Beyond Inflammaging: The Impact of Immune System Aging on Age-Related Muscle Decline, Results From the InCHIANTI Study

Abstract

Aging is characterized by chronic low-level inflammation and is associated with geriatric syndromes such as sarcopenia and frailty. Our aim was to evaluate the longitudinal variation of muscle area, muscle quality, and muscle strength, relative to the variation of leukocyte-derived markers, and to assess the presence of a pathway of associations among derived leukocyte ratios, and the components of muscle health. The InCHIANTI is a longitudinal cohort study of aging that began in 1998 with follow-up visits every 3 years. Out of the 1 453 participants enrolled at baseline, this study includes 1 179 participants with complete data. Muscle strength was assessed by hand grip strength test, whereas muscle density and fat area were considered as indirect markers of muscle quality, derived from peripheral quantitative computed tomography of the calf. Muscle area was associated with neutrophil-to-lymphocyte ratio (NL-ratio), age, gender, comorbidities, and body mass index (BMI). Muscle density variation over time was inversely associated with age, comorbidities, and BMI, while being positively associated with monocyte-to-lymphocyte ratio (ML-ratio) and male gender. Fat area was inversely associated with age, interleukine-6 (IL-6), male gender, and NL-ratio, while being positively associated with ML-ratio, comorbidities, and BMI. Handgrip strength decreased with age, IL-6 levels, comorbidities, and NL-ratio, but increased with ML-ratio, being male, and having a higher BMI. In a path-analysis model, ML-ratio positively correlates with muscle mass, density, and strength, while NL-ratio only correlates inversely with muscle mass and density. NL-ratio and ML-ratio are associated with aging and may be implicated in age-related mechanisms that affect body composition and muscle strength. These ratios may represent a link between aging of the immune system and decline of muscle health with aging. However, further studies are needed to identify their usefulness for early detection of sarcopenia, myosteatosis, and frailty in the older adult.

Aging processes are characterized by chronic low-level sterile inflammation known as “inflammaging” (1). This condition is associated with several health issues, including cardiovascular and neurological diseases (2), as well as sarcopenia, which is the loss of muscle mass and strength (3). In addition, inflammation has been shown to increase the risk of frailty, disability, and ultimately mortality (4). One of the major causes of chronic inflammation is the imbalance between innate and adaptive immunity, which is considered one of the main features of immunosenescence (5). Immune cells overproduce and release various cytokines such as interleukine-6 (IL-6), -1 (IL-1-beta), tumor necrosis factor-alpha (TNF-alpha), and interferon-gamma (INF-gamma) (6). Beyond immunosenescence and inflammaging, there is an increased prevalence of loss in muscle mass and function with age, possibly due to repeated episodes of incomplete muscle regeneration occurring throughout adult life, and/or associated to disease- or age-related muscle wasting (7). The loss of muscle mass is frequently accompanied by fat deposition at the subcutaneous, inter- and intramuscular, and myocellular levels (8). Those conditions are associated with insulin ­insensitivity, inflammation, and skeletal muscle functional deficits (9). Changes in muscle quality may precede a loss of muscle mass (10). These age-related defects in muscle repair mechanisms can be intrinsic to the muscle, or result from changes in extrinsic regulatory systems. The regeneration of skeletal muscle after acute injury or during aging is heavily influenced by interactions with immune cells that invade and proliferate in the damaged tissue (11). It is important to note that the regulation of infiltrating inflammatory cells is a dynamic process influenced by the extent of muscle damage and the time required for repair (6). The response to damage, in the early phase, involves the innate immunity system; namely, IL-6, IL-15, and TNF-alpha increase rapidly, with a chemotactic effect on neutrophils infiltrating the extracellular space around the damaged area (11,12), with release of IL-1 and IL-8, which are chemoattractants for monocytes and macrophages (13). Macrophages go through different phases of activation, first with M1 phenotype (producing large amounts of cytokines and chemokines), and then switching to M2 phenotype with anti-inflammatory and promyogenic activities (14). Following the activation of the innate immune system, adaptive immunity is activated in response to muscle injury. In particular, T lymphocytes (specifically, regulatory T cells) infiltrate the muscle perilesionally, increasing the production of factors that promote and control satellite cell proliferation (6).

