Deregulation of the inflammatory response plays a major role in the age-related decline of physical performance. The causal pathway leading from inflammation to disability has not been fully clarified, but several researches suggest that interleukin-6 (IL-6) causes a reduction of physical performance in elderly through its effect on muscle function. In vitro studies demonstrated that IL-6 inhibits the secretion of insulin-like growth factor I (IGF-I) and its biological activity, suggesting that the negative effect of IL-6 on muscle function might be mediated through IGF-I. We evaluated the joint effect of IGF-I and IL-6 on muscle function in a population-based sample of 526 persons with a wide age range (20–102 yr). After adjusting for potential confounders, such as age, sex, body mass index, IL-6 receptor, and IL-6 promoter polymorphism, IL-6, IGF-I, and their interaction were significant predictors of handgrip and muscle power. In analyses stratified by IL-6 tertiles, IGF-I was an independent predictor of muscle function only in subjects in the lowest IL-6 tertile, suggesting that the effect of IGF-I on muscle function depends on IL-6 levels. This mechanism may explain why IL-6 is a strong risk factor for disability.
- muscle function
- InCHIANTI Study
- insulin-like growth factor I
there is a growing body of evidence suggesting that chronic inflammation is one of the most important biological mechanisms underlying the decline in physical function that is often observed over the aging process (27).
The plasma concentration of interleukin-6 (IL-6), a cytokine that plays a central role in inflammation, tends to increase with age. High serum levels of IL-6 predict disability in the elderly (10, 15,16). Although some preliminary data suggest that IL-6 is associated with accelerated sarcopenia, the mechanism by which chronic inflammation affects physical function has not been fully established.
Several studies suggest that insulin-like growth factor I (IGF-I) is an important modulator of muscle mass and function not only during the developmental period but across the entire life span (8,23). Recent findings of an epidemiological study performed in a large representative sample of older women show that low plasma IGF-I levels are associated with poor knee extensor muscle strength, slow walking speed, and self-reported difficulty with mobility tasks, thus suggesting a role of IGF-I in the causal pathway leading to disability in the elderly (9).
Interestingly, previous studies (12, 21) demonstrated that IL-6 inhibits the secretion of IGF-I and its biological activity and that IL-6 overproduction is a mechanism implicated in IGF-I and insulin-like growth factor-binding protein (IGFBP)-3 downregulation (21). Furthermore, in transgenic mice, an IGF-I deficiency caused by IL-6-related mechanisms determined growth impairment (11).
In light of such evidence, a potential effect of IL-6 and IGF-I in the regulation of the homeostatic mechanisms that maintain an adequate muscle mass can be proposed. For example, it can be hypothesized that an IL-6-mediated decrease in IGF-I production may be a potential mechanism by which chronic inflammation causes impaired physical function.
In the analysis presented here, we evaluated the relationship of plasma concentrations of IGF-I and IL-6 with muscle function in a population-based sample of older persons.
The analyses performed in this paper use data from the InCHIANTI Study, a prospective population-based survey of older persons, designed by the laboratory of Clinical Epidemiology of the Italian National Research Council of Aging (INRCA, Florence, Italy) and carried out in the Chianti geographic area of Tuscany, Italy. As previously reported, this database includes data from 1,453 participants (age range 20–102 yr) randomly selected from the residents in the two municipalities of Greve in Chianti and Bagno a Ripoli using a multistage stratified sampling method (14). The data collection started in September 1998 and was completed in March 2000.
The study population for this analysis consists of 526 subjects (mean age: male = 65 ± 15 yr; female = 66 ± 16 yr), only selected from the Greve in Chianti cohort for which the data processing of all information collected was completed. Of 805 participants, the Greve cohort, a complete clinical examination, was available from a subset of 653 individuals. Subjects affected by diabetes mellitus (n = 38) and major clinical cardiovascular diseases (n = 31) were excluded. Clinical information was obtained by routine laboratory analyses, anamnesis, and physical examination. Diabetes mellitus was diagnosed according to American Diabetes Association criteria (1).
Subjects were asked to provide the commercial names of all drugs that they had taken in the last 5 years. Subjects who were currently taking drugs known to interfere with IGF-I and IL-6 metabolism (n = 58) were excluded from the study.
All subjects gave informed consent to participate in the study, which was approved by the Ethical Committee of our Institution.
Weight and height were measured by standard technique. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height expressed in meters. Waist circumference was measured at the midpoint between the lower rib margin and the iliac crest (most often at the umbilical level) and hip circumference at the level of the greater trochanter. Both were measured to the nearest 0.5 cm with a plastic tape measure, and the waist-to-hip ratio was calculated.
