Inflammation-related indices and low physical performance in older adults with unintentional weight loss: a cross-sectional study

Abstract

Background

Sarcopenia is characterized by age-related loss of muscle mass and function and is associated with chronic low-grade inflammation (inflammaging). Novel inflammation- based indices – including the Uric acid to HDL-cholesterol ratio (UHR), Monocyte to HDL ratio (MHR), Triglyceride to HDL ratio (THR), C-reactive protein (CRP) to albumin ratio (CAR), CRP to HDL ratio (CHR), and Systemic immune-inflammation index (SII) – have emerged as markers of inflammaging. This study investigated the relationship between these inflammatory parameters and Low Physical Performance (LPP) in older adults.

Methods

490 patients aged 65 years and older who applied to the geriatric medicine outpatient clinic of a university hospital with complaints of weight loss were evaluated retrospectively cross-sectionally (2022–2023). LPP was assessed by SARC-F questionnaire, handgrip strength test (HGST), and the 5 times-sit-to-stand-test (STST), and patients were grouped into Low Physical Performance (LPP, n = 259) or Normal Physical Performance (NPP, n = 231) based on these criteria. UHR, MHR, THR, CAR, CHR, and SII were calculated from laboratory values.

Group differences in demographics, comorbidities, geriatric assessment scores, and these inflammatory markers were analyzed. The correlations between new inflammatory markers and standard inflammatory indicators (CRP, neutrophil) were evaluated. Receiver operating characteristic (ROC) analysis determined the ability of each parameter to discriminate LPP.

Results

The LPP group was older than NPP (median 76 vs 71 years, p<0.001) and had a higher prevalence of atrial fibrillation (p=0.002) and dementia (p<0.001), while other comorbidities were similar between groups. All inflammatory indices were elevated in the LPP group: median UHR 0.11 vs 0.09, MHR (higher in LPP), CAR 1.37 vs 1.02, CHR 0.13 vs 0.07 and SII 623.5 vs 479.5 (all p<0.001), and THR was slightly higher (2.19 vs 2.15, p=0.012). Serum uric acid, monocyte count and CRP levels were higher in LPP than in NPP group, while albumin and HDL levels were lower (all p<0.01). UHR, CAR, MHR and SII correlated with one another and with CRP and neutrophils (p<0.001 for all). In ROC analysis, UHR showed a modest discriminative ability for identifying individuals with low physical performance (AUC 0.638, 95% CI 0.586–0.690), demonstrating a moderate ability to discriminate participants with low physical performance at a cutoff value of 0.1204 (sensitivity 44%, specificity 83%). CAR and SII demonstrated similar modest performance (AUC 0.602 and 0.626, respectively), while THR had a weaker association (AUC 0.566). UHR performed best with 83% specificity, while CAR and SII performed best with 71% sensitivity but AUC values indicate only modest discrimination. In multivariate logistic regression analyses adjusting for confounding factors, the UHR, SII, CAR, and CHR remained independently associated with physical performance status (all p<0.05).

Conclusion

Older adults with low physical performance show higher UHR, MHR, THR, CAR, CHR and SII levels, reflecting an increased inflammatory state. Among these indices, UHR, CAR, and SII demonstrated modest discriminative ability, with UHR showing higher specificity and CAR/SII showing higher sensitivity; however, their modest AUC values indicate limited standalone clinical utility, and thus these indices should be considered supportive rather than independent screening tools for identifying older adults at risk for sarcopenia. Prospective studies with long-term outcomes are needed to confirm the predictive validity of these studies and to determine whether interventions targeting modifiable components (such as serum uric acid, HDL, CRP, albumin, and complete blood count parameters) can positively influence sarcopenia progression and functional decline.

Introduction

Sarcopenia is a progressive loss of skeletal muscle mass and strength associated with aging [1]. It has been linked to inflammaging – a chronic, low-grade pro-inflammatory state that develops during aging. Inflammaging is characterized by increased levels of inflammatory mediators and is implicated in age-related diseases such as type 2 diabetes, osteoarthritis, and sarcopenia. With advancing age, the prevalence of conditions related to metabolic dysfunction and inflammaging (e.g. diabetes, obesity, hypertension) rises, suggesting that virtually everyone may face sarcopenia and its complications [2, 3].

The inflammaging involved in the sarcopenia process is exacerbated by changes in glucose and lipid metabolism, insulin resistance, and oxidative stress, leading to increased production of inflammatory cytokines. In addition to primary aging, secondary causes of sarcopenia are associated with lipid metabolism disorders [4]. For example, replacement of type II muscle fibers with fat is a hallmark of sarcopenia, contributing to muscle atrophy [5]. Chronic low- grade inflammation in sarcopenia causes oxidative stress-induced redox imbalance and upregulation of pro-inflammatory mediators [1]. This chronic inflammatory milieu drives tissue degeneration in age-related diseases [6].

However, the contribution of the pro-inflammatory mediators involved to sarcopenia and other age-related conditions is still poorly understood [7]. Low physical performance (LPP), recognized as a core functional component of sarcopenia based on EWGSOP2 criteria, may reflect the impact of systemic inflammation on muscle strength and physical function in older adults. Recent studies have identified inflammatory biomarkers associated with inflammatory conditions that contribute to sarcopenia. These include Uric Acid/HDL-Cholesterol Ratio (UHR), C-Reactive Protein (CRP)/Albumin Ratio (CAR), Monocyte/HDL-Cholesterol Ratio (MHR), CRP/HDL-Cholesterol Ratio (CHR), Systemic Immunity-Inflammation Index (SII), and Triglyceride/HDL-Cholesterol Ratio (THR) [8–10]. These indices are derived from routine laboratory parameters and reflect systemic inflammation in an accessible and cost-effective manner. Given evidence that sarcopenia in older individuals may be a consequence of chronic inflammatory and metabolic derangements, assessing inflammatory biomarkers in those with reduced physical performance may provide valuable insights into the pathophysiology associated with sarcopenia. According to current research, this broad panel of inflammatory markers has not previously been examined collectively in older adults with functional impairments suggestive of sarcopenia risk.

Therefore, this study aimed to evaluate the levels and characteristics of inflammatory indices (UHR, MHR, THR, CAR, CHR, and SII) in adults aged 65 years and older with preserved or reduced physical performance, to determine whether these markers are elevated in the LPPgroup, and to investigate their potential roles as supportive indicators in the comprehensive assessment of sarcopenia in geriatric clinical practice. Furthermore, we aimed to better characterize the inflammation-related changes associated with impaired physical performance by comparing these parameters between the LPP group and those with normal physical performance, with and without weight loss.

Methods

Study population

This retrospective cross-sectional study was conducted at the Department of Geriatrics, Hacettepe University Hospital, following ethical approval from the ‘Hacettepe University Faculty of Medicine Health Sciences Research Ethics Committee’’ (Date: January 9, 2024; Approval No: 2024/01–01, SBA 23/411). Patients aged 65 years and older who presented to the geriatric outpatient clinic with complaints of weight loss between June 1, 2022, and June 30, 2023, were evaluated for inclusion. After ethical approval, patient data between March 1, 2024 and May 30, 2025 were analyzed. We applied the following inclusion criteria: patients aged ≥ 65 years who had undergone a comprehensive geriatric evaluation and documented weight loss. Since all of the patient population included in the study had a weight loss statement at presentation, in order to clearly evaluate the weight loss that could also affect LPP, patients with 5% and more weight loss in 6 months were accepted as patients who applied to the outpatient clinic with unintentional weight loss (UWL) [11].

