Association between lymphocyte-to-monocyte ratio and prostate cancer in men: A population-based study

Abstract

Using the novel inflammatory biomarker lymphocyte-to-monocyte ratio (LMR), this work aimed to look into any potential connections between LMR and prostate cancer (PCa). A cross-sectional research investigation was conducted on 7706 male participants involved in the National Health and Nutrition Examination Survey from 2001 to 2010. Multivariate logistic regression modeling investigated the relationship between LMR levels and PCa risk. Furthermore, threshold analysis, subgroup analysis, interaction testing, and smoothed curve fitting were carried out. A significant negative correlation was seen between LMR and PCa risk (OR = 0.79, 95% CI: 0.65–0.97, P = .0002), even after controlling for potential confounding factors. A significant nonlinear negative correlation with a threshold effect and a breakpoint of 4.86 was found by smooth curve fitting between LMR and PCa. Subgroup analysis revealed a significant interaction (P for interaction = 0.0448) between the negative correlation between PCa and LMR about hypertension. Moreover, additional stratified smoothed curve fitting demonstrated a statistically significant inverse relationship between PCa risk and LMR. According to our findings, there is a substantial inverse relationship between PCa risk and LMR level. The inflammatory response-related index is quick, easy to use, and offers some clinical references. However, more extensive prospective investigations are required to confirm the involvement of LMR levels in PCa.

1. Introduction

Among industrialized nations, prostate cancer is a frequent substance tissue tumor and ranks as the second most prevalent malignancy globally, behind lung cancer.^[1] An American Cancer Society research findings indicate that both the incidence and mortality rate from prostate cancer are rising annually.^[2,3] In the initial stages of its growth, prostate cancer does not exhibit any unique clinical signs and is difficult to identify.^[4] Prostate-specific antigen (PSA) levels can be measured to aid in the early identification of prostate cancer.^[5] Yet, PSA values have limited specificity and are influenced by several other variables.^[6] Furthermore, a significant level of overdiagnosis and overuse of test results for PSA for prostate cancer occurs.^[7]

Several malignancies and inflammation-related indicators are somewhat correlated, which suggests that inflammation-related markers might be a novel early identification of tumor indicators.^[8,9] Tumor cell incidence, growth, and migration are intimately linked to the inflammatory state.^[10] Several immune-inflammatory cells, including lymphocytes, monocytes, neutrophils, and platelets, will exhibit distinct alterations within inflammatory settings, and their respective quantities may indicate when inflammation begins.^[11,12] These immunoinflammatory markers might be helpful in the criteria for tumor diagnosis and prognosis.^[13,14] The possible mechanisms of action of immune-inflammatory cells on carcinogenesis and pathophysiology are currently being investigated in several laboratory experiments and in vivo research.^[15,16] More excellent neutrophil-lymphocyte ratio ratios were linked to more aggressive tumors and higher Gleason scores (GSs) when researching prostate cancer.^[17] These signs are quick and easy to use in hospitals and clinics as useful diagnostic or prognostic indications. However, more research is required as the association between lymphocyte-monocyte ratio levels and prostate cancer (PCa) risk remains unclear.

Evaluating the clinical relevance of lymphocyte-to-monocyte ratio (LMR), one of the immunoinflammatory markers of significant attraction, concerning PCa risk is crucial. Accordingly, to ascertain a meaningful correlation between LMR and PCa, this study examined 5 cycles of individual data from the National Health and Nutrition Examination Survey (NHANES) database.

2. Methods

2.1. Study design and population

Using a stratified multistage probability sampling approach, NHANES is continuously cross-sectional research by the National Center for Health Statistics. The National Center for Health Statistics Research Ethics Review Committee approved the NHANES study procedures. Each survey participant also signed a documented informed authorization form. For the pertinent analysis in our current study, we examined information collected during 5 NHANES cycles over 10 years (2001–2010). Within these instances, 52,192 individuals were considered in our study; however, we excluded female participants, individuals under 40 years old, and participants whose PCa and LMR data were missing. Consequently, 7706 participants were included in our final analysis, as illustrated in Figure 1.

