Differential Systemic Immune-Inflammation Index Levels in People with and without HIV Infection

Abstract

Background:

HIV infection is linked to persistent inflammation despite effective antiretroviral therapy (ART). The Systemic Immune-Inflammation Index (SII) is a marker of inflammation in various conditions.

Methods:

We compared SII values between PWH and PWoH. Clinical blood laboratory data were used to calculate the SII for each participant using the formula [(Platelet count × Neutrophil count) / Lymphocyte count]. Differences in SII values between the groups were analyzed using the Wilcoxon test, and the impact potential confounders was assessed with multivariable regression models.

Results:

The study included 343 PWH and 199 PWoH. Age and race did not significantly differ, but sex distribution did (83.1% male in PWH vs. 55.8% in PWoH, p<0.0001). Among PWH, median [IQR] nadir and current CD4 counts were 199 cells/μL [50, 350] and 650 [461, 858], respectively. Nearly all PWH were on ART, with 97.2% achieving viral suppression. PWH had lower SII values than PWoH (327 [224, 444] vs. 484 [335, 657], p=1.35e-14). PWH also had lower neutrophils and platelets (ps<0.001) and higher lymphocyte counts (p=0.001). These differences remained significant after adjusting for age, sex, and other potential confounders.

Discussion:

Contrary to expectations, PWH had lower SII levels, likely due to altered hematologic parameters influenced by HIV and ART. These findings suggest that SII interpretation in PWH requires consideration of unique hematologic profiles and underscore the need for further research to understand the mechanisms and clinical implications of SII in HIV management.

Introduction

HIV infection is commonly associated with increased systemic inflammation and immune activation that persists despite viral suppression on antiretroviral therapy (ART). This study investigated the Systemic Immune-Inflammation Index (SIl), a widely studied index that combines three blood parameters—platelets, neutrophils, and lymphocytes—to measure systemic inflammation based on readily available clinical laboratory information. The SII has been validated in a variety of conditions, such as rheumatoid arthritis¹, but a gap in the literature exists because there are no published, systematic comparisons of the SII in virally suppressed people with HIV (PWH) to those of people without HIV (PWoH). To our knowledge, only two prior studies examined the SII in the context of HIV. The first study found the SII to be significantly higher in ART-naïve PWH than PWH on ART or PWoH² but suffered from small sample sizes. The second study found hypertension to be associated with significantly higher SII in PWH³, but it was a retrospective study based on a medical record review rather than standardized evaluations and did not include a control group of PWoH. Thus, we sought to characterize the SII and its clinical correlates in a large sample of PWH and compare them to PWoH.

Methods

We performed a cross-sectional comparative analysis of SlI values between people with HIV (PWH) and people without HIV (PWoH). Inclusion criteria were age greater than 21 years and serological evidence of IV infection (for PWH) or a negative test (for PWoH). Exclusion criteria included inability to participate in the study procedures and ongoing major medical or psychiatric disorders that would compromise the study assessments.

HIV infection was diagnosed by enzyme-linked immunosorbent assay with Western blot confirmation. Routine complete blood counts and CD4+ T cells (flow cytometry) were performed at a Clinical Laboratory Improvement Amendments (CLIA)–certified laboratory. HIV RNA was quantified by commercial RT-PCR (Amplicor version 1.5, Roche Diagnostics, Indianapolis, IN, USA; lower limit of quantification 50 copies per mL). Nadir CD4+ T-cell count was the lowest self-reported count or study measurement value.

From the complete blood count, data on platelet count, neutrophil count, and lymphocyte count were used to compute the Sll for each participant using the formula [(Platelet count × Neutrophil count)/Lymphocyte count]. We evaluated several other ratios that have been studied in the literature⁴ – the systemic inflammation response index (SIRI = [neutrophil count × monocyte count] / lymphocyte count; neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR), and monocyte to lymphocyte ratio (MLR). High-sensitivity C-reactive protein (CRP) in mg/L was measured in blood serum by latex immunoturbidimetry in a CLIA-certified laboratory; values were log10-transformed to improve distributions for statistical analysis.

Covariables, potential confounding conditions. As summarized in Table 1, we identified potential demographic, anthropometric, and clinical confounding conditions. Tobacco use history was collected using the timeline follow-back interview⁵, which generated an estimated cumulative lifetime number of cigarettes, cumulative lifetime days of tobacco smoking, and days since last tobacco use. Comorbidities were summarized using the Charlson Comorbidity Index (CCI), a validated index of comorbidity burden⁶. Metabolic syndrome (MetS) was evaluated using standard methods.⁷ Substance use and psychiatric variables (eg, methamphetamine use disorders) were assessed using the Composite International Diagnostic Interview (CIDI).⁸ We calculated the Framingham CVD Risk Score⁹, which incorporates risk factors (age, sex, total cholesterol level, high-density lipoprotein (HDL), cholesterol level, systolic blood pressure treatment for hypertension, smoking status, diabetes status) to predict the likelihood of cardiovascular events

Table 1.