Patients were excluded if they were (a) younger than 65 years, (b) had a known condition that could affect physical performance tests (e.g., chair stand test, hand grip strength test) such as severe dementia, neuromuscular disease, rheumatologic disease, corticosteroid therapy with.

those taking > 5 mg prednisolone or equivalent per day for the last 3 years or those with physical limitations to perform these tests (having had surgery that would prevent movement within the last 1–3 months or having infectious findings in both extremities), (c) had incomplete medical history or incomplete clinical data, or (d) had incomplete laboratory results to calculate inflammatory indices. Data from patients with stroke sequelae, severe osteoarthritis, or extrapyramidal movement disorders (such as Parkinson’s disease) who were disabled or immobile were also excluded. After applying the exclusion criteria, 490 older adults were included in the study.

Physical performance was evaluated according to the revised European Working Group on Sarcopenia in Older People 2 (EWGSOP2) consensus, which defines reduced physical performance as a key indicator of sarcopenia severity. Both upper and lower extremity function were assessed using the handgrip strength (HGS) test and the five-times sit-to-stand (5×STS) test, respectively. HGS was measured with a calibrated digital dynamometer, and the highest value of three attempts with the dominant hand was recorded. According to EWGSOP2 recommendations, sex-specific cut-off points were applied (< 27 kg for men and < 16 kg for women). The 5×STS test was used to assess lower limb performance, and completion time longer than 15 s indicated impaired performance.

Participants were classified based on their results on these tests. Those who scored below the sex-specific cut-off for HGS or required more than 15 s to complete the 5×STS were categorized as having low physical performance (LPP). Participants who met both criteria—HGS above the cut-off and 5×STS completed in less than 15 s—were classified as having normal physical performance (NPP).

This dual-assessment approach is consistent with the EWGSOP2 framework, ensuring inclusion of individuals with upper or lower extremity limitations (e.g., arthritis, pain, or balance impairment) while accurately reflecting overall physical performance capacity. The study flow and patient allocation are shown in Fig. 1.

Fig. 1.

Shows the study flow and patient allocation process

Patients with a SARC-F score ≥ 4 were considered at risk for sarcopenia. Muscle strength was measured using the HGST with a calibrated handheld dynamometer (TKK5401; Takei III Smedley Type Digital Dynamometer Takei Scientific Instruments, Tokyo, Japan) [12].

Measurements were made while participants were standing with their arms positioned parallel to the floor. The highest value of 3 repeated measurements was taken in the analysis. HGST < 16 and < 27 kg for women and men, respectively, were taken as the cut-off values to assess muscle strength [13]. In terms of the ability to predict outcomes associated with sarcopenia, poor physical performance was defined as walking speed ≤ 0.8 m/s during a 4 m walking test using a manual stopwatch [13]. When evaluating walking speed, the average of two measurements was taken. For each patient, we recorded demographic information (age, gender), comprehensive geriatric assessment results, comorbidities, and laboratory values at baseline assessment in Table 1. Comparison of geriatric evaluations in patients with and without UWL and differences between groups with new inflammatory markers are shown in Table 2. Common comorbid conditions included diabetes mellitus (DM), hypertension (HT), coronary artery disease (CAD), chronic kidney disease (CKD), atrial fibrillation (AF), and dementia. Geriatric assessment measures such as Clinical Frailty Scale (CFS), Katz Activities of Daily Living (ADL) index, Lawton Instrumental Activities of Daily Living (IADL) scale, standardized Mini-Mental State Examination (s-MMSE), 15-item Geriatric Depression Scale (GDS), and Mini Nutritional Assessment-Short Form (MNA-SF) were obtained from patient records. Laboratory values including C-reactive protein (CRP), serum albumin, serum uric acid, triglycerides, high-density lipoprotein cholesterol (HDL-C), 25-hydroxyvitamin D, complete blood count parameters [(CBC), total leukocyte count, hemoglobin, hematocrit, platelet count and differential counts] were obtained from the institutional database. These laboratory measurements were performed from fasting blood samples as part of the routine geriatric assessment of patients.

Table 1.

Shows baseline characteristics of the patients

Parameters	Total Group (n = 490)	Group with Normal Physical Performance (NPP, n = 231)	Group with Low Physical Performance (LPP, n = 259)		P value
Age (years)	73 (69–79)	71 (68–76)	76 (72–80)		< 0.001
Sex (n), female	325 (66.3%)	155 (67.1%)	170 (65.6%)
male	165 (33.7%)	76 (32.9%)	89(34.4%)		0.732
CFS	4 (3–5)	3 (3–4)	4 (3–6)		< 0.001
Smoker	127 (25.9)	62 (26.8%)	65 (25.1%)		0.660
Comorbidities
Diabetes mellitus	242 (49.4%)	120(51.9%)	122(47.1%)		0.284
Hypertension	365 (74.5%)	170(73.6%)	195(75.3%)		0.667
Coronary artery disease	153 (31.2%)	64(27.7%)	89(34.4%)		0.112
Chronic kidney disease	36 (7.4%)	13(5.6%)	23(8.9%)		0.168
Dementia	61(12.4%)	13 (5.6%)	48 (18.5%)		< 0.001
Atrial Fibrilliation	44 (9%)	11 (4.8%)	33 (12.7%)		0.002
Geriatric Syndromes
Katz ADL	5 (5–6)	6(6–6)	6(5–6)		< 0.001
Lawton IADL	8 (5–8)	8(8–8)	7(4–8)		< 0.001
MMSE	28 (25–29)	29(26–30)	27(23–29)		< 0.001
Yesavage GDS	2 (0–4)	1 (0–4)	2(0–5)		0.002
MNA-SF	13 (10–14)	13(11–14)	12(9–14)		< 0.001
SARC-F	2 (0–4)	1(0–2)	3(1–5)		< 0.001
Handgrip Strength Test	20.60 (16.30- 26.00)	23(19.10–30.00)	16.17(14.00–23.30.00.30)		< 0.001
				P value	p ( Bonferroni )
UHR (%)	0.10 (0.08–0.13)	0.09 (0.07–012.07)	0.11 (0.08–0.15)	< 0.001	< 0.001 **
MHR (%)	0.0098 (0.0075-	0.0090 (0.072-	0.0100 (0.01–0.01)	< 0.001	< 0.001**
	0.0130)	0.114)
THR (%)	2.16 (1.51–3.02)	2.15 (1.46–3.06)	2.19 (1.60–3.10)	0.012	0.072**
CAR (%)	1.13 (0.70–4.11)	1.02 (0.69–4.14)	1.37 (0.72–4.31)	< 0.001	< 0.001**
SII	552.84 (384.33-	479.47 (351.77-	623.51 (432.39-	< 0.001
	817.18)	684.18)	933.37)		<0.001**
CHR (%)	0.10 (0.05–0.31)	0.07 (0.05–0.27)	0.13 (0.06–0.37)	< 0.001	< 0.001**

Data were expressed as n (%) or median [interquartile range], unless otherwise indicated Data were expressed as n + STD or mean, unless otherwise indicated

CFS Clinical frailty score, ADL Activities of daily living, IADL Instrumental activities of daily living, MNA- SF Mini Nutritional Assessment- Short Form, MMSE-Mini-Mental State Examination, GDS Geriatric depression scale, SARC-f A Simple Questionnaire To Rapidly Diagnose Sarcopenia, BMI Body mass index, TUG Time up and go test, STST 5 times sit to stand test, UHR Uric acid-to-high-density lipoprotein cholesterol ratio, MHR Monocyte-to-high-density lipoprotein cholesterol ratio, THR total cholesterol-to-high- density lipoprotein cholesterol ratio, CAR C-reactive protein to albumin ratio, CHR C-reactive protein to HDL ratio, SII systemic immune-inflammation index, UWL Unintentional weight loss

**Bonferroni correction applied for six comparisons (adjusted significance threshold = 0.05/6 = 0.0083)

p: Mann–Whitney U test; p (Bonferroni): Bonferroni-adjusted p-value

Table 2.

