
| Chair-rise time (s) | 0.71 ± 0.26 | .0063 (none) | Same as Model 1 | |
| Chair rises/10 s | −0.46 ± 0.10 | <.0001 (none) | Same as Model 1 | |
| Grip strength changec (kg) | <0.00 ± 0.00 | .88 (eGFR, HTN) | Same as Model 1 | |
| Gait speed changec (m/s) | −0.05 ± 0.01 | <.0001 (stroke) | Same as Model 1 | |
| Chair-rise time changec (s) | <0.00 ± 0.01 | .5552 (stroke) | Same as Model 1 | |
| Chair rises/10 s changed | −0.31 ± 0.14 | .0325 (baseline value) | Same as Model 1 | |
| Chair-rise change categoricale | 0.41 ± 0.08 | <.0001 (none) | Same as Model 1 | |
Notes: CVD = cardiovascular disease; eGFR = estimated glomerular filtration rate; hsCRP = high-sensitivity C-reactive protein; HTN = hypertension; IL-6 = interleukin-6; P3NP = procollagen type III N-terminal peptide; SPPB = Short Physical Performance Battery. Smoke = smoking status (ever, current, or never); Drink = alcoholic drink status (>1 vs ≤1 drink/d).
aFinal model, base covariates: field center, sex, age, height, weight, and family relatedness—additional covariates were assed as described in the “Statistical Methods.” bFinal Model 2 includes all covariates in final Model 1, and additionally considered inflammatory cytokines. cGrowth curve phenotypes pre-adjusted for age, sex, and field center; age2 and base models are additionally adjusted for weight change from baseline to Visit 2. dChange in number of chair rises/10 s was modeled using a change score (Visit 2 – Visit 1) adjusted for baseline chair rises/10 s. eThe first 3 groups are based on tertiles of change in chair rise time, with the fourth group composed of those who were able to perform the test at baseline, but unable at follow-up. p-Value for this phenotype corresponds to an overall F-test for trend. Pairwise tests are given in Supplementary Table 4. fBetas are expressed per SD log-transformed P3NP (mean ± SD log-transformed P3NP was 1.98 ± .42; Table 1).
We also examined the association between baseline P3NP levels and change in physical function. In final models, higher levels of log-transformed P3NP were associated with greater declines in gait speed, but there was no relationship with change in grip strength (Table 4). Higher levels of log-transformed P3NP were not associated with worse change in chair-rise time (Table 4); however, when change in chair-rise time was analyzed in categories to include those who transitioned from being able to unable to perform the exam, higher log-transformed P3NP levels were significantly associated with categories of change in chair-rise time (p < .0001, trend). When pairwise tests among categories were examined, P3NP levels in those who transitioned to being unable to perform the test (Q4) were significantly different from all other quartiles (p < .0001), but no other comparisons among quartiles reached statistical significance (Supplementary Table 4). Just as in the baseline models, the association between P3NP and change in SPPB, gait speed, and chair-rise were only significant in those aged 65 and older; however, the association with grip strength decline was stronger in those younger than 65 years. Finally, as expected, those who only had cross-sectional data were older, less healthy, lower functioning, and had higher P3NP levels than those who were included in longitudinal analyses. However, this is likely to have biased our longitudinal findings to the null, as individuals who were not included in longitudinal analyses had higher P3NP levels, and given their poorer health, would have been likely to have experienced steep declines in physical function prior to death or being lost to follow-up.
Discussion
In the current study, we found that higher levels of plasma P3NP were strongly associated with weaker grip strength and lower physical function in older adults with a wide range of physical function. We also found that higher baseline levels of P3NP were associated with greater declines in gait speed. These associations remained statistically significant after adjustment for important covariates including age, sex, height weight, chronic disease prevalence, lung function, liver function, kidney function, and markers of inflammation. The associations between higher P3NP and worse SPPB score, gait speed, and chair-rise performance was only significant in those aged 65 and older; however, these outcomes were designed for older adults and have known ceiling effects in younger individuals (32). The association between P3NP and baseline and change in grip strength was actually stronger in those younger than 65 years, which is a more “upstream” outcome than the SPPB (and its components), and does not suffer from the same ceiling affects. Thus, higher P3NP appears to be a marker of poorer physical health in adults, regardless of age. Our findings require confirmation but raise the possibility that circulating P3NP concentration may be a biomarker for current and future physical health. P3NP levels may ultimately help to identify those with “subclinical disability,” who are at higher risk for future physical impairments, functional limitations, and disability.