Key predictors, covariables and outcomes.

Predictors	Covariables/Potential Confounders	Outcomes
HIV serostatus	Demographics, substance use characteristics	SII
	BMI, tobacco smoking, substance use and psychiatric characteristics	SIRI
	MetS, CCI

SII: Systemic Immune-Inflammation Index; SIRI: Systemic Inflammation Response Index

Statistical methods

Because SII values had a skewed distribution, the statistical significance of the differences in values between the two groups was determined using the Wilcoxon test. Demographics, medical history, and HIV disease characteristics were summarized using means and standard deviations, medians and interquartile ranges, or counts and percentages as appropriate. Table 1 summarizes the key predictors, outcomes, covariables, and potential confounders, including demographics, HIV disease characteristics, and comorbidities (CCI, MetS) that, when significant, were evaluated in multivariable regression. Log₁₀ transformation was used for non-normally distributed values (CRP). The relationship between SII and log₁₀ CRP levels was assessed using Pearson correlation. Statistical analyses were performed using JMP Pro 17.0.0 (SAS Institute Inc., Cary, NC, USA, 2021).

Results

Participants. We evaluated 542 participants, comprising 343 PWH and 199 PWoH. Table 2 presents the demographic and clinical characteristics of the participants. Age and race did not differ significantly by HIV serostatus. The sex distribution differed significantly between groups, with 83.1% of PWH being male compared to 55.8% of PWoH (p < 0.0001). Nearly all PWH (98.5%) were on ART, and 97.2% had viral suppression; median [interquartile range, IQR] nadir and current CD4 counts were 199 cells/μL [50,350] and 650 [461,858]. PWH did not differ from PWoH in body mass index (BMI) and Framingham CVD Risk Score. Cumulative tobacco quantity in the (estimated number of cigarettes smoked) was higher in PWH than PWoH (71435 [9537, 18307] versus 2526 [116, 48225], p=0.03). PWH had higher rates of lifetime methamphetamine use disorder (34.7%) than PWoH (5.69%), p=3.37e-11. The average Charlson Comorbidity Index (CCI) was higher in PWH than PWoH (7.06±4.82) versus 2.83±2.71, p=6.53e-4). BDI-II scores were not different in PWH and PWoH.

Table 2.

Demographic and clinical characteristics of PWH and PWoH.

	PWH (n=343)	PWoH (n=199)	P
Demographics
Age – mean (SD)	56.4 (10.4)	57.2 (10.7)	0.387
Sex, female – N (%)	58 (17.1%)	88 (44.9%)	4.86e-12
Race/Ethnicity			0.981
Black	34 (10.8%)	17 (9.71%)	-
Hispanic	136 (43.2%)	80 (45.7%)	-
Non-Hispanic White	140 (44.4%)	78 (44.6%)	-
Other	5 (1.59%)	0 (0%)
HIV Disease and Treatment
Estimated Duration of HIV – mean (SD)	19.8 (10.3)	-	-
Nadir CD4* count - median (IQR)	199 (51, 350)	-	-
Current CD4* - median (IQR)	652 (462, 860)	-	-
On ART – N (%)	324 (98.5%)	-	-
Virally suppressed – N (%)	282 (92.8%)	-	-
Other characteristics
SII	378 (242)	528 (261)	4.15e-11
SIRI	0.845 (0.649)	1.045 (0.667)	0.0006
Log10 CRP	0.215 (0.415)	0.177 (0.486)	0.331
BMI	26.6 (7.54)	27.2 (8.64)	0.468
MetS – N (%)	93 (27.1)	46 (23.1%)	0.302
Framingham CVD Risk Score	15.7 (12.4)	16.9 (11.6)	0.560
Charlson Comorbidity Index (CCI) – mean (SD)	7.06 (4.82)	2.83 (2.71)	6.53e-4
Cumulative tobacco smoking, days – median (IQR)	5114 (1461, 11342)	844 (38, 7129)	0.666
Cumulative tobacco quantity, #cigarettes - median (IQR)	71435 (9537, 18307)	2526 (116, 48225)	0.0367
Days since last tobacco use - median (IQR)	3366 (137, 9080)	1592 (46, 6883)	0.303
Lifetime methamphetamine use disorder – N (%)	86 (34.7%)	7 (5.69%)	3.37e-11
BDI-II	8.67±8.48	6.75±8.48	0.240