Demonstrates general characteristics of patients with and without UWL

	without UWL (n = 309)	with UWL (n = 181)	P value
Sex (n), female male	200 (64.7%) 109 (35.3%)	125 (69.1%) 56 (30.9%)	0.373
ADL	6 (5–6)	6 (5–6)	0.002
IADL	8 (7–8)	8 (3–8)	0.002
CFS	4 (3–4)	4 (3–5)	< 0.001
MNA-SF	13 (12–14)	11 (8–13)	< 0.001
SARC-F	1 (0–3)	2 (1–5)	< 0.001
Low Physical Performance	152 (58.7%)	107 (41.3%)	0.034
UHR (%)	0.10 (0.08–0.13)	0.10 (0.07–3.44)	0.655
MHR (%)	0.0098 (0.0078–0.01269)	0.0100 (0.0075–0.0134)	0.743
THR (%)	2.21 (1.45–3.14)	2.07 (1.56–2.80)	0.396
CAR (%)	1.22 (0.71–4.73)	1.10 (0.70–3.44)	0.261
SII	528.70 (375.38–753.01.38.01)	599.91 (416.04–938.18.04.18)	0.009
CHR (%)	0.10 (0.05–0.36)	0.09 (0.05–0.26)	0.148

Data were expressed as n (%) or median [interquartile range], unless otherwise indicated

UWL Unintentional weight loss, CFS Clinical frailty score, ADL Activities of daily living, IADL Instrumental activities of daily living, MNA-SF Mini Nutritional Assessment- Short Form, SARC-F A Simple Questionnaire To Rapidly Diagnose Sarcopenia, UHR Uric acid-to- high-density lipoprotein cholesterol ratio, MHR Monocyte-to-high-density lipoprotein cholesterol ratio, ROC Receiver-operating characteristic, THR Total cholesterol-to-high-density lipoprotein cholesterol ratio, CAR C- reactive protein to albumin ratio, CHR C-reactive protein to HDL ratio, SII Systemic immune-inflammation index

Comprehensive geriatric assessment

All participants underwent a Comprehensive Geriatric Assessment (CGA). Basic activities of daily living (ADL) [14, 15] (0–6 points) and instrumental activities of daily living (IADL) were used to measure the independence and functional ability of the patients [16]. Basic ADLs consist of six activities; bathing, dressing, going to the toilet, incontinence, transferring and feeding. IADL consists of using the phone, shopping, preparing food, housework, doing laundry, transportation, taking medication and managing financial affairs. MNA-SF (0–14 points) was used to determine the nutritional status of the patients and scores below 11 were considered as malnutrition and malnutrition risk [17]. The weakness status of the patients was defined by the CFS (1–9 points) [18]. According to the CFS, patients who were level 4 and more were accepted as living with frailty. Geriatric syndromes (osteoporosis, dementia, depression, falls, and polypharmacy) identified by the CGA were also recorded. s-MMSE and GDS were performed to assess cognitive function and depressive symptoms, respectively [19–22].Depression was defined as a score of 5 or more. The SARC-F questionnaire was used to determine the risk of sarcopenia. The questionnaire screens patients for self-reported symptoms suggestive of sarcopenia, including lack of strength, walking assistance, getting up from a chair, climbing stairs, and falling. Each of the self-reported parameters has a minimum and maximum score of 0 and 2, with the largest maximum SARC-F score being 10, with a score of 4 or more being considered as a risk of sarcopenia [23].

Calculation of inflammatory indices

The inflammatory parameters were calculated for each patient from the laboratory data as follows:

• UHR (Uric Acid to HDL-C Ratio) = Serum uric acid level (mg/dL) ÷ HDL cholesterol level (mg/dL).

• MHR (Monocyte to HDL-C Ratio) = Absolute monocyte count (10³/µL) ÷ HDL cholesterol level (mg/dL).

• THR (Triglyceride to HDL-C Ratio) = Serum triglyceride level (mg/dL) ÷ HDL cholesterol level (mg/dL).

• CAR (CRP to Albumin Ratio) = C-reactive protein level (mg/L) ÷ serum albumin level (g/dL).

• CHR (CRP to HDL-C Ratio) = C-reactive protein level (mg/L) ÷ HDL cholesterol level (mg/dL).

• SII (Systemic Immune-Inflammation Index) = Neutrophil count × Platelet count ÷ Lymphocyte count (all counts from CBC, with neutrophils, platelets, lymphocytes in 10³/µL.

These indices were derived for each patient to quantify systemic inflammation and metabolic imbalance. They were chosen based on prior studies indicating their relevance in chronic inflammatory states and age-related diseases.

Statistical analysis

Statistical analyses were performed using IBM SPSS Statistics (Version 24 for Windows; IBM Corp., Chicago, IL, USA). We evaluated the distribution of continuous variables using the Kolmogorov–Smirnov test.

For continuous variables that followed a normal distribution, comparisons between the PS and NS groups were made using the Independent-Samples T- test, and results are presented as mean ± standard deviation. For continuous variables not normally distributed, we used the Mann–Whitney U test, and data are presented as median with interquartile range (IQR).

Categorical variables were compared using the chi-square test and are presented as number (percent). Spearman’s rank correlation test was employed to analyze the relationships between the inflammatory parameters (UHR, MHR, THR, CAR, CHR, SII) and other variables, including traditional inflammatory markers (CRP, Neutrophil) and clinical measures.

In order to demonstrate the discriminatory power of the diagnostic performance of new inflammation indices in determining LPP, we constructed ROC curves and calculated the area under the ROC curve (AUC) for each index. Areas with AUC < 0.6 were not considered significant to determine the cut-off value. Cutoff values for sensitivity and specificity were determined using the Youden Index [24]. A two-tailed p value < 0.05 was considered statistically significant for all analyses. Spearman Correlation analysis was recorded for correlation analyses, τ > 0.2 or < − 0.2 (|τ|>0.2) were considered to reflect the value of the underlying entity [25]. A Bonferroni correction was applied to reduce the likelihood of Type I error in multiple testing for the six inflammatory indices (adjusted significance level: p < 0.0083). In correlation analyses, a Benjamini–Hochberg (False Discovery Rate, FDR) correction was applied to control the false discovery rate due to multiple comparisons; q < 0.05 was considered significant. To compare the distribution of inflammatory indices between groups, kernel density plots were generated using the ggplot2 package in R. Each index was visualized with overlaid density curves for the LPP and NPP groups, shown in different colors for clarity. Optimal cut-off values determined by ROC analysis were indicated with vertical dashed lines on each plot. This approach allowed for visual assessment of distribution differences and the discriminative capacity of the cut-off values across groups. To further account for potential confounding effects, multivariable logistic regression analyses were performed to supplement the univariate comparisons. Covariates that demonstrated clinical relevance or exhibited statistically significant between-group differences in univariate analyses, including age, sex, dementia, atrial fibrillation, frailty status, nutritional status, and cancer diagnosis, were incorporated into the adjusted model. The magnitude and directionality of associations between inflammatory indices (UHR, MHR, THR, CAR, CHR, SII) and LPP were assessed through both unadjusted and fully adjusted odds ratios with corresponding 95% confidence intervals. In the adjusted models (Models 2 and 3), UHR and MHR values were multiplied by 100 due to their very small decimal range, to facilitate interpretability of the odds ratios. Model 1 used unscaled raw values. CAR and CHR were entered in their original units in all models.

Ethical considerations

The study protocol was approved by the ‘Hacettepe University Faculty of Medicine Health Sciences Research Ethics Committee’’ (Date: 09.01.24; Approval No: 2024/01–01, SBA 23/411). The study was conducted in accordance with the principles of the Declaration of Helsinki. Informed consent was not required from individuals due to the retrospective nature of the study using de-identified patient data. No interventions or blood draws were performed for the purposes of this research, and all data were derived from routine clinical care records.