To our knowledge, this is the first study to examine the association of P3NP with objectively measured physical function. The Framingham Offspring Study examined the cross-sectional relationship between P3NP levels and quadriceps muscle strength in 806 adults mean aged 57 and older and found no association (26). There are 2 important differences between our study and the Framingham study that may account for the discrepancy. First, the participants in the Framingham study were considerably younger (mean age: 58 vs 73 years). This is important because the loss of muscle accelerates with age, and most of the Framingham participants may not have experienced significant muscle tissue loss and turnover. Likewise, fibrosis of the kidneys, lungs, liver, and heart also increase with age; thus, the younger Framingham participants may have experienced less systemic fibrosis. The second major difference was that the Framingham study used a radioimmunoassay, while we used a sandwich ELISA—thus the differences in results may be partially attributable to assay methodology.
The association of P3NP levels with measures of physical function was particularly strong, even after adjustment for important covariates, including chronic disease status and inflammatory cytokine levels. For example, a 1 SD higher log-transformed P3NP concentration was associated with 1 point lower SPPB score, which is considered to be “substantially meaningful” (33); and a 0.05 m/s lower gait speed, which—depending on starting gait speed—has meaningful health consequences (34). P3NP was also strongly associated with chair-rise performance, an important phenotype in older adults (30), as well as longitudinal changes in gait speed and transitioning from being able to complete to unable to complete 5 chair-rises. Given the strength of these associations, the biological mechanisms underlying this association warrant further investigation, but we have some hypotheses. P3NP is released into circulation during the late stages of collagen synthesis and is an indicator of liver (19,20,35), lung (15,36), and heart (13,14,35) fibrosis and disease severity. Thus, P3NP may be a marker of “systemic fibrosis.” In the Cardiovascular Health Study, higher levels of P3NP were related to higher risks of incident CVD, heart failure (37), and mortality (13). In our study, P3NP levels were highly associated with cardiovascular and kidney disease traits, but not lung disease or function. Additionally, P3NP levels also increase during periods of rapid muscle growth, such as during adolescence (21), upon initiating a high-intensity resistance training intervention (22), and with anabolic hormone treatment (23). Instead, muscle mass decreases with age; thus, in older adults, higher P3NP levels may reflect the replacement of healthy skeletal muscle with fibrotic tissue (25). As in other tissues, this process may be driven by age-related impairments to the injury repair process, characterized by excess collagen deposition during extracellular matrix formation (38,39). These mechanisms warrant further investigation in future studies that have quantified other ECM proteins, and have more direct measures of fibrosis.
Our study has potential limitations. First, P3NP was measured at a single time point; thus, it is unclear if trajectories of P3NP will track with trajectories of physical function across the adult life span. We also did not have objective measures of physical activity and could not assess if activity status was a confounder or effect-modifier. We measured circulating levels of P3NP so the source of P3NP is unclear. Finally, we were unable to examine potential underlying mechanisms for the reported associations. Nonetheless, our study has several noteworthy strengths, including the extensive phenotyping of the LLFS cohort—which allowed us to carefully consider important covariates; the robustness of the P3NP assay used; the use of objective measures of physical function; and a relatively long duration of follow-up time.
In conclusion, our analysis provides preliminary evidence that circulating P3NP level may be associated with physical function status and functional decline over a 7-year period among older adults.
Funding
This work was supported by the National Institute on Aging (U19-AG063893-01, U01-AG023712, U01-AG23744, U01-AG023746, U01-AG023749, and U01-AG023755). A.J.S. was supported by a career development award from the and National Institute on Aging (K01 AG057726). M.M.M. was supported by the National Heart, Lung, and Blood Institute training grant at the University of Pittsburgh (NHLBI T32-HL083825-11).
Conflict of Interest
None declared.