The neutrophil-to-lymphocyte ratio (NL-ratio) and the monocyte-to-lymphocyte ratio (ML-ratio) are composite markers derived from the absolute peripheral count of leucocyte; they have been used as markers of immune–inflammatory activities and specifically the balance between innate and adaptive immunity, which is dynamically remodeled with immunosenescence (5). Recently, in the InCHIANTI longitudinal study, The NL-ratio was shown to follow a temporal trend in positive association with aging (15), thus suggesting that the ratio could be considered a synthetic indicator of the immune system aging, and as a marker of healthy aging. Moreover, in the Enhanced Nathan Kline Institute-Rockland Sample, a large cross-sectional study, NL-ratio had a strong effect on changes in grip strength with aging, although the association became weaker and nonsignificant at high NL-ratio values. In other words, the NL-ratio modulated the downward trend in muscle function occurring with aging (16,17). The association of ML-ratio with strength and muscle mass has been less studied: it has been reported as biomarker of sarcopenia in a cross-sectional study of Chinese community-dwelling older people (18). In addition, ML-ratio has been studied as an effect modulator combined with sarcopenia in the progression and in the response to treatment of different types of cancer (19,20).

The interaction between the immune system and the muscle has been previously regarded as unidirectional (11), but it has been recently demonstrated that muscle cells have immunoregulatory properties, for example, myocytes exhibit major histocompatibility complexes, and display innate immunity receptors. Moreover, muscle cells can modulate the immunological microenvironment, through secretion of several mediators (21).

The association between structural or functional muscle features and candidate markers of immunosenescence has been addressed in several population studies. Most of these studies were cross-sectional analyses that provided inconsistent results, possibly due to the heterogeneity of proposed biomarkers and different characteristics of the populations studied.

In the Newcastle 85+ longitudinal study, no significant differences in muscle function and mass were found between senescent-like phenotype group (defined by higher frequency of CD4 and CD8 senescent-like effector memory cells) and the less-senescent phenotype (22). In another study of over 2 000 participants from the National Health and Nutrition Examination Survey (NHANES), C-reactive protein (CRP), but not NL-ratio, was inversely associated with peak force only in old male participants (23). In a smaller study of hospitalized kidney transplant patients, neither CRP nor NL-ratio ratio were associated with muscle function and mass (24). In a cross-sectional analysis of a community-based Chinese study, inflammatory indexes derived from leukocyte count were not associated with sarcopenia, regardless of the diagnostic criteria used. However, there was a slight correlation with muscle mass and handgrip strength (18). Finally, a meta-analysis of cross-sectional studies reported strong inverse associations between muscle strength, muscle mass, and muscle function with circulating CRP concentrations (25). While many cross-sectional analyses provide supportive evidence of the connection between markers of inflammation and muscle phenotypes, the relationship has not been fully explored in longitudinal studies.

This study aims to assess trajectories of leukocyte-derived markers as tools to study the pathophysiology of muscle aging processes. We used data from the InCHIANTI study, a large and representative study of the Italian population, and evaluated the longitudinal variation of muscle area, muscle quality, and muscle strength, relative to variation of leukocyte-derived markers; moreover, in a path-model analysis we also looked for the presence of a pathway of associations among derived leukocytes ratios, and the components of muscle health.

Method

Sample Description

The InCHIANTI study protocol has been described in detail elsewhere (26). Briefly, 1 453 participants aged 21–102 were recruited from 2 cities in Tuscany for a baseline visit conducted between 1998 and 2000. The participants were followed every 3 years with follow-up visits in 2001–2003, 2004–2006, 2007–2009, and 2013–2014. The study protocol was approved by the Italian National Institute of Research and in the United States the protocol was given an exemption status by the National Institutes of Health Intramural Research Program Institutional Review Board (Exemption #11976); informed consent was obtained from participants at each visit.