Results

Study population characteristics

The study included 490 older patients (median age 73 years, range 65–95; 66.3% female). Of these, 259 patients (52.9%) were classified as having LPP group, and 231 (47.1%) were NPP group (Fig. 1). The LPP group was older than the NPP group (median age 76 [72–80] vs. 71 [68–76] years, p < 0.001). The proportion of females was similar between groups (65.6% in PS vs. 67.1% in NS, p = 0.732), indicating no sex difference in physical performance prevalence in our cohort.

Comorbidity profiles of the two groups were comparable in terms of chronic diseases such as diabetes mellitus (DM), hypertension (HT), coronary artery disease (CAD), and chronic kidney disease (CKD), dementia and atrial fibrilliation (AF) (Table 1). The prevalence of DM, HT, CAD and CKD did not differ between LPP and NPP groups (p > 0.10 for each). However, dementia and AF were more common in the LPP group: AF was present in 12.7% of LPP patients versus 4.8% of NS (p = 0.002), and clinically diagnosed dementia in 18.5% of LPP vs. 5.6% of NPP (p < 0.001) group. These findings suggest that LPP in older adults may associate with a higher burden of conditions like AF and dementia, even though other comorbidities were similar.

Since all patients included in the evaluation had weight loss, those with a weight loss of 5% and above in 6 months were accepted as UWL for classification, a total of 181 people had UWL (36.9%) and when subgroup analyses were performed, it was seen that those with UWL were more dependent in daily living and instrumental living activities (all p = 0.002), were more fraile than the other group and had a higher risk of malnutrition and sarcopenia (all p values < 0.001) (Table 2). However, the risk of LPP, including performance tests, was higher in the group without UWL (p = 0.034).

Among laboratory parameters (Table 1) serum uric acid, monocyte count, and inflammatory markers differed between groups, whereas lipid profiles showed differences in HDL but not triglycerides. Patients with LPP had higher median serum uric acid (5.70 vs. 5.10 mg/dL, p < 0.001) and higher monocyte counts (median 0.56 vs. 0.51 × 10³/µL, p = 0.002) compared to controls. They also had elevated CRP levels (median 6.79 vs. 3.59 mg/L, p = 0.001) and lower serum albumin (median 4.20 vs. 4.35 g/dL, p < 0.001), consistent with a more inflammatory and less nourished state in LPP patients. HDL cholesterol was lower in the LPP group (median 53 vs. 56 mg/dL, p = 0.001). In contrast, median triglyceride levels were not different (115 vs. 113 mg/dL, p = 0.565), and 25(OH) vitamin D levels tended to be lower in LPP (median 28.9 vs. 30.4 µg/L) but without statistical significance (p = 0.095).

Functional and geriatric assessment

As expected, patients in the LPP group had worse performance on physical tests and functional scales (Table 1). The LPP group was more frail when assessed by CFS (4 [3–6] vs. 3 [3, 4], p < 0.001). Although Katz (ADL) was the same in both groups as a median, there was a significant difference in maximal values (median 6 vs. 6, p < 0.001, and Lawton (IADL) was lower in the LPP group (median 7 vs. 8, p < 0.001), indicating that LPP patients.

were more dependent on daily and instrumental activities. Cognitive screening scores, s- MMSE, were lower in LPP (median 27 vs. 29, p < 0.001), consistent with a higher prevalence of dementia. The risk of malnutrition measured by MNA-sf was higher in LPP (median 12 vs. 13, p < 0.001) and depressive symptoms (according to GDS) were more pronounced (median 2 vs. 1, p = 0.002). The SARC-F questionnaire score, which is part of the sarcopenia case finding, was higher in LPP (median 2 vs. Objective physical performance measures reinforced these differences: median handgrip strength was lower in LPP (16.2 kg vs. 23.0 kg, p < 0.001), 4-m walking test was slower (5.00 s vs. 3.78 s for 4 m, p < 0.001), Timed Up and Go (TUG) was longer (12.03 s vs. 8.75 s, p < 0.001) and STST was slower (16.07 s vs. 12.00 s, p < 0.001).

These results confirm that the LPP group had physical dysfunctions consistent with LPP. In addition, the LPP group had a higher mean body mass index (BMI) (median 29.0 vs. 26.7 kg/m², p = 0.006), suggesting that some subjects with sarcopenic obesity may reflect the inclusion of patients with lower muscle mass (despite higher weight).

Inflammatory markers – group differences

All inflammatory indices examined were higher in the LPP group than in controls (Table 1). UHR was higher in LPP with a median of 0.11 compared to 0.09 in NPP (p < 0.001). MHR was also higher (median of 0.0100 vs. 0.089, p < 0.001). The median of CAR in LPP was 1.37, with the NPP median above 1.02 (p < 0.001). The CHR was also higher in LPP (LPP median 0.13 vs. NPP 0.07; p < 0.001 for group difference). SII was higher in LPP, median 623.5 (×10⁹/L) and 479.5 in NPP (p < 0.001). THR showed a smaller but statistically significant difference (LPP median 2.19 vs. NPP 2.15, p = 0.012). According to the Mann–Whitney U test results, UHR, MHR, CAR, CHR, THR, and SII values ​​were significantly higher in the LPP patient group than in NPP group (all p < 0.001). These differences remained significant after Bonferroni correction, except for THR (p < 0.0083). THR lost significance after Bonferroni correction. Figure 2 shows the distribution of selected inflammatory parameters between LPP and NPP groups. Overall, these findings suggest that older adults with LPP have an elevated inflammatory profile as reflected by these composite indices. Diagnostic performance of inflammatory indices is shown in Table 3. The largest relative differences were observed for UHR, MHR, CAR, CHR, and SII, all of which were elevated in LPP individuals, while THR showed only a modest increase.

Fig. 2.

Shows the distribution of inflammatory parameters between the ‘‘Low physical performance (LPP) and Normal physical performance (NPP) groups.’’

Table 3.

Presents ‘’Diagnostic performance of inflammatory indexes’’

Parameter	AUC (95%	P value	Sensitivity	Specificity	Cutoff
	CI)		(%)	(%)
UHR (%)	0.638	< 0.001	44	83	0.1204
MHR (%)	0.616	< 0.001	45	74	0.011226
CAR (%)	0.602	< 0.001	71	45	0.79
CHR	0.608	< 0.001	68	51	0.0708
SII	0.626	< 0.001	71	51	481.79
THR (%)	0.566	0.012	–	–	–

Abbreviations: AUC Area under the receiver-operating characteristic curve, CI Confidence interval, UHR Uric acid-to-high-density lipoprotein cholesterol ratio, MHR Monocyte-to-high-density lipoprotein cholesterol ratio, ROC Receiver-operating characteristic, THR total cholesterol-to-high-density lipoprotein cholesterol ratio, CAR C-reactive protein to albumin ratio, CHR C-reactive protein to HDL ratio, SII Systemic immune-inflammation index

When subgroup analyses were examined (Table 2), 181 (36.9%) of those who lost 5% or more weight in 6 months had LPP. While no significant difference was found between the groups with and without UWL in terms of gender (p = 0.373); the group with UWL was more dependent in daily and instrumental life activities (p = 0.002). Although the same values were.

seen in the table, the difference between the two groups was caused by the difference in the lower and upper cut-off values. The group with UWL was more frail (p < 0.001); and had a greater risk in terms of malnutrition and sarcopenia (p < 0.001). The risk of sarcopenia was significantly higher in patients admitted to the hospital with UWL (p = 0.034). A significant difference was observed only with SII among inflammatory markers such as UHR, MHR, THR, CAR, SII and CHR between the groups with and without UWL (p = 0.009). Although individuals with unintentional weight loss had higher frailty, poorer nutritional status, and greater sarcopenia risk, physical performance was relatively more preserved in this group. This pattern may reflect the fact that patients applying to the outpatient clinic with UWL were still functionally independent, less frail in daily life activities, and potentially in an earlier stage of weight-loss–related decline. As a result, despite higher SARC-F scores and lower MNA-SF values indicating risk, their objective performance measures may not yet have deteriorated to the same extent as in older adults without UWL.

