Introduction

Testosterone is an important steroid hormone primarily secreted by the Leydig cells in the testes in males. It is a major component of male androgens. Testosterone is essential for developing male reproductive function, maintaining male secondary sexual characteristics, maintaining nitrogen balance, promoting protein synthesis and maintaining skeletal muscle growth1,2. However, studies have found that 20–50% of American males experience testosterone deficiency (TD), which significantly affects aspects such as sexual function, physical growth, and quality of life in males3,4,5. Therefore, in clinical practice, we need to focus on the occurrence of TD.

In recent years, several studies have identified a significant association between inflammatory status and testosterone levels6. A study conducted by Maggio et al. found that IL-6, TNF-α, and IL-1β have an impact on testosterone levels. Amar Osmancevic et al. found a correlation between IL-6, high-sensitivity C-reactive protein, and both total testosterone and bioavailable testosterone levels7. Another study found that participants with testosterone deficiency had higher levels of inflammatory biomarkers compared to those without testosterone deficiency8. Moreover, studies have also found that pro-inflammatory cytokines can regulate testosterone release by modulating the hypothalamic-pituitary-gonadal axis9. However, these findings have been based on the relationship between individual inflammatory cytokine responses and testosterone levels.

The pan-immune-inflammation-value (PIV) is a novel immune-inflammatory biomarker that assesses both local and systemic immune inflammation. It is calculated based on platelet, neutrophil, monocyte, and lymphocyte counts10,11. Several studies have already confirmed the association between PIV and malignant tumors such as breast cancer12, colorectal cancer13, and lung cancer14. Additionally, research has also demonstrated that PIV is a predictive factor for overall mortality in patients with anti-neutrophil cytoplasmic antibody-associated vasculitis15. However, the association between the immune-inflammatory biomarker PIV and TD has not been well assessed. We therefore undertook a cross-sectional, population-based study to examine the association between TD and PIV in those participating in the National Health and Nutrition Examination Survey (NHANES).

Materials and methods

Study participants

NHANES is an ongoing survey of the U.S. population that provides rich information on nutrition and health. It utilizes a multistage and probability sampling approach to collect data. Before data collection, the survey protocol of NHANES has obtained approval from the Institutional Review Board of the National Center for Health Statistics (NCHS), CDC, and has been approved by the NCHS Research Ethics Review Board. Informed consent was obtained from all participants, and the study followed the guiding principles of the Declaration of Helsinki.

For this survey, we utilized the 2011–2016 continuous cycles of the NHANES dataset in the United States. We excluded 15,151 female participants, 6506 participants under the age of 20, 836 participants with missing testosterone data, and 20 participants with missing PIV data from a total of 29,902 participants. We use imputation method for missing covariates. Ultimately, the study included a total of 7389 individuals (Fig. 1).

Fig. 1
figure 1

Flowchart of participant selection process (NHANES 2011–2016).

Full size image

Study variables

The exposure variable in this study is PIV, which is calculated as (neutrophil count × monocyte count × platelet count) divided by lymphocyte count. Serum testosterone levels were measured using precise isotope dilution liquid chromatography and tandem mass spectrometry. According to the clinical practice guidelines of the Endocrine Society and the guidelines of the American Urological Association, TD < 300 ng/dl is considered the threshold for low testosterone16,17. Covariates include age, race, education, BMI, smoking, alcohol consumption, diabetes, hypertension, triglycerides, and total cholesterol. BMI is categorized as < 25 kg/m2, 25–30 kg/m2, and > 30 kg/m2. Smokers are defined as individuals who have smoked at least 100 cigarettes in their lifetime, otherwise, they are considered non-smokers. Drinkers are defined as individuals who have consumed at least 12 types of alcoholic beverages in their lifetime, otherwise, they are classified as non-drinkers. Diabetes patients are defined as those who have been informed by a doctor about their diabetes diagnosis, are currently taking anti-diabetic medication, or are using insulin for blood sugar control. Hypertension is defined as having an average systolic blood pressure > 140 mmHg and/or diastolic blood pressure > 90 mmHg, as reported by a doctor, or being informed by a doctor about having hypertension. Triglycerides and total cholesterol are determined based on standard clinical biochemical values.