Figure 1.

Flow chart.

2.2. Variables

In the present investigation, PCa was the dependent factor, while the amount of LMR was the independent variable. “Have you ever been told by a doctor or health professional that you had prostate cancer?” was posed to participants by the trained person being interviewed, and if participants replied favorably, their medical condition was classified as prostate cancer. With lymphocytes and monocytes expressed as ×10³ cells/μL, the LMR was computed by counting both types of blood using an automatic analysis of the blood system.^[18] Additional factors in our study, such as age, race, marital status, education level, smoked at least 100 cigarettes in life, alcohol use status, hypertension, diabetes, poverty to income ratio, serum albumin, serum cholesterol, serum triglycerides, body mass index, might have an impact on the association between LMR and PCa. Competent interviewers employed a computer-assisted personal interview method to get pertinent data from the subjects, and the required tests were conducted using the proper standardized scientific analyses. Visit www.cdc.gov/nchs/nhanes/ for more detailed information on these variables and additional guidance on measurements and calculations.

2.3. Statistical analysis

To lessen the unpredictable nature of the dataset, relevant statistical analyses were carried out by the Centers for Prevention and Control of Diseases standards. Appropriate sample weights were also used. The existence or absence of PCa was used to characterize baseline tables for the study population scientifically, and weighted linear regression models and averages with or without a standard deviation were used to describe continuous variables.^[19] Multiple linear regression analysis determined the beta values and 95% confidence intervals between LMR and PCa. The multiple linear regression analysis was conducted using 3 models: model 1, which was unadjusted for variables; model 2, which was adjusted for age and race; and model 3, which was corrected for all covariates. Subgroup analyses of the link between LMR and PCa were carried out by adding interaction effects to test for heterogeneity of correlations throughout subgroups and by correcting for variables that classified factors as prespecified possible effect modifiers. A threshold effects analysis model was used to study connections and inflection points between LMR and PCa, along with smoothed curve fitting to analyze nonlinear interactions. R version 4.2.2 (http://www.R-project.org) and Empower software (www.empowerstats.com) were used for all analyses. P < .05 was used as the statistically significant threshold.

3. Results

3.1. Baseline characteristics of participants

There were 7706 participants, 48.38% of whom were under 60, 41.13% between 60 and 80, and 10.49% of whom were 80 or older. The level of Mean LMR ± SD was 3.746 ± 1.718. Baseline characteristics of participants according to whether they had PCa as a column-stratified variable are shown in Table 1. Whether or not one had PCa was statistically associated with age, race, marital status, hypertension, serum albumin, serum cholesterol, serum triglycerides, and LMR (P < .05). Compared to non-PCa patients, PCa patients tended to be older, non-Hispanic White, married, with hypertension, lower serum albumin levels, lower serum cholesterol levels, lower serum triglycerides levels, and lower LMR levels.

Table 1.

Characteristics of the study population based on prostate cancer.

	Prostate cancer (N = 350)	Non-prostate cancer (N = 7356)	P-value
Age (years)			<.001
<60	4.00%	50.49%
60–80	64.00%	40.05%
≥80	32.00%	9.46%
Race (%)			<.001
Mexican American	6.00%	17.63%
Other Hispanic	2.57%	6.00%
Non-Hispanic White	64.00%	55.03%
Non-Hispanic Black	25.14%	18.09%
Other races	2.29%	3.25%
Marital status (%)			<.001
Married	72.29%	68.20%
Widowed	13.14%	6.61%
Divorced	8.00%	10.82%
Separated	2.00%	2.78%
Never married	2.29%	6.94%
Living with partner	2.29%	4.65%
Education level (%)			.704
Less than high school	15.52%	16.90%
High school	37.07%	37.70%
More than high school	47.41%	45.41%
Smoked at least 100 cigarettes in life (%)			.550
Yes	61.03%	62.62%
No	38.97%	37.38%
Alcohol use status (%)			.169
Yes	56.25%	63.89%
No	43.75%	36.11%
Hypertension (%)			<.001
Yes	58.00%	43.06%
No	42.00%	56.94%
Diabetes (%)			.980
Yes	16.37%	16.32%
No	83.63%	83.68%
Poverty to income ratio (years)	2.84 ± 1.47	2.80 ± 1.62	.612
SAL (g/dL)	4.17 ± 0.31	4.25 ± 0.31	<.001
SCH (mg/dL)	189.74 ± 42.19	198.38 ± 42.35	<.001
STR (mg/dL)	139.18 ± 75.13	170.67 ± 136.16	<.001
BMI (kg/cm2)	28.25 ± 5.05	28.72 ± 5.59	.127
LMR (1000 cells/uL)	3.24 ± 1.74	3.77 ± 1.71	<.001