Correlation and ROC analyses

The relationship between standard inflammation markers (such as CRP, uric acid, albumin, etc.) and new inflammation markers was evaluated. Table 4 presents the correlation coefficients between biomarkers related to inflammation parameters. The correlations between UHR, MHR, CAR and SII reached significant statistical significance (Spearman’s τ ranged from 0.57 to 0.93, all p < 0.001). For example, UHR showed a significant (p < 0.001) correlation with MHR (τ = 0.57) and THR (τ = 0.55). UHR also had the expected negative correlation with HDL-C (τ = − 0.76, p < 0.001) and a high positive correlation with serum uric acid (τ = 0.81, p < 0.001) (since UHR components include serum uric acid and HDL).

Table 4.

Presents the correlation coefficients between inflammatory parameters and related biomarkers

	τ value	P value	q (FDR)
UHR-HDL	−0.757	< 0.001	< 0.001 **
UHR-s Uric Acid	0.810	< 0.001	< 0.001 **
UHR-Neutrophil	0.235	< 0.001	< 0.001 **
UHR- MHR	0.570	< 0.001	< 0.001 **
UHR-THR	0.554	< 0.001	< 0.001 **
UHR-CHR	0.271	< 0.001	< 0.001 **
CAR- CRP	0.963	< 0.001	< 0.001 **
CAR- Albumin	−0.368	< 0.001	< 0.001 **
CAR-CHR	0.930	< 0.001	< 0.001 **
MHR-HDL	−0.757>	< 0.001	< 0.001 **
MHR-s Uric Acid	0.810	< 0.001	< 0.001 **
MHR- Neutrophil	0.235	< 0.001	< 0.001 **
MHR- CHR	0.271	< 0.001	< 0.001 **
MHR-THR	0.458	< 0.001	< 0.001 *
MHR-CAR	0.208	< 0.001	< 0.001 **
THR-DM	0.320	< 0.001	< 0.001 **
THR-HDL	−0.643	< 0.001	< 0.001 **
SII-Neutrophil	0.692	< 0.001	< 0.001 **
SII- CRP	0.102	0.024	0.03 **
SII-CAR	0.117	0.01	0.01 **
SII-UHR	0.148	0.001	0.001 **
SII-CHR	0.137	0.002	0.002 **
SII-MHR	0.200	< 0.001	< 0.001 **

Correlation findings are shown as Spearman correlation coefficient (unadjusted P value); significant correlations

with |τ| >0.2 or < − 0.2 are underlined, P < 0.05 is shown as significant

q: p-value adjusted using the False Discovery Rate (FDR) correction

**Bold values indicate correlations that remained statistically significant after FDR adjustment (q < 0.05)

HDL High-density lipoprotein cholesterol, CRP C-reactive protein, UHR Uric acid-to-high-density lipoprotein cholesterol ratio, MHR Monocyte-to-high-density lipoprotein cholesterol ratio, THR Total cholesterol-to-high-density lipoprotein cholesterol ratio, CAR C-reactive protein to albumin ratio, CHR C-reactive protein to HDL ratio, SII Systemic immune-inflammation index

Similarly, MHR had a negative correlation with HDL (τ = − 0.76, p < 0.001) as expected by definition, while it had a high positive correlation with uric acid (τ = 0.81, p < 0.001). There was also a high correlation with CAR and CHR (τ = 0.93, p < 0.001). MHR and CHR were also moderate correlated than other inflammatory indices (τ = 0.27, p < 0.001). The least correlation was generally seen between SII and other inflammatory indices. Correlation analyses revealed no correlation that lost significance after False Discovery Rate (FDR, Benjamini–Hochberg) correction (all q < 0.05, after FDR correction).

All of these new inflammatory indices were associated with classical inflammatory markers such as CRP and neuthrophil. For example, CAR is mathematically related to CRP and albumin; as expected, CAR was highly correlated with CRP level (τ = 0.96), while it had a moderate and expected inverse correlation with albumin (τ = − 0.37), reaching statistical significance (p < 0.001). MHR showed a moderate positive correlation with CHR (τ = 0.27, p < 0.001) and a similar correlation with CAR (τ = 0.20, p < 0.001). SII, which combines blood cell counts, showed a positive correlation with the neutrophil component (τ = 0.69, p < 0.001) and also with CRP (τ = 0.10, p = 0.024). These correlations confirm that higher values of UHR, MHR, CAR, CHR and SII reflect increased systemic inflammation, consistent with elevations in traditional inflammatory biomarkers. We assessed the potential utility of each inflammatory index in distinguishing patients with LPP from those without. Figure 3 shows the ROC curves for all inflammatory parameters in detecting LPP. Table 3 summarizes the results.

Fig. 3.

Demonstrates the ROC curves for all inflammatory parameters in detecting Low physical performance (LPP)

of the ROC analysis. UHR emerged as the most promising discriminator between the indices with AUC of 0.638 (95% confidence interval approximately 0.59–0.69, p < 0.001). The optimal cut-off value for UHR was approximately 0.1204, providing 44% sensitivity and 83% specificity for detecting PS. Although the sensitivity for UHR at this threshold is modest, the high specificity suggests that patients with a UHR above 0.12 are likely to have sarcopenia, although many sarcopenic patients will have a lower UHR. The AUC of MHR was 0.616 (p < 0.001). A cut-off value of approximately 0.011 for MHR provided 45% sensitivity and 74% specificity. The CAR index showed an AUC of 0.602 (p < 0.001) and had an optimum CAR cut-off of approximately 0.79 (given CRP in mg/L and albumin in g/dL). At this threshold, CAR had a sensitivity of 71% and a specificity of 45% for sarcopenia - CAR had lower specificity but much higher sensitivity compared to UHR and MHR. CHR had an AUC of 0.608 (p < 0.001); the best cut-off (0.07 mg/L per mg/dL unit) provided 68% sensitivity and 51% specificity. The AUC value of SII was 0.626 (p < 0.001), the optimum cut-off point was.

481.8 (corresponding to approximately 4.82 × 10¹¹ in SI units, combining neutrophils, platelets, lymphocytes), providing 71% sensitivity and 51% specificity. Although THR was significantly higher in those with LPP (p = 0.012), it was not considered significant in terms of sensitivity and specificity since the AUC value was < 0.6 when calculated, and cut-off values were not determined (Fig. 4). We explicitly acknowledge that the AUC values (0.60–0.64) reflect only moderate discriminatory ability with limited clinical utility. These modest AUC values indicate that the inflammatory indices have fair but not excellent performance in distinguishing LPP from NPP. According to standard interpretation guidelines for AUC values, 0.60–0.64 represents fair discrimination, falling short of the 0.70–0.80 range considered acceptable for clinical screening tools and well below the > 0.80 threshold for good to excellent discrimination.

Fig. 4.

Shows the density distributions of the six inflammatory indices—UHR, MHR, THR, CAR, CHR, and SII—overlaid for Low physical performance (LPP, teal) versus Normal physical performance (NPP, orange), with optimal cut-off points indicated by dashed vertical lines in each panel; no cut-off was calculated for THR because AUC < 0.6

Multivariable binary logistic regression analyses

In the unadjusted model (Model 1), higher UHR, SII, and CAR values were independently associated with an increased likelihood of LPP (UHR OR = 1.163, 95% CI 1.100–1.229, p < 0.001; SII OR = 1.090, 95% CI 1.037–1.145, p = 0.001; CAR OR = 1.204, 95% CI 1.057–1.372, p = 0.005). In contrast, higher CHR values were significantly associated with a lower odds of LPP (OR = 0.984, 95% CI 0.969–0.999, p = 0.035), suggesting that a higher lipid-related inflammatory ratio may be protective (Table 5).