^*CD4+ T lymphocytes/μl (median (IQR) CVD=cardiovascular disease

SII and related parameters. PWH had lower SIl values than PWoH (327 [224,444] versus 484 [335,657], Wilcoxon p=1.35e-14; Figure 1). Among PWH with an undetectable viral load, SII values compared to PWoH were 376±239 versus 528±261 (p=2.79e-11). PWH also had lower neutrophil counts than PWoH (2.85 [2.19,3.65] versus 3.37 [2.59,4.12], p=2.46e-6). Platelet counts were lower in PWH than in PWoH (225 [185,268] versus 246 [215, 280], p=3.05e-5), and PWH had higher lymphocyte counts than PWoH (1.90 [1.52,2.42] versus 1.74 [1.45,2.06], p=0.0013). Differences in other commonly used blood cell count-based biomarkers of systemic inflammation by HIV serostatus were also assessed. PWH and PWoH did not differ with respect to MLR (0.26±0.10 versus 0.28±0.11, p=0.0536) but had lower levels for NLR (1.65±0.89 versus 2.09±0.88; p=3.98e-8), and PLR (125±52.6 versus 151±60.8, p=2.1e-8), mirroring the SII. The SIRI showed a similar pattern to the SII (PWH 0.845±0.649 versus PWoH 1.045±0.667, p=0.0006). Effect sizes (Eta squared [η²]) and p-values in a multivariable model (whole model p=0.0007) were as follows: HIV serostatus (η²=0.026, p=0.0003), sex (η²=0.011, p=0.021), and their interaction (η²=0.001, p=0.495).

Figure 1.

SII levels were significantly lower in PWH than in PWoH

SII, demographics, HIV disease characteristics, and other covariables.

In the sample as a whole, females had higher SII values than males (404 [287, 599] versus 348 [244, 502], p=0.0225). However, when evaluated separately, SII values did not differ by sex in PWH alone (females 389±205 versus males 368±237, p=0.551) or PWoH (females 530±267 versus males 532±262, p=0.928). In a multivariable model predicting SII from HIV serostatus, sex, and their interaction, only serostatus was significant (p=1.54e-8). SII did not differ by ethnicity (Black [N=51], Hispanic [N=216], White [N=218], Other [N=5], p=0.527). SII was not related to age (r=0.0491, p=0.278). Among PWH, higher SII was correlated with lower CD4 lymphocyte counts (r=−0.205, p=0.0002) and CD8 counts (r=−0.266, p<.0001). A multivariable analysis including HIV serostatus and CD4 as predictors of SII could not be done since CD4 counts were not measured in PWoH. Nadir CD4 counts were not related to SII (r=−0.0539, p=0.320). PWH with viral suppression (n=297, 97.7%) did not differ in SII from those without viral suppression (327 [223, 449] versus 305 [77.1, 516], p=0.569). Among PWH,

SII was unrelated to the ARV regimen type (Table 3, p=0.526). The total months of exposure to antiretroviral medications was (median [interquartile range]), 182 [82.8, 250]. The correlation of total months with SII was small and nonsignificant: r=−0.0566, p=0. 330.The CCI was unrelated to SII (For PWH, r=0.0491, r=0.705; for PWoH, r=−0.1013 p=0.689). SII was not related to BMI (r=−0.0101, p=0.845). For PWH, the SII did not differ for those with and without MetS (365±230 versus 383±247, p=0.533). The SII was unrelated to cumulative tobacco smoking in days (Spearman rho=0.006, p=0.944), cumulative tobacco quantity in #cigarettes (rho =−0.02, p=0.786), and days since last tobacco use (rho =−0.06, p=0.498). SII levels did not differ for participants with a lifetime history of methamphetamine use disorder (371 [271, 543] versus those without 304 [204, 488], p=0.203). The SII did not correlate with BDI-II scores (r=0.00305, p=0.965)

Table 3.

SII median (IQR) by ARV regimen type. 3-class (eg, protease inhibitor [PI] + nucleoside reverse transcriptase inhibitor [NRTI] + non-nucleoside reverse transcriptase inhibitor [NNRTI]); 4+ class (eg, integrase inhibitor [II] + PI + NRTI + NNRTI), other (eg, fusion inhibitor).

Regimen type	N (%)	Median (IQR)
3-class	39 (13.6%)	282 (191, 444)
4+ class	2 (0.7%)	279 (220, 338)
NNRTI/II	2 (0.7%)	557 (379, 736)
NNRTI/NRTI	35 (12.2%)	355 (253, 510)
NRTI	3 (1.0%)	283 (160, 543)
NRTI/II	166 (58.0%)	313 (223, 468)
PI/II	3 (1.0%)	378 (193, 410)
PI/NRTI	31 (10.8%)	286 (219, 383)
Other	5 (1.7%)	405 (260, 484)

Correlation of the SII with C-reactive protein, another frequently studied inflammatory biomarker.