Statistical analysis

We have chosen ‘wtmec2yr’ as the weighted variable for the years 2011–2016. This is consistent with previous research methods18. By using weighted t-tests, we analyzed continuous variables, while weighted chi-square tests were employed for categorical variables. Continuous variables were presented as mean ± standard error, and categorical variables were expressed as proportions with 95% confidence intervals (CI). Initially, we assessed the differences between individuals with normal testosterone levels and those with low testosterone levels. To assess the association between PIV and testosterone deficiency, we employed weighted logistic regression analysis and constructed three models: in Model 1, no adjustment was made for any covariates; in Model 2, adjustment was made for age and race; and in Model 3, adjustment was made for age, race, education, BMI, smoking, alcohol consumption, diabetes, hypertension, triglycerides, and total cholesterol. Furthermore, after adjusting for all covariates, we employed Generalized Additive Models (GAM) and smooth curve fitting to observe nonlinear correlations. In cases of nonlinear correlations, segmented regression models were used to estimate threshold effects. The recursive experimental approach was utilized to identify the best-fitting model with the maximum likelihood value and determine inflection points. Lastly, we conducted subgroup analyses using a stratified multivariable logistic regression model to assess the correlation between PIV and testosterone deficiency. Statistical analysis was performed using R (version: 4.2.0). A p-value < 0.05 was considered statistically significant.

Results

Baseline characteristics

According to the inclusion criteria of this study, a total of 7389 participants were included. The average age of the participants was 47.1 ± 0.40 years, and the average PIV level was 313.67 ± 5.03. In our study, 2100 participants (28.42%) were diagnosed with TD.

The clinical characteristics of the participants are presented in Table 1. Statistical significance was observed in age, BMI, diabetes, hypertension, triglycerides, and PIV between TD and non-TD patients (p < 0.05). Compared to individuals with normal total testosterone levels, TD patients were older, had a BMI > 30/m2, higher triglyceride and PIV levels, and were more likely to have diabetes and hypertension.

Table 1 Basic characteristics of the participants, weighted.
Full size table

Correlation between PIV and TD

To further explore the relationship between PIV and testosterone deficiency, we analyzed their association using a weighted logistic regression model. In the unadjusted model (OR 0.81, 95% CI 1.46, 2.26) and the partially adjusted model (OR 1.68, 95% CI 1.34, 2.11), there was a positive correlation between PIV and the risk of testosterone deficiency. After full adjustment, this positive correlation remained statistically significant, with participants in the highest quartile of PIV having a 51% increased risk of testosterone deficiency compared to those in the lowest quartile of PIV (Table 2).

Table 2 The correlation between PIV and the prevalence of TD, weighted.
Full size table

Used a generalized additive model to investigate whether there is a linear relationship between PIV and testosterone deficiency

We utilized a generalized additive model and conducted an analysis through smooth curve fitting and threshold effect analysis, while adjusting for all covariates. We observed a non-linear relationship between PIV and testosterone deficiency. On the left side of the threshold, the rate of change in testosterone deficiency significantly alters as PIV values change, with statistical significance. Furthermore, we determined a turning point at 565.89 with a log-likelihood ratio of < 0.001. These results indicate a non-linear relationship between PIV and testosterone deficiency (Fig. 2; Table 3).

Fig. 2
figure 2

The association between PIV and the prevalence of TD. The red solid line represents the smoothed curve fit between the variables. The upper and lower blue bands represent the fitted 95% confidence intervals.

Full size image
Table 3 Threshold effect of PIV on testosterone deficiency.
Full size table

Subgroup analysis

Our subgroup analysis results showed a significant association between PIV and TD in each subgroup stratified by age, BMI, diabetes, and hypertension (p < 0.05). In the interaction analysis, we observed no significant differences in the relationship between PIV and TD based on age, diabetes, and hypertension, indicating that these factors do not significantly impact PIV correlation with testosterone deficiency (p for interaction > 0.05). However, obese may influence the correlation between PIV and TD (p for interaction < 0.05) (Fig. 3).

Fig. 3
figure 3

Subgroup analysis of the association between PIV and the prevalence of TD.

Full size image

Sensitivity analysis

We excluded discrete values with PIV > 3000 and conducted a sensitivity analysis between PIV and TD. The results indicated a positive correlation between PIV and TD risks, demonstrating statistical significance (p < 0.001). The sensitivity analysis results were consistent with the main analysis, suggesting that the conclusions of this study are not influenced by outliers (Table 4; Fig. 4).

Table 4 Conducting a sensitivity analysis on the correlation between PIV and testosterone deficiency.
Full size table

Discussion

In this cross-sectional study, after investigating demographic, examination, and individual data of participants in the NHANES database from 2011 to 2016, we found a significant association between increased PIV and an elevated risk of testosterone deficiency. Our research results indicate that PIV shows a certain degree of stability in patients with testosterone deficiency. In the future, PIV may become an effective indicator for assessing cognitive impairment.