Table 5.

Multivariable binary logistic regression analysis of low physical performance

	B	SE	Wald	P value	OR	95%CI
Unadjusted Model 1
UHR	0.151	0.028	28.787	< 0.001	1.163	1.100–1.229.100.229
SII	0.086	0.025	11.594	0.001	1.090	1.037–1.145
CAR	0.186	0.067	7.776	0.005	1.204	1.057–1.372
CHR	−0.016	0.008	4.445	0.035	0.984	0.969–0.999
MHR	0.086	0.085	1.007	0.316	1.089	0.922–1.288
Model 2
UHR  SII  CAR  CHR  MHR	0.149 0.068 0.168 −1.570 0.078	0.031 0.028 0.066 0.786 0.095	23.233 5.723 6.497 3.996 0.678	< 0.001 0.017 0.011 0.046 0.410	1.161 1.070 1.183 0.208 1.082	1.093–1.234 1.012–1.131 1.040–1.346 0.045–0.970 0.897–1.304
Age	0.029	0.018	2.495	0.114	1.029	0.993–1.067
Sex	−0.108	0.232	0.214	0.643	0.898	0.570–1.416
CFS	0.336	0.126	7.134	0.008	1.399	1.093–1.790
IADL(Lawton)	−0.167	0.064	6.7715	0.010	0.846	0.746–0.960
Model 3
UHR  SII  CAR  CHR  MHR	0.142 0.091 0.186 −0.016 0.052	0.033 0.031 0.078 0.010 0.092	18.478 8.644 5.703 2.560 0.324	< 0.001 0.003 0.017 0.110 0.569	1.152 1.096 1.205 0.984 1.054	1.080–1.230 1.031–1.165 1.034–1.403 0.965–1.004 0.880–1.261
Dementia	1.086	0.399	7.402	0.007	2.963	1.355–6.480
Cancer	0.480	0.286	2.811	0.094	1.617	0.922–2.834
AF	0.749	0.506	2.194	0.139	2.116	0.785–5.702
TUG	0.118	0.031	14.267	< 0.001	1.125	1.058–1.195
4 m Walking Test	0.011	0.051	0.047	0.829	1.011	0.915–1.117

Abbreviations: CI Confidence interval, UHR Uric acid-to-high-density lipoprotein cholesterol ratio, MHR Monocyte-to-high-density lipoprotein cholesterol ratio, CAR, C-reactive protein to albumin ratio, CHR C-reactive protein to HDL ratio, SII Systemic immune-inflammation index, CFS Clinical frailty score, IADL Instrumental activities of daily living, AF Atrial fibrillation, TUG Time up and go test, SE Standard error, OR Odd ratio, 95% CI 95% confidence interval

Model 1, unadjusted; Model 2, adjusted for age, sex, CFS and IADL; Model 3, adjusted for multimorbidities (dementia, cancer, AF) and physical performance tests (4 m walking test, TUG)

UHR & MHR ×100 in Models 2–3; CAR & CHR in original units

After adjustment for age, sex, frailty (CFS), and Lawton–Brody IADL (Model 2), the associations for UHR (OR = 1.161, 95% CI 1.093–1.234, p < 0.001), CAR (OR = 1.183, 95% CI 1.040–1.346, p = 0.011), and SII (OR = 1.070, 95% CI 1.012–1.131, p = 0.017) remained statistically significant, and CHR continued to be significantly associated with lower odds of LPP (OR = 0.984, 95% CI 0.969–1.000, p = 0.046).

In the fully adjusted clinical model (Model 3), which additionally included multimorbidity and objective performance measures (atrial fibrillation, dementia, malignancy, 4-meter walking test, TUG Test, the associations for UHR (OR = 1.152, 95% CI 1.080–1.230, p < 0.001), CAR (OR = 1.205, 95% CI 1.034–1.403, p = 0.017), and SII (OR = 1.096, 95% CI 1.031–1.165, p = 0.003) persisted, whereas the association for CHR was attenuated and did not reach statistical significance (OR = 0.984, 95% CI 0.965–1.004, p = 0.110). Of the clinical covariates, dementia (OR = 2.963, 95% CI 1.355–6.480, p = 0.007) and poor physical performance (Timed-Up-and-Go OR = 1.125, 95% CI 1.058–1.195, p < 0.001) were strongly associated with LPP.

To facilitate interpretability due to the small decimal range of UHR and MHR, these indices were multiplied by 100 in the adjusted models (Models 2 and 3), whereas CAR and CHR were analyzed in their original units. The shift in the magnitude of OR values across models therefore reflects differences in unit scaling rather than a change in effect direction.

Association between inflammation-based indices and LPP

In multivariable binary logistic regression analyses across three models (Model 1: unadjusted; Model 2: adjusted for age, sex, and functional measures; Model 3: additionally adjusted for comorbidities, cognition, and gait performance), the inflammation-based indices—CHR, UHR, CAR, and SII—showed varying associations with LPP. In the unadjusted model, UHR and CAR demonstrated significant associations (OR = 1.16 and 1.20, respectively), and these associations remained significant after adjustment for confounding variables in Models 2 and 3. CHR showed a consistently lower odds of LPP with OR values ≈ 0.98 across the models, reaching statistical significance in Model 1 and Model 2, but becoming non-significant after full adjustment in Model 3. SII showed OR values ≈ 1.09–1.10 across the models, maintaining a similar effect size throughout all adjustments. However, the absolute magnitude of effect estimates is modest, suggesting that inflammation-based ratios alone may not fully capture the multifactorial risk underlying impaired physical performance (Fig. 5).

Fig. 5.

This forest plot illustrates the effects of four health indicators (CHR, UHR, CAR, SII) on a health outcome across three different statistical models. The vertical dashed line (OR = 1) represents the “no effect” threshold; values to the left indicate protective effects while values to the right indicate increased risk. For each indicator, squares represent point estimates (odds ratios) and horizontal lines show 95% confidence intervals. The findings reveal that CHR shows a slight protective effect (0.984), UHR demonstrates a notable risk increase (1.15–1.17), CAR shows virtually no effect (≈ 1.0), and SII is associated with a moderate risk increase (1.05–1.10). The consistency of results across all three models supports the reliability of these findings

Discussion

Our study investigated the association between novel inflammation-based indices and poor physical performance (LPP) in older adults presenting to a geriatric outpatient clinic complaining of unintentional weight loss. To our knowledge, this is the first comprehensive study to collectively examine these inflammation-based biomarkers in relation to poor physical performance (LPP) in an elderly population. We assessed six indices (UHR, MHR, THR, CAR, CHR, and SII) in 490 participants aged 65 years and older and found that all were higher in the LPP group. This finding supports the role of systemic inflammation in the functional decline associated with sarcopenia persistence. Our cross-sectional design allowed us to identify associations between inflammation markers and LPP; however, it does not allow us to make causal inferences about the temporal order or direction of these associations.