Log₁₀ CRP levels were numerically but not statistically significantly higher in PWH than in PWoH (0.215±0.415 versus 0.177±0.486, p=0.331). Higher SII significantly correlated with higher log₁₀ CRP levels for PWH (r=0.188, p=4.74e-4) and PWoH (r=0.266, p=1.49e-4). In a multivariable analysis including HIV serostatus, CRP, and their interaction as predictors of SII, both serostatus (p=3.45e-12) and CRP were significant (p=1.78e-7), but not the interaction (p=0.487).

Discussion

We unexpectedly found that PWH had lower SII as well as other cellular ratio levels than PWoH despite the well-documented chronic inflammation associated with HIV infection, as indicated by higher CRP levels in virally suppressed PWH. Given the SII calculation formula, this unexpected result can be explained by the lower neutrophil and platelet counts and higher lymphocyte counts observed. All values for the commonly used cellular ratios appear to be within the normal range. These hematological parameters are influenced by HIV’s effect on bone marrow, leading to altered production and survival of these cells. HIV can also cause thrombocytopenia through mechanisms such as immune-mediated platelet destruction and impaired platelet production in the bone marrow^10,11. ART can partially reverse cytopenia, but cell counts may remain lower compared to HIV-negative individuals. The reduction in specific hematological components contributed to the lower SII we observed in PWH. Chronic HIV infection or ART may have differential impacts on cell subpopulations, as seen with our sample’s lower neutrophil and platelet counts but higher lymphocyte numbers in PWH.

Lower CD4 counts are associated with abnormal bone marrow parameters¹², consistent with our finding that higher SII correlated with lower CD4 and CD8 lymphocyte counts in PWH. Individuals with long-standing HIV infection might develop adaptive mechanisms to regulate inflammation over time, impacting SII levels. ART might play a general role in modulating these effects by reducing inflammation and partially restoring immune function. However, we found no SII differences attributable to specific regimen types.

The SII was not significantly related to the CCI and other comorbidities or to a lifetime history of methamphetamine use disorder or depressed mood. Although published studies show that tobacco smoking is related to increased inflammation in the general population¹⁴, we did not find an association between the SII and tobacco use variables in our sample.

Our study has several limitations. The cross-sectional design precludes us from drawing conclusions about the causality of the observed associations. Additionally, the relatively small number of unsuppressed PWH limits the generalization of the findings to all PWH. While the study included 542 participants, the generalizability of the findings may be limited due to the demographics. The demographic characteristics, such as the higher proportion of males among PWH and the underrepresentation of females, may introduce bias and limit the applicability of the findings across different populations. Although the study controlled for several potential confounders, such as age, sex, BMI, and metabolic syndrome, other unmeasured factors may influence the results. For instance, lifestyle factors, co-infections, nutritional status, and genetic predispositions were not comprehensively accounted for. Relying on routine clinical laboratory measures and self-reported data (e.g., nadir CD4 counts) may introduce measurement bias. More advanced and comprehensive techniques for measuring immune function and inflammation, such as multi-parameter flow cytometry and cytokine profiling, could provide more detailed insights into PWH’s immune status and inflammatory processes.

CRP levels were slightly but non-significantly higher in PWH than in PWoH. This differs from the findings of a meta-analysis showing that high-sensitivity C-reactive protein (hsCRP) levels are significantly elevated in PWH on ART (including those who achieve viral suppression) compared to PWoH¹⁵.

Our findings indicate that the SII, a widely used marker of systemic inflammation, behaves differently in PWH than in PWoH. While the SII was lower in PWH, it was still correlated with key immune function indicators, such as CD4 and CD8 counts. These results underscore the importance of considering the unique hematologic effects of HIV and ART when interpreting SII in this population. Elucidating the mechanisms behind the observed differences in SII could inform the development of targeted interventions to manage systemic inflammation in PWH. Investigating the molecular mechanisms underlying bone marrow suppression and immune dysregulation in chronic HIV infection could uncover potential therapeutic targets to normalize hematopoiesis and reduce inflammation. Future longitudinal studies are needed to explore the causal relationships between HIV, ART, and SII and to determine the clinical utility of SII in managing HIV infection. Longitudinal studies should also explore the SII’s potential as a clinical tool for monitoring disease progression and the efficacy of therapeutic interventions in PWH. Longitudinal studies tracking changes in myeloid and lymphoid populations over time in PWH on ART could help elucidate the dynamic effects of chronic HIV infection and ART on these cell lineages.