Mean ± SD for continuous variables: the P value was calculated by the weighted linear regression model.

(%) for categorical variables, the P value was calculated by the weighted chi-square test.

BMI = body mass index, LMR = lymphocyte-to-monocyte ratio, SAL = serum albumin, SCH = serum cholesterol, STR = serum triglycerides.

3.2. Relationship between LMR and PCa risk

The creation of univariate and multivariate logistic regression models is displayed in Table 2. With all P-values <.05, our findings demonstrated that LMR was adversely correlated with prostate cancer in models 1, 2, and 3. For models 1, 2, and 3, the odds ratios (95% confidence intervals) were, in order, 0.77 (0.71, 0.84), 0.92 (0.85, 0.99), and 0.79 (0.65, 0.97). The LMR was split into quartiles, as indicated in Table 2, to conduct additional sensitivity and trend analyses. When comparing groups Q2, Q3, and Q4 to group Q1, the PCa risk-adjusted odds ratios (95% confidence intervals) in model 1 varied between 0.58 (0.44, 0.76), 0.47 (0.35, 0.62) and 0.30 (0.22, 0.42) (P for trend < .0001). The risk-adjusted odds ratios (95% confidence intervals) for PCa in model 3 were 0.91 (0.50, 1.66), 0.54 (0.27, 1.09), and 0.27 (0.11, 0.68) for the Q2, Q3, and Q4 groups, correspondingly, in comparison to the Q1 group (P for trend = .0018). These patterns also exist in model 2 (P for trend = .0047).

Table 2.

Associations between LMR and prostate cancer.

	OR (95%CI), P-value
	Model 1	Model 2	Model 3
LMR	0.77 (0.71, 0.84) < 0.0001	0.92 (0.85, 0.99) 0.0343	0.79 (0.65, 0.97) 0.0265
LMR (Quartile)
Q1	1.00	1.00	1.00
Q2	0.58 (0.44, 0.76) <0.0001	0.81 (0.61, 1.07) 0.1417	0.91 (0.50,1.66) 0.7532
Q3	0.47 (0.35, 0.62) <0.0001	0.82 (0.60, 1.12) 0.2201	0.54 (0.27, 1.09) 0.0865
Q4	0.30 (0.22, 0.42) <0.0001	0.58 (0.40, 0.83) 0.0029	0.27 (0.11, 0.68) 0.0054
P for trend	<.0001	.0047	.0018

Model 1: No adjustments;

Model 2: Minimally adjusted for age, race.

Model 3: Fully adjusted for age, race, marital status, education level, smoked at least 100 cigarettes in life, alcohol use status, hypertension, diabetes, poverty to income ratio, SAL, SCH, STR, BMI.

BMI = body mass index, LMR = lymphocyte-to-monocyte ratio, SAL = serum albumin, SCH = serum cholesterol, STR = serum triglycerides.

Moreover, we employed a smoothed curve-fitting model to describe the nonlinear connection between PCa and LMR. The findings indicated that LMR and PCa had a nonlinear negative connection (Fig. 2). Although there was a breakpoint of 4.86 and an apparent threshold effect, this was not statistically significant (log-likelihood ratio test = 0.103, Table 3).

Figure 2.

Smooth curve fitting of LMR to PCa. The solid red line represents the fitted curve between the variables. The blue solid line represents the 95% confidence interval of the fitted results. LMR = lymphocyte-to-monocyte ratio, PCa = prostate cancer.