UHR and CHR represent key metabolic-inflammatory interactions, as elevated UHR reflects increased uric acid together with reduced HDL, both associated with pro-inflammatory metabolic disturbances frequently present in hypertension, diabetes, and CKD [26–31]. Low-grade inflammation, reduced HDL function, and elevated CRP — common in aging — likely contribute to the higher UHR and CHR observed in LPP [32–34]. The similar sex distribution between groups suggests that sex-related HDL differences are unlikely to confound these findings [35]. Likewise, the higher CHR levels in LPP may reflect impaired lipoprotein function under chronic inflammation [36]. In parallel, the Systemic Immune-Inflammation Index (SII) captures immune-inflammatory activation via neutrophils, platelets and lymphocytes, which are central to inflammatory diseases such as ulcerative colitis, chronic obstructive pulmonary disease and cancer [37, 38].The elevated SII in LPP supports the concept that systemic immune-inflammatory stress contributes to muscle performance decline. Furthermore, the association between THR and diabetes reinforces the well-established interaction between metabolic dysregulation and inflammation in age-related muscle impairment [39]. Importantly, after applying Bonferroni and FDR corrections to minimize type I error inflation, the associations for all markers except THR remained statistically significant, and their correlations with inflammatory parameters were preserved, supporting the robustness of these findings.

Logistic regression models showed consistent associations for UHR and SII after adjusting for age, multimorbidity, and sex. This suggests that these indices reflect functional impairment independent of major geriatric confounding factors. However, the low magnitude of the observed effect sizes also suggests that systemic inflammation is unlikely to be the primary driver of early functional decline and that muscle-specific metabolic and microenvironmental factors may play a more central role in functional decline than systemic inflammation. Similarly, in our multivariate logistic regression analyses, adding physical performance indicators such as cancer diagnosis and walking speed did not significantly change the odds ratios of inflammatory indices, and walking speed itself was not independently associated with LPP. These findings suggest that inflammatory activity alone may not be a dominant determinant of early functional decline in this outpatient population, where subclinical and multifactorial mechanisms may be more significant than overt cachexia or advanced frailty [40, 41].

CAR has been proposed as a prognostic marker in inflammatory and malignant diseases, reflecting both inflammation and nutritional status [39, 40]. High CAR in LPP in older adults experiencing weight loss (a population at high risk of malnutrition) likely represents a combination of nutritional deficiencies and multimorbidity rather than a direct muscle-specific mechanism. The attenuation pattern of CAR across adjusted models further supports this: although CAR showed a positive association with LPP in the unadjusted model, this association remained statistically significant after adjustment for age, comorbidities, frailty and functional parameters (Model 2), but became attenuated and no longer significant in the fully adjusted model including mobility and physical performance measures (Model 3). This aligns with previous research showing that CAR mainly reflects systemic vulnerability and chronic disease burden rather than sarcopenia-specific muscle changes. For example, Yeung et al. reported that CAR showed a weak association with muscle strength but no independent relationship with EWGSOP2-defined sarcopenia in community-dwelling older adults [42]. Similarly, Xu et al. demonstrated that associations between CAR (and other ratios such as NLR and PLR) and sarcopenia disappeared after adjusting for age, sex, and BMI [43]. Capurso et al. also found that CAR reflects inflammatory and nutritional burden in hospitalized frail older adults, without direct association to sarcopenia itself [44].

A similar pattern was observed for CHR, which showed a significant association in the unadjusted model and after adjustment for age, comorbidities, frailty and functional parameters (Model 2), but became attenuated and non-significant after full adjustment in Model 3.The lower odds associated with CHR in our models may suggest that reduced HDL — common in malnutrition, low-grade inflammation, chronic disease and muscle wasting — could contribute more strongly than elevated CRP to the observed association. However, this interpretation should be considered exploratory, given the attenuation after full adjustment. This interpretation is consistent with the findings of Yeung et al., Xu et al., and Capurso et al., who reported that the discriminatory ability of CHR was limited after adjustment for major geriatric confounders. Therefore, elevated CHR in the context of LPP likely reflects broader cardiometabolic vulnerability rather than a protective metabolic state [42–44].

The modest AUC values ​​observed for these indices (0.60–0.64) indicate limited independent discriminatory value for identifying LPP. Therefore, these markers should only be interpreted as supplementary laboratory indicators alongside direct measures of muscle strength and performance. For all these reasons, inflammation-induced ratios alone are unlikely to serve as a powerful diagnostic or screening tool but may have complementary value when interpreted in conjunction with biochemical and functional assessments.

In our study, the unintentional weight loss (UWL ≥ 5% over 6 months) group was expectedly more dependent in activities of ADL and IADL, more frail, and at greater risk of malnutrition and sarcopenia (all p < 0.001), reflecting a clinically vulnerable population (Table 2). Although such characteristics are typically associated with increased inflammatory activation and could theoretically amplify the observed associations between inflammation-based indices and lower physical performance (LPP), this did not meaningfully affect our main findings. UWL was present only in a limited subgroup (36.9% of the cohort), which likely reduced the impact of weight loss–related inflammation on systemic biomarkers.

Interestingly, individuals with UWL demonstrated higher physical performance compared to those without significant weight loss (LPP: 36.9% vs. 58.7%, p = 0.034), suggesting that UWL did not drive LPP classification in this outpatient population. Although all inflammatory parameters were numerically higher in the UWL subgroup, as expected, only SII remained statistically elevated (p = 0.009), whereas UHR, CHR, CAR, and MHR did not differ significantly between groups (Table 2). This pattern is consistent with previous reports showing that weight loss in older adults may be associated with increased inflammatory signaling and early catabolic changes, particularly reflected in neutrophil–lymphocyte–platelet interactions rather than lipid-associated ratios [45]. Likewise, early unintentional weight loss in geriatric populations often reflects a preliminary phase of nutritional decline, in which functional capacity and mobility remain preserved despite nutritional risk and frailty indicators [46]. The similar sex distribution across groups and preserved independence in part of the UWL cohort also support that early weight loss may not immediately translate into functional impairment [13, 46]. Accordingly, these findings suggest that the inflammatory differences observed between groups largely reflect functional vulnerability rather than an isolated effect of weight loss itself [45, 46].

Consistent with this interpretation, dementia (OR = 2.963, 95% CI 1.355–6.480; p = 0.007) and impaired functional mobility reflected by TUG performance (OR = 1.125, 95% CI 1.058–1.195; p < 0.001) remained strongly associated with low physical performance in the fully adjusted model, indicating that the higher LPP prevalence in the non-UWL group likely reflects greater functional vulnerability rather than the absence of weight loss.

Although UWL is frequently considered a marker of vulnerability in older adults, several mechanisms may explain why low physical performance (LPP) was more prevalent in the non-UWL group, despite greater nutritional and frailty risk in the UWL cohort. First, weight loss in geriatric outpatients may often represent an early phase of catabolic decline, where low-grade chronic inflammation (“inflammaging”) begins before overt loss of lean mass or measurable performance impairment appears, placing patients at nutritional risk without functional deterioration [45].Second, early weight loss in older adults typically leads to a preferential reduction of adipose tissue rather than muscle mass, resulting in minimal decline in gait speed or mobility despite clinically relevant weight change [46].Third, unintentional weight loss has heterogeneous etiologies in older adults, including medication effects, depressive symptoms, reduced appetite, subclinical disease, and transient inflammatory states, and therefore does not necessarily represent a highly catabolic, muscle-wasting process such as cachexia [47].In such situations, clinical screening tools such as SARC-F or MNA-SF may identify nutritional risk without a corresponding decline in objective physical performance, particularly in patients who are cognitively preserved and aware of weight changes but remain mobile and largely independent.

Taken together, these observations suggest that the differences in inflammation-based biomarkers observed between groups primarily reflect functional impairment itself rather than inflammation secondary to weight loss, and that UWL alone does not explain the association between inflammatory indices and LPP in our multivariable models. Longitudinal studies including both weight-losing and weight-stable populations will be essential to confirm the consistency and generalizability of these findings.