Blood samples were collected in the morning after the participants had been fasting for at least 8 h. Glucose level was immediately quantified by an enzymatic colorimetric assay using a modified glucose oxidase-peroxidase method (Roche Diagnostics, Mannheim, Germany) and a Roche-Hitachi 917 analyzer.
Several 0.5-ml aliquots of serum were processed immediately and stored at −80°C and subsequently used for the assessment of hormones and cytokines. Plasma insulin (intra-assay coefficient of variation 3.1 + 0.3%; Sorin Biomedica, Milan, Italy) and free IGF-I (intra-assay coefficient of variation 3.8 + 0.4%; Diagnostic System Laboratories; Milan, Italy) concentrations (28) were determined by RIA. The degree of insulin resistance was calculated according to the homeostasis model assessment (HOMA; see Ref.22), which is widely considered a valid index for assessment of insulin resistance over a wide range of severity of this problem (6) and is highly correlated with insulin-mediated glucose uptake calculated by the euglycemic hyperinsulinemic glucose clamp (6, 22).
IL-6 and IL-6 receptors were quantified with immunoassay kits (BioSource Cytoscreen human IL-6 and human sIL-6R UltraSensitive kits, Camarillo, CA; see Ref. 5). The minimum detectable concentrations were 0.10 pg/ml for IL-6 and 0.8 pg/ml for IL-6 receptor, and the interassay coefficient of variation was 7% for both kits.
Genomic DNA was obtained from blood lymphocytes using a standard salting-out DNA extraction technique to evaluate the IL-6 −174 C/G promoter polymorphism. Amplification of the −174 C/G site was performed according to Olomolaiye et al. (24). The PCR amplification was followed by an overnight restriction digest of 15 μl PCR product with Nla III enzyme. The presence of a cytosine (C allele) at nucleotide −174 was revealed by the detection of an Nla III cutting site by electrophoresis in 2% agarose gel.
Muscle strength and power.
Isometric grip muscle strength was assessed using a hand-held dynamometer (Nicholas Manual Muscle Tester, model no. BK-5474; Fred Sammons, Burr Ridge, IL) following a standardize measurement protocol that has been shown to provide highly reliable data (2). Each muscle group was tested two times, and separate measures were obtained for the left and the right side for each hand. The best measure for the strongest side was used in the analysis.
Explosive muscle power of lower extremity in extension (physical work delivered to the external environment in a unit of time) was evaluated using the device proposed by Bassey and Short (3). Briefly, the extension movement takes 0.25–0.40 s in a push through 0.165 m against a flat pedal. At the end of the push, the leg is extended fully. The movement is made seated so that the forces are contained between the buttocks and the foot. The seat position is adjusted for leg length, and the push is transmitted by a lever and chain to spin a flywheel. The gearing is such that resistance to the movement remains nearly constant throughout the extension. The final angular velocity of the flywheel is measured by an optoswitch and is used to calculate the average leg extensor power in the push (3).
To approximate normal distributions, log-transformed values for plasma IL-6, triglycerides, insulin, and insulin resistance (HOMA) were used in the analyses. Differences in continuous variable between males and females were tested with the Student's t-test. Pearson product-moment correlations were calculated to test associations among variables. Because IL-6 serum level was an effective modifier of the relationship between IGF-I and muscle function, anthropometrical, functional, and metabolic measures were calculated for tertiles of IL-6 and compared using a one-way ANOVA. Furthermore, multivariate linear regression analyses were used to test the independent association of age, sex, free IGF-I and IL-6, and IL-6 genotype with muscle function within each IL-6 tertile. To smooth the huge difference in the ratio between young and old subjects, age was categorized in tertile for multiple linear regression analysis. Statistical analyses were performed using the SPSS software package.
Clinical characteristics of the study group are reported in Table1. Study participants were slightly overweight with a mainly central body fat distribution. Women had lower plasma IL-6 levels and lower handgrip strength and total power than men. In contrast, no gender difference in plasma IGF-I levels was found.
In the whole group (n = 526) age was negatively correlated with handgrip (r = −0.47; P< 0.001) and total (r = −0.59; P < 0.001) power. Plasma IL-6 levels were positively correlated with age and BMI and negatively correlated with total power and handgrip (Table2). Conversely, IGF-I was negatively correlated with age and BMI and positively correlated with total power and handgrip (Table 2). Both IL-6 (r = −0.19;P < 0.05) and IL-6 receptor (r = −0.11; P < 0.05) serum concentrations were negatively correlated with IGF-I serum concentrations. As expected, IL-6 and IL-6 receptor were positively correlated (r = 0.16;P < 0.01). No differences in IL-6 plasma levels according to IL-6 promoter polymorphism were found (P = not significant). Indeed, when stratifying subjects according to tertiles of plasma IL-6 levels, IL-6 promoter polymorphism was a significant predictor of IL-6 plasma levels (independently of age, sex, and BMI) in subjects at the highest tertile. In particular, lower plasma IL-6 levels were found in subjects carrying the allele C compared with no carrier subjects (3.72 ± 2.53 vs. 4.84 ± 3.83 pg/ml; P < 0.05).