Supplementary Material
References
- 1.Cesari M, Kritchevsky SB, Penninx BW, et al. Prognostic value of usual gait speed in well-functioning older people—results from the Health, Aging and Body Composition Study. J Am Geriatr Soc. 2005;53:1675–1680. doi: 10.1111/j.1532-5415.2005.53501.x [DOI] [PubMed] [Google Scholar]
- 2.Guralnik JM, Ferrucci L, Pieper CF, et al. Lower extremity function and subsequent disability. J Gerontol A: Biol Sci Med Sci. 2000;55:M221–M231. doi: 10.1093/gerona/55.4.M221 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Fried TR, Bradley EH, Williams CS, Tinetti ME. Functional disability and health care expenditures for older persons. Arch Intern Med. 2001;161:2602–2607. doi: 10.1001/archinte.161.21.2602 [DOI] [PubMed] [Google Scholar]
- 4.Janssen I, Shepard DS, Katzmarzyk PT, Roubenoff R. The healthcare costs of sarcopenia in the United States. J Am Geriatr Soc. 2004;52:80–85. doi: 10.1111/j.1532-5415.2004.52014.x [DOI] [PubMed] [Google Scholar]
- 5.Studenski S, Perera S, Patel K, et al. Gait speed and survival in older adults. J Am Med Assoc. 2011;305:50–58. doi: 10.1001/jama.2010.1923 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Seeman TE, Merkin SS, Crimmins EM, Karlamangla AS. Disability trends among older Americans: National Health and Nutrition Examination Surveys, 1988–1994 and 1999–2004. Am J Public Health. 2010;100:100–107. doi: 10.2105/ajph.2008.157388 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Federal Interagency Forum on Aging-Related Statistics. Older Americans 2016: Key Indicators of Well-Being. Federal Interagency Forum on Aging-Related Statistics. Washington, DC: Government Printing Office; 2016. [Google Scholar]
- 8.Zuo X, Luciano A, Pieper CF, et al. Combined inflammation and metabolism biomarker indices of robust and impaired physical function in older adults. J Am Geriatr Soc. 2018;66:1353–1359. doi: 10.1111/jgs.15393 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Pahor M, Guralnik JM, Ambrosius WT, et al. ; LIFE Study Investigators . Effect of structured physical activity on prevention of major mobility disability in older adults: the LIFE study randomized clinical trial. J Am Med Assoc. 2014;311:2387–2396. doi: 10.1001/jama.2014.5616 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Rejeski WJ, Brubaker PH, Goff DC Jr, et al. Translating weight loss and physical activity programs into the community to preserve mobility in older, obese adults in poor cardiovascular health. Arch Intern Med. 2011;171:880–886. doi: 10.1001/archinternmed.2010.522 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sierra F. Moving geroscience into uncharted waters. J Gerontol A Biol Sci Med Sci. 2016;71:1385–1387. doi: 10.1093/gerona/glw087 [DOI] [PubMed] [Google Scholar]
- 12.Stone PJ. Potential use of collagen and elastin degradation markers for monitoring liver fibrosis in schistosomiasis. Acta Trop. 2000;77:97–99. doi: 10.1016/s0001-706x(00)00118-2 [DOI] [PubMed] [Google Scholar]
- 13.Agarwal I, Glazer NL, Barasch E, et al. Fibrosis-related biomarkers and risk of total and cause-specific mortality: the cardiovascular health study. Am J Epidemiol. 2014;179:1331–1339. doi:101093/aje/kwu067 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mansour IN, Bress AP, Groo V, et al. Circulating procollagen type III N-terminal peptide and mortality risk in African Americans with heart failure. J Card Fail. 2016;22:692–699. doi: 10.1016/j.cardfail.2015.12.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Su Y, Gu H, Weng D, et al. Association of serum levels of laminin, type IV collagen, procollagen III N-terminal peptide, and hyaluronic acid with the progression of interstitial lung disease. Medicine (Baltimore). 2017;96:e6617. doi: 10.1097/MD.0000000000006617 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Gonzalez-Lopez L, Rocha-Muñoz AD, Olivas-Flores EM, et al. Procollagen type I and III aminoterminal propeptide levels and severity of interstitial lung disease in Mexican women with progressive systemic sclerosis. Arch Bronconeumol. 2015;51:440–448. doi: 10.1016/j.arbres.2014.06.018 [DOI] [PubMed] [Google Scholar]
- 17.Madahar P, Duprez DA, Podolanczuk AJ, et al. Collagen biomarkers and subclinical interstitial lung disease: the multi-ethnic study of atherosclerosis. Respir Med. 2018;140:108–114. doi: 10.1016/j.rmed.2018.06.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zheng M, Cai WM, Weng HL, Liu RH. ROC curves in evaluation of serum fibrosis indices for hepatic fibrosis. World J Gastroenterol. 2002;8:1073–1076. doi: 10.3748/wjg.v8.i6.1073 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Hu XR, Cui XN, Hu QT, Chen J. Value of MR diffusion imaging in hepatic fibrosis and its correlations with serum indices. World J Gastroenterol. 2014;20:7964–7970. doi: 10.3748/wjg.v20.i24.7964 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Valva P, Casciato P, Diaz Carrasco JM, et al. The role of serum biomarkers in predicting fibrosis progression in pediatric and adult hepatitis C virus chronic infection. PLoS ONE. 2011;6:e23218. doi: 10.1371/journal.pone.0023218 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Crofton PM, Wade JC, Taylor MR, Holland CV. Serum concentrations of carboxyl-terminal propeptide of type I procollagen, amino-terminal propeptide of type III procollagen, cross-linked carboxyl-terminal telopeptide of type I collagen, and their interrelationships in schoolchildren. Clin Chem. 1997;43:1577–1581. [PubMed] [Google Scholar]