Table 3.

Threshold effect.

Outcome	Prostate cancer risk
Model I
A straight-line effect	0.8 (0.68, 1.02)
Model II
Fold points (K)	4.86
<K-segment effect 1	0.71 (0.54, 0.93)
>K-segment effect 2	1.02 (0.82, 1.28)
Effect size difference of 2 vs 1	1.44 (0.97, 2.13)
Equation predicted values at break points	‐3.70 (‐4.24, ‐3.17)
Log likelihood ratio tests	0.103

Result variable: prostate cancer risk.

Exposure variables: LMR.

Results are expressed as OR (95% CI).

Adjusted for age, race, marital status, education level, smoked at least 100 cigarettes in life, alcohol use status, hypertension, diabetes, poverty to income ratio, SAL, SCH, STR, BMI.

BMI = body mass index, LMR = lymphocyte-to-monocyte ratio, SAL = serum albumin, SCH = serum cholesterol, STR = serum triglycerides.

The solid red line represents the fitted curve between the variables. The blue solid line represents the 95% confidence interval of the fitted results.

3.3. Subgroup analysis

As demonstrated in Figure 3, our study also conducted subgroup analysis and an interaction test according to age, smoked at least 100 cigarettes in life, alcohol use status, hypertension, and diabetes to confirm the connection between LMR and PCa in the entirety of the population and to ascertain the population a particular category.

Figure 3.

Subgroup analysis.

While there was not a significant relationship for individuals who had smoked at least 100 cigarettes in life, alcohol use status, or diabetes (P for interactions > .05), we did find a significant interaction for the hypertension subgroup (P for interaction = .0448). In hypertension individuals in the 60 to 80 years old range, the correlation between LMR and PCa remained negative. Our results indicate that there is a hypertension reliance in the link between LMR and PCa and that only patients with hypertension have this negative correlation (P for interaction < .05). No individuals without hypertension do.

Furthermore, further stratified smoothed curve fitting revealed a strong negative correlation between LMR and PCa risk, regardless of whether hypertension and diabetes subgroups of the population were present or not (as seen in Fig. 4 A and B).

Figure 4.

Further subgroup analysis.

4. Discussion

A negative correlation was discovered between LMR and PCa in this cross-sectional investigation based on an analysis of data from the NHANES database for 2001 to 2010. Furthermore, the hypertension subgroup analysis revealed a substantial interaction between the 2, indicating hypertension reliance. Finally, a strong negative association between LMR and PCa risk was discovered in this study, which is crucial for PCa early detection and prevention.

We understand that an increasing number of research conducted in the past several years has found a connection between PCa and inflammation.^[20–22] Prostate samples varied in morphology and neutrophil-lymphocyte ratio (NLR), according to a retrospective analysis; in individuals with prostate cancer, more excellent GS was linked to higher NLR.^[23] According to cohort research, NLR was substantially correlated with death specific to prostate cancer in all males.^[24] Another example comes from research that demonstrated the significance of NLR as a prognostic factor in prostate cancer and the correlation between greater NLR levels before treatment and higher PSA levels, GSs, and later clinical stages.^[25] Research by Luo that was cross-sectional revealed that, after controlling for all other variables, systemic immune-inflammation index (SII) was significantly positively linked with a 7% higher risk of prostate cancer (OR 1.07, 95% CI 0.99–1.15, P = .094).^[26] As a further illustration, a clinical investigation discovered that high SII was linked to decreased biochemical recurrence-free survival and specific unfavorable pathologic characteristics in patients with nonmetastatic prostate cancer and poor overall survival in patients with metastatic-castration resistant prostate cancer.^[27] SII is regarded as a reasonably priced biomarker for tracking PC condition survival, according to the findings of a meta-analysis.^[28] Another study found that SII performed exceptionally well and offered a superior clinical evaluation approach for PCa while having diagnostic significance in identifying isolated PCa.^[29] Our research revealed a strong inverse relationship between LMR levels and PCa risk in models 1, 2, and 3. A threshold effect with a breakpoint of 4.86 was also discovered between LMR levels and PCa. In summary, our results demonstrate that this is the first study to establish a robust negative association between LMR levels and PCa, even though there have been other publications on the relationship between inflammation-related markers and PCa risk.