Laboratory Assay

Assessments of the number of red blood cells, white blood cells, platelets, hemoglobin concentration, and hematometric values were performed through an automated system at the Laboratory of Clinical Chemistry and Microbiological Assays, SS. Annunziata Hospital, Azienda Sanitaria 10, Florence, Italy, using a Hematology SE 9000 Autoanalyzer (Sysmex, Kobe, Japan, provided by DASIT, Milano, Italy) at baseline and follow-up 1, a Coulter LH 750 Hematology Autoanalyzer (Beckman Coulter Inc., Brea, CA) at follow-up 2 and 3, and a Sysmex XE 2100 (DASIT) at follow-up 4. NL-ratio and ML-ratio were derived from leukocytes absolute number count (27).

Serum high-sensitivity CRP was measured by immunonephelometric assay and monoclonal antibodies in duplicate with the Dade Behring BN II Nephelometer (Dade Behring Inc., Deerfield, IL). Baseline serum IL-6 was measured in duplicate ultra-sensitive ELISA (CytoScreen Human IL-6; BioSource International Inc., Camarillo, CA). A regression equation was applied to derive IL-6 sandwich ELISA results. Plasma IL-6 at the follow-ups was measured by a solid-phase high-sensitivity quantitative sandwich ELISA (Quantikine HS Human IL-6 Immunoassay; R&D Systems Inc., Minneapolis, MN) (28).

Body Mass Index and Muscle Strength

Body mass index (BMI) was calculated as weight (in kg) divided by height (in m²). Grip strength was measured with a handheld dynamometer (hydraulic hand BASELINE; Smith and Nephew). Participants were asked to perform the task twice with each hand, and the maximum strength attained during the 4 trials was used for the present analyses.

Tibial Peripheric Quantitative Computed Tomography

The peripheral quantitative computed tomography (pQCT) was performed by the XCT 2000 device (Stratec Medizintechnik, Pforzheim, Germany) (29). The images obtained from the pQCT were analyzed using the BonAlyse software (BonAlyse Oy, Jyvaskyla, Finland) (30).

The following bone parameters were derived from the pQCT images measured at 38% tibia length (31): calf muscle cross-sectional area (cm²), fat cross-sectional area (cm²), muscle density (mg/cm³); all the estimates are measured from a transverse scan performed at 38% of the tibia length from the distal tip of the tibia.

Diseases and Comorbidities

Based on self-report, the diagnosis of major medical conditions was ascertained according to preestablished criteria that combine reported doctor diagnosis, eventually supported by medical records, physical examination, blood tests, and drugs prescription (32). Comorbidities score was calculated by summing the number of diseases reported at baseline and all follow-up visits (angina, cancer, hepatic diseases, acute myocardial infarction, congestive heart failure, stroke, Parkinson’s disease, peripheral artery disease, diabetes, chronic obstructive pulmonary disease [COPD], asthma, and arthrosis).

Statistical Analysis

Baseline characteristics were compared across times of the study for the variables of interest, and differences among times were evaluated using analysis of variance for continuous variables. Association in the variation of: cross-sectional muscle area, cross-sectional fat area, muscle density, and handgrip strength, at baseline and at follow-up, were tested by mixed model analysis of covariance. Model A reported the unconditional means model, which evaluated just the random effect for the intercept without any predictors; Model B reported the unconditional growth model, which considers the effect of chronological age; Model C was the fully adjusted model, considering all the potential confounders. Based on the association found from linear mixed model, were built 2 models of path analysis: (i) looking at the association among chronological aging, NL-ratio, and ML-ratio (exogenous variables), and cross-sectional muscle area, muscle density, and handgrip strength (endogenous variables); (ii) a model with same exogenous variables (chronological age, NL-ratio, and ML-ratio) but analyzing the mediation role of cross-sectional fat area on handgrip strength. Path analysis is a statistical technique that executes a sequential multiple regression analysis. This method allows for the simultaneous solving of equations to break down the total effect of independent variables. The resulting effects are then categorized as either direct or indirect. Direct effects affect outcome variables without the need for an intermediary variable, while indirect effects affect outcome variables through an intermediary variable (33).