The joint effect of IL-6 and IGF-I on muscle function was tested in linear regression models predicting, respectively, handgrip and lower extremity muscle power and adjusted for age, sex, IL-6 receptor serum concentration, IL-6 promoter polymorphism, and smoking (model 1 in Table 3). IGF-I and IL-6 were independent predictors of total power, whereas IL-6, but not IGF-I, was an independent predictor of handgrip. However, both in the model predicting handgrip and in the model predicting lower-extremity muscle power, a test for interaction between IL-6 and IGF-I was statistically significant (model 2 in Table 3). Furthermore, in the model including a term for the IL-6 times IGF-I interaction, the independent effect of IGF-I on handgrip became borderline statistically significant. To better investigate how IGF-I and IL-6 plasma levels reciprocally condition their effects on muscle function, all further analyses were performed stratified according to tertiles of the plasma IL-6 level.
Subjects with higher plasma IL-6 levels (3rd tertile) were older and, although the wide age range, had lower free IGF-I concentrations and lower total power and handgrip than subjects in the lowest tertile (Table 4). Furthermore, subjects with elevated plasma IL-6 levels (3rd tertile) had a greater BMI and higher IL-6 receptor levels and severity of insulin resistance than subjects in the lowest tertile. A significant association between IL-6, handgrip, and total power was found only in subjects in the third tertile of plasma IL-6 levels (Fig. 1). In contrast, significant association of free IGF-I with handgrip and total power occurred only in subjects at the lowest plasma IL-6 level tertile (Fig. 1). These findings suggest the existence of a relationship between IL-6 and IGF-I that reciprocally conditions their effect on muscle function.
To further examine this issue, we conducted multivariate linear regression analyses testing the independent association of age, sex, BMI, IL-6, IGF-I, IL-6 receptor, and genotype with handgrip and total power within each IL-6 tertile (Table 5). After adjusting for age, sex, and multiple covariates, IGF-I was independently associated with both handgrip and total power in subjects with the lowest IL-6 levels (1st tertile) but not in subjects with the highest plasma IL-6 levels. Conversely, plasma IL-6 levels were independently associated with both handgrip and total power only in subjects with the highest IL-6 plasma levels (3° tertile).
Our study provides evidence that higher plasma IL-6 levels and lower plasma IGF-I levels are associated with lower muscle strength and power. However, the reciprocal relationship between IL-6 and IGF-I in their joint effect on muscle function is more complex than previously understood.
The role of IL-6 in the development of disability in older person has been documented widely (10, 15, 26, 27). Elevated plasma IL-6 levels are associated with high mortality in the elderly (16), and higher plasma IL-6 levels are often found in older persons who are disabled in activities of daily living (10). It has been proposed that the high risk of disability associated with high IL-6 serum levels is explained by the catabolic effect of IL-6 on muscle, which results in accelerated sarcopenia. However, the true effect of IL-6 on muscle has not been investigated fully.
IL-6 plays a central role in the inflammatory response (4). In addition to its multiple effects at inflammation sites, IL-6 also induces the synthesis of the hepatic acute phase inflammation proteins, such as C-reactive protein, haptoglobin, and fibrinogen, while inhibiting the synthesis of others, such as IGF-I (4).
IGF-I has an impact on muscle mass and function (8, 9,23). In regard to the latter, a population study including frail and healthy older women demonstrated that low plasma IGF-I levels were associated with poor knee extensor muscle strength, slow walking speed, and self-reported difficulty with mobility tasks, suggesting a role for IGF-I in disability (9).