In addition to UHR and CAR, MHR has gained attention as an indicator of chronic low-grade inflammation and cardiometabolic risk. Elevated MHR has been associated with poorly controlled hypertension, Metabolic dysfunction–associated steatotic liver disease (MASLD), coronary heart disease and CKD, supporting its role as a marker of systemic inflammation [28, 32, 33]. In our study, higher MHR levels in older adults with LPP, and its relationship with CVD, are consistent with these findings. However, the absence of significant associations between MHR and HT, DM or CKD in our population may reflect the clinical characteristics of our outpatient cohort, which included relatively fitter older adults rather than frail, hospitalized patients where systemic inflammation is typically more pronounced.

Despite the strengths of this study, several methodological considerations require attention when interpreting these results. Consistent with the EWGSOP2 framework, we classified LPP based on impaired performance in either HGS or the 5×STS test; however, reduced 5×STS performance may be influenced not only by muscle weakness but also by conditions such as arthritis, pain-related mobility limitations, or balance disorders. Consequently, our operational definition likely captured a broader and more heterogeneous functional impairment phenotype than the muscle strength-based definition of “probable sarcopenia”, which may have diluted muscle-specific inflammatory associations. Furthermore, although inflammation-derived ratios showed statistically significant associations with LPP, the modest AUC values observed in ROC analysis (0.60–0.64) indicate limited standalone clinical utility. UHR demonstrated relatively high specificity and CAR/SII moderate sensitivity; nevertheless, these markers should be considered complementary indicators rather than independent tools for sarcopenia-related functional screening.

Bonferroni correction was applied to address the risk of inflated type I error resulting from multiple biomarker comparisons, and the persistence of significance for all inflammatory markers except THR supports the robustness of these finding. However, it should also be recognized that all participants were evaluated due to unintentional weight loss, a condition frequently associated with multimorbidity and systemic inflammation [13, 48]. Clinically significant weight loss (≥ 5% over 6 months) occurred in only a minority of the cohort, which reduces the risk of confounding. However, this limitation restricts the generalizability of the findings to community-dwelling older adults without weight loss and to those with sarcopenic obesity, in whom biomarker behavior may differ.

Additionally, the sample size in our study may not have been sufficient to determine optimal cut-off thresholds for biomarkers such as UHR and SII. Larger cohorts are needed to better define discriminative boundaries for clinical application. Finally, it remains unknown whether modification of these inflammatory components, including reduction of uric acid levels, improvement of HDL function, or lowering of CRP concentrations, can beneficially influence physical performance. This question should be addressed in future interventional studies.Lastly, the retrospective cross-sectional design does not allow for conclusions regarding temporality or causality. It remains uncertain whether elevated inflammatory parameters contribute to declining muscle function or whether reduced activity, metabolic dysregulation, and nutritional compromise subsequently drive inflammatory activation. Well-designed, longitudinal, multicenter studies with repeated assessments are needed to determine whether changes in these biomarkers track the progression from preserved strength to LPP and eventually to confirmed sarcopenia, and whether therapeutic modulation of uric acid, CRP, HDL metabolism, or albumin alters clinical outcomes in older adults.

Taken together, inflammation-related indices were consistently higher in older adults with LPP, suggesting that metabolic-inflammatory dysregulation contributes to functional decline. While UHR and SII emerged as the most robust predictors (retaining significance after correction and adjustment), neither demonstrated sufficient discriminatory accuracy for independent clinical use. Rather, these indices should be considered as supplementary measures complementing direct assessments of muscle performance in geriatric practice. In summary, our findings expand the existing evidence linking chronic low-grade inflammation to reduced physical performance in aging. However, inherent limitations of our study (including heterogeneity in LPP classification, modest AUC values, weight loss-selected cohort, and cross-sectional design) highlight the need for future research that not only tracks progression from preserved function to severe sarcopenia, frailty, and morbidity and mortality over time, but also clarifies underlying causal pathways and examines whether targeted interventions to alter uric acid, CRP, HDL, or albumin can improve physical function in older adults. Until such data become available, these inexpensive and readily available indices may provide clinically relevant contextual information for assessing sarcopenia-related impairment but do not provide diagnostic capacity.

Conclusions

In conclusion, inflammation-derived indices (UHR, MHR, CAR, CHR, SII, and to a lesser extent THR) were elevated in older adults with low physical performance, indicating a clear association between systemic inflammation and functional decline in aging. Among these parameters, UHR demonstrated the highest specificity and SII and CAR showed relatively stronger sensitivity for identifying individuals with LPP, supporting their relevance as reflections of metabolic-inflammatory dysregulation rather than diagnostic indicators on their own. These inexpensive, widely accessible biomarkers may therefore provide complementary clinical insight when interpreted alongside direct functional assessments, particularly in geriatric outpatient settings where rapid screening is critical.

Although our findings support the concept that chronic low-grade inflammation contributes to performance decline with age, the modest discriminative performance of these indices indicates that they cannot replace physical performance tests in the assessment of sarcopenia. Future large-scale longitudinal studies are required to determine whether changes in these biomarkers parallel alterations in physical performance over time, to clarify their underlying mechanisms, and to evaluate whether modification of upstream components such as uric acid, HDL metabolism, CRP, or albumin can help maintain muscle health and functional independence in older adults. Until such evidence is available, inflammation-based ratios should be regarded as contextual and supportive clinical tools rather than as standalone diagnostic measures for sarcopenia-related physical impairment.

Abbreviations

ADL

Katz Activities of Daily Living index

AF

Atrial fibrillation

AUC

Area Under the curve

BMI

Body mass index

CAD

Coronary artery disease

CAR

C-reactive protein to albumin ratio

CFS

Clinical Frailty Scale

CGA

Comprehensive Geriatric Assessment

CHR

CRP to HDL ratio

CKD

Chronic kidney disease

CRP

C-reactive protein

DM

Diabetes mellitus

EWGSOP2

the revised European Working Group on Sarcopenia in Older People 2 consensus

FDR

False discovery rate

GDS

Geriatric Depression Scale

HGST

Handgrip strength test

HT

Hypertension

IADL

Lawton Instrumental Activities of Daily Living scale

LPP

Low Physical Performance

MASLD

Metabolic dysfunction–associated steatotic liver disease

MHR

Monocyte to HDL ratio

MNA-SF

Mini Nutritional Assessment-Short Form

NPP

Normal physical performance

ROC

Receiver Operating Characteristic Curve

SARC-F A

Simple Questionnaire To Rapidly Diagnose Sarcopenia

s-MMSE

standardized Mini-Mental State Examination

SII

Systemic immune-inflammation index

STST

the 5 times-sit-to-stand-test

THR

Triglyceride to HDL ratio

TUG

Time up and go test

UHR

Uric acid to HDL-cholesterol ratio

UWL

Unintentional weight loss

Authors’ contributions

Prof. Dr. Mustafa Cankurtaran and Dr. Elif Gecegelen supervised and monitored all stages of the study from its inception, including study design and manuscript development. Dr. Cansu Atbaş, Didem Karaduman, and Mete Üçdal contributed to patient screening, assessment of eligibility and exclusion criteria, and patient follow-up. Statistical analyses were performed by Dr. Elif Gecegelen and Didem Karaduman, and the use and verification of the R statistical software were jointly reviewed by Mete Üçdal. Dr. Elif Gecegelen, Didem Karaduman, and Mete Üçdal contributed to drafting the manuscript. All authors critically reviewed the manuscript and approved the final version. All authors read and approved the final manuscript.

Funding

This study did not receive any specific funding from public, commercial, or not-for-profit funding agencies.

Data availability

The datasets generated and analyzed during the current study are not publicly available due to institutional data protection policies, but are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Ethical approval for this study was obtained from the Hacettepe University Faculty of Medicine Health Sciences Research Ethics Committee (Date: January 9, 2024; Approval No: 2024/01–01, SBA 23/411). All procedures were conducted in accordance with the principles of the Declaration of Helsinki and relevant national regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.