Results

Study results include data from 1 179 participants at baseline visit, 889 at first follow-up, 772 at the second, and 600 at the final follow-up visit. Except for CRP, absolute count of neutrophils and lymphocytes, all other variables were significantly different among the various times of the study (BMI, comorbidities, NL-ratio, ML-ratio, IL-6, muscle area, muscle density, subcutaneous fat area, handgrip, and monocytes; Table 1).

Table 1.

Descriptive of the InCHIANTI Study Population According to Time of the Study

	Baseline	Follow-up 1	Follow-up 2	Follow-up 3
	1 179	889	772	600	p Value
Age	67.94 ± 15.24	65.70 ± 14.97	63.55 ± 15.24	61.87 ± 15.05	<.001
Chronological age	67.94 ± 15.24	68.74 ± 14.98	69.63 ± 15.25	71.01 ± 15.07	<.001
Sex female	648 (54.96)	478 (53.77)	409 (52.98)	311 (51.83)	.19
Body mass index	27.15 ± 4.11	26.26 ± 3.96	27.04 ± 4.17	27.20 ± 4.15	.01
Comorbidities	1.29 ± 1.37	1.73 ± 1.61	1.92 ± 1.68	2.33 ± 1.74	<.001
NL-ratio	2.12 ± 0.93	1.98 ± 0.84	2.12 ± 1.13	2.10 ± 1.18	.05
ML-ratio	0.18 ± 0.08	0.20 ± 0.08	0.27 ± 0.11	0.28 ± 0.14	<.001
C-reactive protein	4.34 ± 7.11	3.75 ± 8.83	4.33 ± 8.63	3.51 ± 6.29	.21
Interleukin-6	3.27 ± 2.31	2.89 ± 2.48	3.50 ± 2.87	3.41 ± 2.62	<.001
Muscle area	36.77 ± 8.42	38.12 ± 8.58	36.62 ± 8.54	36.61 ± 9.62	<.001
Muscle density	72.62 ± 3.75	72.04 ± 3.60	72.87 ± 3.52	73.02 ± 3.67	<.001
Fat area	11.97 ± 7.82	12.08 ± 8.42	12.80 ± 8.00	13.53 ± 8.53	<.001
Handgrip	32.04 ± 13.98	35.63 ± 13.06	35.76 ± 13.60	30.61 ± 12.08	<.001
Monocytes	0.32 ± 0.11	0.38 ± 0.14	0.48 ± 0.16	0.50 ± 0.16	<.001
Neutrophils	3.71 ± 1.22	3.58 ± 1.13	3.60 ± 1.18	3.66 ± 1.28	.87
Lymphocytes	1.92 ± 0.65	1.96 ± 0.64	1.88 ± 0.70	1.98 ± 0.78	.37

Notes: Data were reported as mean ± SE; for trend p value was reported to assess differences among times of the study. Age represents the age at the enrollment, whereas chronological age was the effective age of the participants in the specific follow-up; comorbidities = sum of all the diseases reported by participants in every follow-up; NL-ratio = neutrophils (n, K/µL)/lymphocytes (n, K/µL) ratio; ML-ratio = monocytes (n, K/µL)/lymphocytes (n, K/µL) ratio; interleukine-6 (IL-6) was reported pg/mL; C-reactive protein—high sensitivity (µg/mL); muscle area (cm²), muscle density (mg/cm³), and fat area (cm²) were estimated at the 38% of the tibial length using peripheral quantitative computed tomography. Muscle strength was measured by handgrip test (kg).

Cross-sectional Muscle Area

The mean muscle area across individuals and time was 36.33 ± 0.23 (cm²). Within-person variance was 12.48 ± 0.37 (cm²), while between-person variance was 62.73 ± 2.68 (cm²; Supplementary Table 1; Model A). When the time effect (chronological age) was considered, muscle area declines with age on average by −0.20 ± 0.01 cm² per year. However, there was still 51.00 ± 1.04 cm² unexplained within-person residual variance. In the fully adjusted model, muscle area was inversely associated with NL-ratio (β ± standard error [SE]: −0.28 ± 0.11; p value = .02) and age (β ± SE: −0.15 ± 0.01; p value < .001); whereas muscle area was positively associated with male sex (β ± SE: 7.86 ± 0.35; p value < .001), comorbidities (β ± SE: −0.28 ± 0.08; p value < .001), and BMI (β ± SE: 0.59 ± 0.03; p value < .001; Figure 1 and Supplementary Table 1). No second-order interaction effect could be found between the variables considered in the model. Lastly, markers of inflammation as ML-ratio, IL-6, and CRP were not associated with muscle area.