The role of IL-6 in the regulation of IGF-I action has been documented widely (11, 12, 21). In particular, in nontransgenic mice, with all the potential difference vs. humans, treatment with IL-6 resulted in a significant decrease in IGF-I levels, associated with a growth defect partially reversed by the administration of a monoclonal antibody of the murine IL-6 receptor (11). Moreover, Fernandez-Celemin and Thissen (13) demonstrated that IL-6 controls IGF-I biological activity through a stimulation of hepatic IGFBP-4 gene expression. In humans, contrasting findings on the potential role of IGF-I on muscle function have been reported (9,7, 17, 19, 20, 25, 26). In fact, a cross-sectional study in 245 healthy elderly women showed that plasma IGF-I levels did not explain the age-associated decline of muscle function, a phenomenon that is considered part of normal human ageing (7). Papadakis et al. (26) found a significant univariate association of plasma IGF-I with handgrip, knee flexor, and extensor strength; however, after adjusting for age, such associations were no longer statistically significant (17). In contrast, in 140 older persons selected for being healthy and functionally active, lower circulating levels of IGF-I have been found to be related to lower values of quadriceps maximal muscle power and optimal shortening velocity, independent of age and anthropometric measures (20). Moreover, “in vitro” data suggest that older muscle stimulated with mechanical overload shows less activation of the satellite cells and develops less hypertrophy as a result of lower expression of the autocrine form of IGF-I and of the inability to upregulate IGF-I receptor (25).
To our knowledge, the joint effect of IL-6 and IGF-I levels on physical function has not been investigated previously. Our study provides evidence of a multifaceted relationship between plasma IL-6 and IGF-I that conditions the effect of these two mediators on muscle function. Based on the findings of our analysis, and in accordance with some literature, it is reasonable to hypothesize that a high level of inflammation, documented by high plasma IL-6 levels, might negatively affect muscle function through three different mechanisms as follows:1) IL-6 directly affects muscle strength; 2) IL-6 inhibits the synthesis of IGF-I; and 3) IL-6 blocks the effect of IGF-I.
These possibilities, although not definitively proved by our cross-sectional analysis, are supported by the following findings of our study. 1) IL-6 was an independent predictor of handgrip and muscle power, especially in subjects with the highest tertile of plasma IL-6 levels, suggesting that the effect of IL-6 on muscle becomes important only above a certain threshold concentration. This finding is compatible with other studies showing that the risk of future mobility disability in older persons with high plasma IL-6 level is evident only above a specific threshold concentration (15). 2) In the analysis stratified according to tertile of plasma IL-6 levels, subjects in the upper tertile had lower plasma IGF-I concentrations and lower total power and handgrip strength than subjects in the lowest tertile of plasma IL-6 levels.3) After adjusting for potential confounders, such as IL-6 receptor levels (18) and IL-6 genotype (29), IGF-I was independently associated with both handgrip and total power in subjects with the lowest plasma IL-6 levels (1st tertile) but not in subjects with the highest IL-6 plasma levels. This finding suggests that high IL-6 levels may downregulate the IGF-I receptors to the point that the effect of IGF-I on muscle is lost. On the contrary, the full effect of IGF-I on muscle is detected when IL-6 plasma levels are low.
The fact that high plasma IL-6 levels are associated with handgrip and total power independent of plasma IGF-I levels prompts us to hypothesize that other IL-6-related factors, in addition to IGF-I inhibition, might be involved. Among such factors, a potential role of IL-6 receptor (18) and IL-6 genotype (29) cannot be ruled out. In fact, IL-6 receptor has been shown to regulate both local and systemic IL-6-mediated events (18); furthermore, the IL-6 promoter polymorphism has been shown to be associated with different IL-6 plasma levels in healthy subjects and to affect the rate of IL-6 gene transcription (29).
Nevertheless, some caution should be used in interpretation of our data because of the fact that many other factors that proceed on to activity restriction such as diet and socioeconomic markers should be taken into account before weighing the potential role of IL-6 and IGF-I joint effects on muscle function.
In conclusion, our study suggests that, in older subjects with high circulating levels of IL-6, the production of IGF-I is reduced and the activity of the circulating IGF-I on muscle might be partially blocked. However, because our findings were obtained from cross-sectional data, this hypothetical cause-effect relationship can be emphasized but not definitely proven.
This is another important contribution to the emerging evidence that markers of inflammation are significant predictors of which elderly patients will develop frailty. The antagonistic link between markers of inflammation and decreasing IGF-I levels and action certainly provide a rationale for intervening to raise IGF-I levels in the frail elderly. These data also suggest why the results of such an intervention might be different between frail individuals and the “healthy elderly,” who have been the main subjects of the previous studies of growth hormone in older persons.
Our findings may also shed light on the controversial results reported by previous studies that examined the relationship between IGF-I and muscle mass and function. Further studies will be needed to confirm our findings in a longitudinal prospective and to gain insight into the pathophysiological mechanism by which IL-6 and IGF-I affect muscle function over the aging process.
Address for reprint requests and other correspondence: G. Paolisso, Dept. of Geriatric Medicine and Metabolic Diseases, IV Divisione di Medicina Interna, Piazza Miraglia 2, I-80138 Napoli, Italy (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published November 5, 2002;10.1152/ajpendo.00319.2002
- Copyright © 2003 the American Physiological Society