No convincing explanation exists for the potential processes behind the association between inflammation and PCa.^[30–32] The immune system, tumor cells, and tumor stromal cells regulate their interactions during the formation and carcinogenesis of tumors.^[33,34] Formation of tumors can worsen the inflammatory response directly or indirectly, in addition to propagating by altering the microenvironment, and can also accelerate the growth and spread of tumor cells.^[35] Among the most prevalent immune cells in the tumor microenvironment, tumor-associated macrophages get most of their energy from monocytes in the surrounding blood circulation and only a tiny amount from locally expanding and proliferating macrophages.^[36,37] Therefore, one of the key elements contributing to the recruitment of monocytes into the tumor locally and in the peripheral circulation is the production of chemokines by tumor cells.^[38,39] Furthermore, the prostate tumor cells are aberrant epithelial cells against which the immune system of the body may mount a robust defense.^[40] An essential part of this process involves lymphocytes. Subsets of lymphocytes do matter, though, when it comes to antitumor cell function. This shows that the amount of lymphocytes is significantly connected with whether tumor cells can evade the body’s immune response, and this, in turn, affects how quickly patients’ cancers of the prostate develop. According to the research above, lymphocytes and monocytes may directly or indirectly impact the immunological microenvironment due to the danger of LMR on PCa. This may represent one of the influencing mechanisms.

Our research is the first to concentrate on the relationship between PCa risk and LMR. Our information came from a carefully planned and logically sampled NHANES database. First, the use of a sizable sample size, the accuracy of the data, and the justifiable removal of missing data values are the results of our research most vital points. Second, after adjusting for various covariates during data analysis, we incorporated the exposure variable LMR levels in various continuous and categorical variable formats in the multiple regression analysis model. Sensitivity analysis was also carried out to increase the accuracy of our findings. Smoothed curve fitting was ultimately used to investigate the nonlinear relationship between LMR and PCa risk. Additional threshold effect analysis, subgroup analyses, and trend tests were also carried out to determine how well our findings would apply to the general public and increase the reliability of our findings. Our study does, however, have several shortcomings. There could be some recollection bias because some of the conclusions of the research indicators were collected through questionnaires compared to objective measurements. Therefore, this cross-sectional study can only identify the association between LMR and PCa risk rather than concluding the causative link. Consequently, to further elucidate the causal link, we must do a significant number of prospective investigations and pertinent basic research in the following study. We considered many confounders that may affect the results throughout the analytic process, although there may continue to be certain factors that do. Therefore, further study is required in the follow-up. In summary, our investigation revealed a hitherto undiscovered negative correlation between LMR level and PCa risk; however, more research is still needed to determine the exact mechanism behind this link and its clinical implications for PCa diagnosis, treatment, and prognosis.

5. Conclusions

According to our research, there is a strong link between PCa risk and LMR levels. Although LMR is anticipated to be one of the many practical new measures for clinical PCa risk assessment, further research is necessary since causation cannot be shown in cross-sectional studies.

Acknowledgments

We are grateful to the employees at the National Center for Health Statistics of the Centers for Disease Control for Health Statistics for organizing, compiling, and developing the NHANES data and building the public database.

Author contributions

Conceptualization: Pingzhou Chen, Xiang Wu.

Data curation: Pingzhou Chen, Zhijie Huang.

Formal analysis: Pingzhou Chen.

Investigation: Pingzhou Chen.

Methodology: Pingzhou Chen.

Project administration: Xiang Wu.

Resources: Pingzhou Chen.

Software: Zhijie Huang.

Validation: Zhijie Huang, Xiang Wu.

Visualization: Xiang Wu.

Writing – original draft: Pingzhou Chen.

Writing – review & editing: Pingzhou Chen, Zhijie Huang, Xiang Wu.