Figure 1.

Linear mixed model (LMM) analysis, variation across time of the study for muscle area, muscle density, subcutaneous adipose tissue, and handgrip strength, according to sex, and to variation of neutrophil-to-lymphocyte ratio, monocytes-to-lymphocyte ratio, chronological age, markers of inflammation (interleukin-6, C-reactive protein), body mass index. The effects were reported as LMM estimates (β ± SE); estimates statistically significant were border in red; whereas in green nonstatically significant association were reported. Figure is created with Biorender (www.biorender.com, Toronto, ON, Canada) licensed to ADI.

Muscle Density

The mean muscle density across individuals and time was 72.23 ± 0.10 (mg/cm³). Within-person variance was 4.09 ± 0.12 (mg/cm³), while between-person variance was 10.58 ± 0.50 (mg/cm³; Supplementary Table 2; Model A). When the time effect (chronological age) was considered, muscle density declines with age on average by a mean of −0.12 ± 0.01 (mg/cm³) per year change. However, there was still 80.49 ± 0.26 (mg/cm³) unexplained within-person residual variance (Supplementary Table 2; Model B). In the fully adjusted model, muscle density was inversely associated with age (β ± SE: −0.15 ± 0.01; p value < .001), comorbidities (β ± SE: −0.12 ± 0.04; p value = .003), and BMI (β ± SE: −0.17 ± 0.01; p value < .001); whereas muscle density was positively associated with ML-ratio (β ± SE: 1.85 ± 0.55; p value < .001) and male sex (β ± SE: 0.82 ± 0.13; p value < 0.001; Figure 1 and Supplementary Table 2). There were no second-order interaction effect between the variables considered in the model. No markers of inflammation including NL-ratio, IL-6, and CRP were statistically associated with muscle area.

Cross-sectional Fat Area

The mean cross-sectional fat area across individuals was 12.24 ± 0.22 (cm²). Within-person variance was 4.95 ± 0.15 (cm²), while between-person variance was 61.68 ± 2.51 (cm²; Supplementary Table 3; Model A). When the time effect (chronological age) was considered, muscle cross-sectional fat area declined with age on average by a mean per year reduction of −0.04 ± 0.01 cm². However, there was still 198.51 ± 32.64 cm² unexplained within-person residual variance (Supplementary Table 3; Model B). In the fully adjusted model, cross-sectional fat area was inversely associated with age (β ± SE: −0.06 ± 0.01; p value < .001), IL-6 (β ± SE: −0.07 ± 0.03; p value = .01), male sex (β ± SE: −7.93 ± 0.37; p value < .001), and NL-ratio (β ± SE: −0.15 ± 0.07; p value = .02); whereas cross-sectional fat area was positively associated with ML-ratio (β ± SE: 4.06 ± 0.70; p value < .001), comorbidities (β ± SE: 0.20 ± 0.06; p value < .001), and BMI (β ± SE: 0.50 ± 0.03; p value < .001; Figure 1 and Supplementary Table 3). There were no second-order interactions observed between the variables considered in the model.