- 22.Fragala MS, Jajtner AR, Beyer KS, et al. Biomarkers of muscle quality: N-terminal propeptide of type III procollagen and C-terminal agrin fragment responses to resistance exercise training in older adults. J Cachexia Sarcopenia Muscle. 2014;5:139–148. doi: 10.1007/s13539-013-0120-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bhasin S, He EJ, Kawakubo M, et al. N-terminal propeptide of type III procollagen as a biomarker of anabolic response to recombinant human GH and testosterone. J Clin Endocrinol Metab. 2009;94:4224–4233. doi: 10.1210/jc.2009-1434 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Goodpaster BH, Park SW, Harris TB, et al. The loss of skeletal muscle strength, mass, and quality in older adults: the Health, Aging and Body Composition Study. J Gerontol A Biol Sci Med Sci. 2006;61:1059–1064. doi: 10.1093/gerona/61.10.1059 [DOI] [PubMed] [Google Scholar]
- 25.Curcio F, Ferro G, Basile C, et al. Biomarkers in sarcopenia: a multifactorial approach. Exp Gerontol. 2016;85:1–8. doi: 10.1016/j.exger.2016.09.007 [DOI] [PubMed] [Google Scholar]
- 26.Berry SD, Ramachandran VS, Cawthon PM, et al. Procollagen type III N-terminal peptide (P3NP) and lean mass: a cross-sectional study. J Frailty Aging. 2013;2:129–134. [PMC free article] [PubMed] [Google Scholar]
- 27.Newman AB, Glynn NW, Taylor CA, et al. Health and function of participants in the Long Life Family Study: a comparison with other cohorts. Aging (Albany NY). 2011;3:63–76. doi: 10.18632/aging.100242 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Sebastiani P, Hadley EC, Province M, et al. A Family Longevity Selection Score: ranking sibships by their longevity, size, and availability for study. Am J Epidemiol. 2009;170:1555–1562. doi: 10.1093/aje/kwp309 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Guralnik JM, Simonsick EM, Ferrucci L, et al. A Short Physical Performance Battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol. 1994;49:M85–M94. doi: 10.1093/geronj/49.2.m85 [DOI] [PubMed] [Google Scholar]
- 30.Santanasto AJ, Glynn NW, Lovato LC, et al. ; LIFE Study Group . Effect of physical activity versus health education on physical function, grip strength and mobility. J Am Geriatr Soc. 2017;65:1427–1433. doi: 10.1111/jgs.14804 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Sebastiani P, Thyagarajan B, Sun F, et al. Age and sex distributions of age-related biomarker values in healthy older adults from the Long Life Family Study. J Am Geriatr Soc. 2016;64:e189–e194. doi: 10.1111/jgs.14522 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Balasubramanian CK. The community balance and mobility scale alleviates the ceiling effects observed in the currently used gait and balance assessments for the community-dwelling older adults. J Geriatric Phys Ther (2001). 2015;38:78–89. doi: 10.1519/jpt.0000000000000024 [DOI] [PubMed] [Google Scholar]
- 33.Perera S, Mody SH, Woodman RC, Studenski SA. Meaningful change and responsiveness in common physical performance measures in older adults. J Am Geriatr Soc. 2006;54:743–749. doi: 10.1111/j.1532-5415.2006.00701.x [DOI] [PubMed] [Google Scholar]
- 34.Miller ME, Magaziner J, Marsh AP, et al. ; LIFE Investigators . Gait speed and mobility disability: revisiting meaningful levels in diverse clinical populations. J Am Geriatr Soc. 2018;66:954–961. doi: 10.1111/jgs.15331 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Lee CH, Lee WC, Chang SH, Wen MS, Hung KC. The N-terminal propeptide of type III procollagen in patients with acute coronary syndrome: a link between left ventricular end-diastolic pressure and cardiovascular events. PLoS ONE. 2015;10:e114097. doi: 10.1371/journal.pone.0114097 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Lammi L, Kinnula V, Lähde S, et al. Propeptide levels of type III and type I procollagen in the serum and bronchoalveolar lavage fluid of patients with pulmonary sarcoidosis. Eur Respir J. 1997;10:2725–2730. doi: 10.1183/09031936.97.10122725 [DOI] [PubMed] [Google Scholar]
- 37.Agarwal I, Glazer NL, Barasch E, et al. Fibrosis-related biomarkers and incident cardiovascular disease in older adults: the Cardiovascular Health Study. Circ Arrhythm Electrophysiol. 2014;7:583–589. doi: 10.1161/CIRCEP.114.001610 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Mahdy MAA. Skeletal muscle fibrosis: an overview. Cell Tissue Res. 2019;375:575–588. doi: 10.1007/s00441-018-2955-2 [DOI] [PubMed] [Google Scholar]
- 39.Kolb M, Bonella F, Wollin L. Therapeutic targets in idiopathic pulmonary fibrosis. Respir Med. 2017;131:49–57. doi: 10.1016/j.rmed.2017.07.062 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.