To our knowledge, our study is the first to explore the association between PIV and TD. Previous research has already established a correlation between inflammation and testosterone. For instance, studies have shown that testosterone can suppress NF-KB signaling, thereby inhibiting P65 signaling (NF-kB and p65 are associated with the production of pro-inflammatory cytokines and the pathogenesis of many inflammatory diseases), leading to a reduction in pro-inflammatory markers such as TNF-α, IL-2, among others19,20,21. Mohamad’s summary of all evidence from human observations and animal studies suggests that TD can promote an increase in levels of inflammatory cytokines. In addition, anti-inflammatory activity may be a function of testosterone in the body22. Serum levels of TNF-α and MIP1α were significantly higher in men with below-normal testosterone levels23. Additionally, A study conducted by Zhang et al. investigated the association between the Dietary Inflammatory Index and male hormones, finding that men adhering to an inflammatory diet appeared to have a greater risk of TD24. Traish et al. also found that testosterone regulates the function of immune system cell components such as macrophages and neutrophils and reduces the release of pro-inflammatory cytokines (TNF-α, IL-6) associated with vascular endothelial damage, eliciting anti-inflammatory effects in tissues and organs25. In our study, we found a non-linear positive correlation between PIV and the increased risk of TD, supporting previous research that high PIV levels are associated with an elevated risk of TD. While the exact mechanism of the association between PIV and testosterone levels remains unclear, a possible conjecture is that there is interaction between the immune system and the reproductive system. During inflammation, pro-inflammatory cytokines typically stimulate the hypothalamic-pituitary-adrenal axis, wherein hypothalamic corticotropin-releasing hormone controls the release of adrenal hormones. Furthermore, it is worth noting that in our interaction analysis, when we restricted participants to obese males, the risk of TD slightly increased. Obese individuals may have a higher risk of developing TD compared to non-obese individuals, which may suggest that higher PIV increases the risk of testosterone decline in obese patients. This is consistent with previous epidemiological surveys indicating an inverse relationship between total testosterone in the blood and obesity26.

Testosterone is referred to as the primary male sex hormone and is believed to possess immune-regulatory properties and influence the occurrence and progression of chronic inflammation27,28. There are also several classic inflammatory markers widely used in clinical practice. Jian Zhou et al. conducted a cross-sectional study investigating the association between neutrophil/lymphocyte ratio and testosterone29. Additionally, a study involving 524 patients found a clinically significant clinical association between C-reactive protein levels, testosterone levels, and prostate cancer30. In these traditional inflammatory markers, a single index only reflects the activity of a specific pathway or cell type. In inflammatory responses in diseases, an elevation in neutrophils (pro-inflammatory) alongside a decrease in lymphocytes (immunosuppression) may coexist, indicating potential limitations of a single index. Furthermore, in chronic inflammation, platelets may modulate immune responses by releasing microparticles, necessitating a multi-index assessment of this complex interaction. Combining immune and inflammatory markers through PIV may have a synergistic amplifying effect, as it incorporates neutrophils, monocytes, platelets, and lymphocytes. Moreover, PIV has been demonstrated as a novel composite circulating immune biomarker that can better reflect inflammatory status31. Moreover, it has demonstrated superior predictive value in several studies. For instance, Ahmet Bilgehan Şahin et al. investigated the predictive ability of neutrophil-to-lymphocyte ratio, monocyte-to-lymphocyte ratio, and PIV for chemotherapy response and survival rates in breast cancer patients. Their ROC curve analysis confirmed that among these markers, PIV was more effective and accurate in predicting outcomes for breast cancer patients32. Similarly, Meikai Zhu et al. compared the predictive value of PIV, systemic immune inflammation index (SII), derived neutrophil-to-lymphocyte ratio, and other markers for prostate cancer and clinically significant prostate cancer. They found that PIV had better diagnostic value33. A study analyzed the prognostic value of inflammatory markers (Neutrophil-to-Lymphocyte Ratio, Platelet-to-Lymphocyte Ratio, SII, PIV) in predicting the outcome of lung cancer. The results demonstrated that PIV is a comprehensive and convenient prognostic indicator for overall survival in lung cancer patients34. In conclusion, PIV has demonstrated excellent predictive ability in various studies. Furthermore, PIV is a widely available method with non-invasive characteristics, ease of accessibility, and low cost. Therefore, it holds great potential for broad clinical applications.

Our study has several strengths. Firstly, we utilized a large sample of data that represents the U.S. population, and careful consideration was given to the sample design and weighting, enhancing the representativeness of the data. Secondly, we adjusted for confounding variables in weighted logistic regression analysis to explore the correlation between PIV and TD. Lastly, PIV is a composite blood marker that is easily accessible and cost-effective in clinical practice. However, our study also has limitations. It is a cross-sectional study that can only explore the correlation between the variables and cannot establish causality. Thus, further research is needed to determine the causal relationship. Although we attempted to adjust for confounding factors, we cannot completely eliminate the potential influence of other unmeasured confounders.

Conclusion

In conclusion, our study results indicate a non-linear positive correlation between PIV and the risk of TD. However, further prospective research is needed to investigate the temporal relationship between PIV and TD and to design experimental research (such as intervening in PIV to observe changes in testosterone) to explore causal relationships.