Handgrip Strength

The mean handgrip strength across individuals and time was 30.55 ± 0.36 (kg). Within-person variance was 35.87 ± 6.50 (cm²), while between-person variance was 152.20 ± 6.49 (kg; Supplementary Table 4; Model A). When the time effect (chronological age) was considered, a mean reduction of −0.65 ± 0.02 (kg) was observed. However, there was still 678.76 ± 89.17 (kg) unexplained within-person residual variance (Supplementary Table 4; Model B). The fully adjusted model considered all variables of interest. Handgrip strength was inversely associated with age (β ± SE: −0.47 ± 0.02; p value < .001), IL-6 (β ± SE: −0.34 ± 0.05; p value < .001), comorbidities (β ± SE: −0.56 ± 0.08; p value < .001), and NL-ratio (β ± SE: −0.33 ± 0.14; p value = .03); whereas handgrip strength was positively associated with ML-ratio (β ± SE: 2.92 ± 1.28; p value = .03), male sex (β ± SE: 15.11 ± 0.36; p value < .001), and BMI (β ± SE: 0.06 ± 0.04; p value < .001; Figure 1 and Supplementary Table 4, Model C). No second-order interaction effect could be found between the variables considered in the model.

PATH Analysis

The path model provided a good fit (goodness-of-fit index by Jöreskog and Sörbom = 0.99; SRMR = 0.033; Bentler–Bonett NFI = 0.98). The total, direct effects of chronological age, NL-ratio, ML-ratio, muscle area, and density on handgrip strength are provided in Figure 2. Both muscle area (β ± SE = 0.44 ± 0.01; p < .001) and muscle density (β ± SE = 0.17 ± 0.01; p < .001) had significant direct effects on handgrip strength; specifically, lower muscle mass and density were associated with lower strength. Aging indirectly affected the relationships between strength (β ± SE = −0.34 ± 0.01; p < .001), muscle mass (β ± SE = −0.25 ± 0.02; p < .001), and density (β ± SE = −0.50 ± 0.01; p < .001). Older patients with lower muscle mass and density had also a lower strength. Higher ML-ratio was directly associated with greater strength (β ± SE = 0.07 ± 0.02; p < .001), muscle mass (β ± SE = 0.06 ± 0.02; p = .003), and density (β ± SE = 0.07 ± 0.02; p < .001). Finally, NL-ratio was associated with muscle mass (β ± SE = 0.08 ± 0.02; p < .001) and density (β ± SE = −0.07 ± 0.02; p < .001), but not with handgrip strength (−0.02 ± 0.02; p = .94).

Figure 2.

Path analysis, associations among chronological aging, neutrophil-to-lymphocyte ratio (NL-ratio), and monocyte-to-lymphocyte ratio (ML-ratio; exogenous variables), and cross-sectional muscle area, muscle density, and handgrip strength (endogenous variables). In figure were indicated the effects of chronological age (black), NL-ratio (red), and ML-ratio (green) on endogenous variables; moreover, were reported the path-analysis estimates (β ± SE); p value were also reported according to this categorization: ***p value < .001; **p value < .01; *p value < .05. Figure is created with Biorender (www.biorender.com, Toronto, ON, Canada) licensed to ADI.

In a similar model exploring the associations between exogenous variables (chronological age, NL-ratio, and ML-ratio), subcutaneous adipose tissue area, and grip strength, chronological age acted directly on muscle strength and indirectly through subcutaneous adipose tissue, but both ML and NL ratios were not significantly associated with the subcutaneous adipose tissue (data not shown).

Discussion

The main results of this study demonstrated that the longitudinal NL-ratio and ML-ratio variations predict longitudinal changes in fat area and in muscle components (muscle area, muscle density, and strength) independently of age, sex, body composition, comorbidities score, and markers of inflammation. These associations were confirmed in an “a priori-conceptualized” path-analysis model, where our findings provide empirical evidence for the hypothesized relationships between age, NL-ratio, and ML-ratio with handgrip strength, and the effects are mediated by the area and density of the muscle.

Muscle is increasingly recognized as an organ with immune and endocrine competence; in effects, muscle cells express membrane-bound factors (major histocompatibility complexes and innate immunity receptors) (21), and release secretory cytokines that exert autocrine, paracrine, or endocrine effects (among the others: IL-6, IL-15, and TNF-alpha) (34). Mechanical loading and muscle injury activate neutrophils, which are the first wave of immune cells to infiltrate the muscle (35), to initiate phagocytosis of necrotic debris and activate a pro-inflammatory response (36). The subsequent step is the activation of monocytes/macrophages, with M1 phenotype in the early phase and M2 in a later phase to modulate myogenesis (14). Afterwards, adaptive immunity plays an important role in myogenesis, through the action of regulatory T cells and promotion of satellite cell proliferation (6).

