Distribution and reference interval establishment of neutral‐to‐lymphocyte ratio (NLR), lymphocyte‐to‐monocyte ratio (LMR), and platelet‐to‐lymphocyte ratio (PLR) in Chinese healthy adults

Abstract Background Neutral‐to‐lymphocyte ratio (NLR), lymphocyte‐to‐monocyte ratio (LMR), and platelet‐to‐lymphocyte ratio (PLR) are associated with coronavirus disease 2019 (COVID‐19) and many diseases, but there are few data about the reference interval (RI) of NLR, LMR, and PLR. Methods The neutrophil count, lymphocyte count, monocyte count, and platelet count of 404,272 Chinese healthy adults (>18 years old) were measured by Sysmex XE‐2100 automatic hematology analyzer, and NLR, LMR, and PLR were calculated. According to CLSI C28‐A3, the nonparametric 95% percentile interval is defined as the reference interval. Results The results of Mann‐Whitney U test showed that NLR (p < .001) in male was significantly higher than that in female; LMR (p < .001) and PLR (p < .001) in male were significantly lower than that in female. Kruskal‐Wallis H test showed that there were significant differences in NLR, LMR, and PLR among different genders and age groups (p < .001). The linear graph showed that the reference upper limit of NLR and PLR increased with age and the reference upper limit of LMR decreases with age in male population. In female population, the reference upper limit of NLR in 50–59 group, LMR in >80 group, and PLR in 70–79 group appeared a trough; the reference upper limit of NLR in >80 group, LMR in 50–59 group, and PLR in 40–49 group appeared peak. Conclusion The establishment of RI for NLR, LMR, and PLR in Chinese healthy adults according to gender and age will promote the standardization of clinical application.

/ or gender. The detailed screening procedures of the study participants are shown in Figure 1.

| Laboratory methods
The venous blood collected from each subject was anticoagulated with EDTA-K2 and tested within 2 h. According to the principle of five blood cell classification, Sysmex XE-2100 automatic blood cell analyzer and related reagents (Sysmex) were used to determine the test items: neutrophil count, lymphocyte count, monocyte count, and platelet count in strict accordance with the instructions. NLR, LMR, and PLR are calculated as follows: NLR = neutral count/lymphocyte count, 25 LMR = lymphocyte count/monocyte count, 26 and PLR = platelet count/lymphocyte count. 27 Use special reagents and standard methods. Two levels of control materials (e-CHECK (XE)) F I G U R E 1 Establishing reference interval of neutrophil-tolymphocyte ratio, lymphocyte-to-monocyte ratio, and platelet-tolymphocyte ratio on the bias of Clinical and Laboratory Standards Institute CA28-A3 were used daily quality control. During the whole study period, internal quality control (IQC) was conducted by Westgard multi rule quality control method. The total coefficient of variation (CV) of the two control materials was less than 5%.

| Statistical analysis
Kolmogorov-Smirnov test was used to test the normality of the data; the normal distribution measurement data were expressed as X±s, and the comparison between the two groups was conducted by t test; the non-normal distribution measurement data was expressed as the median (p25-p75), and the comparison between the two groups was conducted by Mann-Whitney U test, and the comparison among the multiple groups was conducted by Kruskal-Wallis H test; reference interval was established from nonparametric 95% percentile according to CLSI C28-A3. 13 Statistical analysis was conducted by using IBM SPSS statistics version 20.0 software and MedCalc 19.1 software. p < .05 was statistically significant.

| RE SULTS
SPSS statistical results showed that among the 404,272 subjects included in the study, 205,592 were male, accounting for 50.85%; 198,680 were female, accounting for 49.15%; and the ratio of male and female was 1.03. Kolmogorov-Smirnov test showed that NLR, LMR, and PLR data showed skewed distribution (p < .05) ( Figure 2).
In order to explore the influence of gender on NLR, LMR, and PLR, we used Mann-Whitney U test to compare NLR, LMR, and PLR between different genders. The results showed that NLR in male (p < .001) was significantly higher than that in female; LMR in male (p < .001) and PLR in male (p < .001) were significantly lower than in women (Table 1).
In order to explore the influence of age on NLR, LMR, and PLR, we used Kruskal-Wallis H test to compare the NLR, LMR, and PLR of male and female on the basis of Mann-Whitney U test results. In the male group, there were significant differences in NLR (p < .001), LMR(p < .001), and PLR(p < .001) in different age groups. In the female group, there were also significant differences in NLR (p < .001), LMR (p < .001), and PLR (p < .001) in different age groups (Tables 2-4).
Further between-group comparisons by Mann-Whitney U test showed statistically significant differences between most age partitions. In the male group for NLR, the average NLR of 18- According to the nonparametric method recommended by CLSI CLSI C28-A3, we established NLR, LMR, and PLR reference intervals

TA B L E 4
Age-dependent and sexdependent reference values for PLR

F I G U R E 3
Values of upper reference limits for three indicators in different age partitions. A, The reference upper limit of NLR increased with age in male population. In female population, the reference upper limit of NLR in 50-59 group showed a trough and the reference upper limit of NLR in >80 group showed peak. B, The reference upper limit of LMR decreases with age in male population. In female population, the reference upper limit of LMR in >80 group showed a trough and the reference upper limit of LMR in 50-59 group showed peak. C, The reference upper limit of PLR increased with age in male population. In female population, the reference upper limit of PLR in 70-79 group showed a trough and the reference upper limit of PLR in 40-49 group showed peak