The ML-ratio and the NL-ratio represent the (im)balance between innate and adaptive immunity. These indices are derived from leukocyte counts in circulation, and were proposed as markers of inflammation (27) and as prognostic factors and predictors of cancer (37), dementia (2), congestive heart failure, and other cardiovascular diseases (15).

In 2 different epidemiological studies, the InCHIANTI and the Rotterdam study, the NL-ratio was proposed as a reliable marker of biological age (15), predicting mortality risk and comorbidities in aging individuals (15,38). This suggests the ratio might be a fingerprint of immunosenescence and inflammaging.

The NHANES has provided aggregated cross-sectional data that offer evidence of racial disparities in the NL-ratio; the ratio was also associated with several chronic disease outcomes (39). However, in a study derived from NHANES, an association was found for CRP, but not for NL-ratio, with muscle strength (23). In a Turkish outpatient population of older participants, the multivariate analysis found cross-sectionally that the NL-ratio is an independent predictor of sarcopenia (40).

Our study aim was to evaluate the NL-ratio and ML-ratio as more than just reliable markers of inflammaging. We hypothesized that they could be part of the pathway that links aging with immunosenescence, and with the reduction of muscle quality and strength. Across the times of the study, changes of both ratios showed strong associations with the quality and strength of muscle. However, the NL-ratio was inversely associated with muscle mass, muscle density, and strength, while the ML-ratio worked in opposing direction, with a positive association.

We are unable to explain the opposite directions of the association of the 2 ratios with the variables under study. Objectively, however, the absolute number of monocytes, among the leukocytes considered in our study, are the only leukocyte type increasing during the course of the study in a statistically significant manner (about 40%, from enrollment to the last follow-up), whereas neutrophils and lymphocytes show minimal fluctuations, with a trend toward decreasing numbers for neutrophils and a slight increase for lymphocytes. In theory both ratios indicate the level of inflammation, or rather a disruption of the balance between innate and adaptive immunity (27). However, neutrophils are the first cell line to be activated in case of infection or local damage, and their action of short duration (12); in contrast, monocytes survive and are active for a longer time, and their increase both in absolute number as well as in percentage is much bigger (14), and following extensive muscle damage they can almost double their number (41).

Interestingly, our results showed that CRP and IL-6 did not influence the leukocyte ratio associations. In fact, CRP was not statistically significant in its association with muscle quality and function, and IL-6 was inversely associated with subcutaneous adipose area and muscle strength, but not with muscle area and density. Therefore, NL-ratio and ML-ratio are better predictors of quality and function of the muscle, and represent more reliable markers of muscle inflammation compared to CRP and IL-6.

Based on the results of path analysis, an alternative hypothesis can be derived, suggesting that exogenous variables, such as chronological age and the 2 ratios, are strongly associated with endogenous variables, including muscle area, muscle density, and muscle strength. More precisely, chronological age and ML-ratio have a direct impact on strength, as well as an indirect impact through the mediation of muscle area and density, while NL-ratio only affects strength through mediation of muscle area and density.

Taken together, our results confirm that NL-ratio and ML-ratio can serve as markers of inflammation, while also potentially linking immune system aging and aging-related muscle impairment. These ratios may therefore provide insights into the underlying mechanisms that contribute to muscle decline in older adults. These conclusions find support in the work by Broadbent et al., who reported that “integrated and interacting network genes are shared in blood neutrophils and the muscles” (42). In some detail, the authors found that in response to a physiological stimulus as ­exercise, at least 2 out of 3 genes strongly preserved are coexpressed both in muscles and neutrophils. They belong to the mitochondria-related gene network, and this suggests that their action is directed to maintain or regain cellular homeostasis (42).

As expected, aging was inversely correlated with muscle area, muscle density, subcutaneous adipose tissue area, and handgrip test, both in longitudinal analysis as well as in path analysis. These results could be explained by the fact that the distribution of fat tissue itself changes with aging, increasing visceral adipose tissue (43), intermuscular adipose tissue, intramuscular, and intramyocellular lipid, while decreasing subcutaneous adipose tissue (8). Adipose tissue can itself experience senescence (44), and this may contribute to the low-grade chronic inflammation seen in the microenvironment of a dysfunctional adipose tissue (45).

Limitations

This project has several limitations. NL-ratio, ML-ratio, and cytokines were derived from circulating level of leukocytes and this may not reflect what happens locally in the muscle; therefore, for all the markers we consider a generalization of the phenomenon. It is worth mentioning that path analysis cannot be used to prove causality; moreover, path analysis is a model-testing procedure, and cannot be used to develop models; therefore, models should always be based on theory and knowledge. Another limitation to be considered is the reverse causality bias (46,47), and the inverse association between age and tibial subcutaneous fat reported could be due to data distortion. Moreover, absolute white blood cell (WBC)-subpopulation counts were not considered in the model of analysis, because our group has previously demonstrated in the InCHIANTI study that the ratios are strong and independent predictors of age and age-related diseases (15). Finally, we assess muscle mass, muscle density, and subcutaneous fat area at the 38% of tibia length, whereas we use hand grip as a proxy generalization of muscle strength. But in a well-characterized cohort, handgrip strength and leg extension strength of older adults showed similar associations in gait speed (48).

Conclusion

Our data indicate that NL-ratio and ML-ratio are associated with the aging process, affecting body composition and muscle strength. The clinical implications of the 2 ratios go beyond the simple concept of markers of “inflammaging,” suggesting the possibility that they may represent a link between the aging of the immune system and the senescence processes affecting the muscle; that is, the ratios reflect events triggering for the activation of muscle repair phenomena. We are still far from being able to suggest a defined role of these ratios in the prevention and treatment of phenomena related to sarcopenia, myosteatosis, and ultimately to the frailty syndrome of the older adult. Further confirmation and validation studies are needed for appropriate age, sex, and race-specific reference cutoffs.

Funding

The InCHIANTI study was supported as a “targeted project” (ICS 110.1/RS97.71) by the Italian Ministry of Health and by the U.S. National Institute on Aging (contracts N01-AG-916413, N01-AG-5-0002, and N01-AG-821336, and grant R01-AG-027012). Supported in part by the Intramural Research Program of the National Institute on Aging.

Conflict of Interest

The authors declare no conflict of interest. L.F. is member of the Editorial Board of the Journal.

Data Availability

The data sets used and/or analyzed during the current study are available from the responsible authors for the InCHIANTI study (L.F.) on reasonable request. Data of the InCHIANTI study is available to all researchers upon justified request using the proposal form available on the InCHIANTI website (https://www.nia.nih.gov/inchianti-study, accessed on April 13, 2023).

Author Contributions

R.P.: conceptualization, interpretation of data, drafted the work. R.P.: conceptualization, interpretation of data, drafted and revised the work. A.D.I.: conceptualization, acquisition, analysis, drafted the work. S.B.: design of the work, interpretation of data, revised the work. A.M.: analysis, interpretation of data, drafted the work. G.I.: analysis, interpretation of data, drafted the work. E.S.: acquisition, analysis, interpretation of data, drafted the work. D.T.: acquisition, analysis, interpretation of data, drafted the work. T.T.: acquisition, analysis, interpretation of data, revised the text. L.F.: design, acquisition, analysis, interpretation of data, revised the text. All authors have read and agreed to the present version of the manuscript.

Ethics Approval

The InCHIANTI study baseline was approved by the Ethical committee at INRCA, Ancona (protocol 14/CE, February 28, 2000) as the FU1 (protocol 45/01, January 16, 2001). InCHIANTI study FU2 and FU3 were approved by the Local Ethical Committee at Azienda Sanitaria Firenze (protocol n◦ 5/04, May 12, 2004). The study was conducted in accordance with the Declaration of Helsinki, and approved by the Ethics Committee of INRCA in Ancona (Italy). Clinical Trial Registration: NCT01331512.

Consent to Participate

Written informed consent was obtained from the participants to participate at the Study at each time (baseline visit and follow-ups).