Background

Immune cells constitute the fundamental elements of the human immune system, with natural killer (NK) cell, monocyte, and dendritic cell (DC) subsets circulating in the peripheral blood to perform critical functions related to immune defense, surveillance, and immunoregulatory processes [1,2,3]. NK cells, a crucial component of innate immunity, exhibit potent cytotoxic activity against virally infected and malignantly transformed cells, through their unique ability to recognize and eliminate target cells without prior sensitization. Clinical investigations have established that abnormal functional and quantitative characteristics in NK cell populations are strongly associated with tumor progression and immunodeficiency syndromes [4, 5]. Monocytes are versatile immune effectors that coordinate innate and adaptive immune responses through their phagocytosis and cytokine production capabilities. Emerging evidence has demonstrated that quantitative and functional alterations in monocyte populations significantly correlate with the pathogenesis of various inflammatory disorders and cardiovascular pathologies [6,7,8]. DCs, recognized as the most potent antigen-presenting cells, serve as critical mediators of T lymphocyte activation and initiation of antigen-specific immune responses through their sophisticated antigen processing and presentation capabilities [9]. The dysregulation of DC subsets has been implicated in the disruption of immune homeostasis, which contributes to the development of immune tolerance and autoimmune disorders [10, 11].

Establishing reference ranges for the proportions and absolute counts of peripheral blood NK cell, monocyte, and DC subsets remains challenging because of the significant variations influenced by geographical, demographic, and biological factors, including regional differences and ethnic diversity. Previous attempts to establish reference intervals faced methodological challenges, as some of the studies did not fully comply with international standardization protocols or lacked comprehensive validation processes, potentially affecting the reliability and generalizability of their findings [12, 13]. Zhejiang Province, a densely populated region in southeastern China with a permanent population of > 60 million, has a unique demographic profile that merits the development of population-specific reference intervals. Therefore, there is a significant clinical need to establish validated reference ranges for NK cell, monocyte, and DC subsets that accurately reflect the particular immunological characteristics of the population in Zhejiang Province.

The Clinical and Laboratory Standards Institute (CLSI) EP28-A3c guidelines provide an internationally recognized framework for establishing medical reference intervals that offer standardized protocols for the definition, development, and verification of reference ranges [14]. In this study, we used the methodologies specified in CLSI EP28-A3c, combined with robust detection techniques and systematic classification approaches, to establish age-stratified reference ranges for peripheral blood NK cells, monocytes, and DC subsets in a healthy adult population from Zhejiang Province. Establishing these population-specific reference values is expected to significantly improve the diagnostic precision and therapeutic management of immune-related disorders, as well as help to advance the field of personalized medicine through the provision of region-specific immunological parameters.

Methods

Participants

This study was conductedat the Affiliated Dongyang Hospital of the Wenzhou Medical University, a tertiary general hospital in Central Zhejiang Province, China. Healthy adults who underwent routine health examinations at our institution between April 2023 and October 2024 were recruited. The inclusion criteria were: (1) apparent good health status; (2) completion of routine health screening and health questionnaire; (3) aged between 18 and 65 years; and (4) permanent residency in Zhejiang Province. The exclusion criteria were: (1) recent use of immunosuppressive agents (e.g., corticosteroids); (2) chronic infections or acute infections within the preceding 2 months; (3) confirmed to have current hepatitis B virus, hepatitis C virus, human immunodeficiency virus, or syphilis infection; (4) history of autoimmune disorders, allergic diseases, cardiovascular diseases, hematological disorders, or malignancies; (5) currently pregnant or lactating; (6) vaccination within the preceding 3 months; (7) history of blood donation, transfusion, or significant blood loss within the preceding 6 months; (8) substance use patterns including chronic alcohol abuse, heavy smoking (≥ 20 cigarettes per day), significant malnutrition, or long-term vegetarianism; (9) clinically significant abnormalities in complete blood count or hepatic/renal function tests.

This study was conducted in accordance with the ethical principles of the Declaration of Helsinki, and approved by the Institutional Review Board of the Affiliated Dongyang Hospital of Wenzhou Medical University. All of the volunteers provided written informed consent before being enrolled.

Sample processing and flow cytometry

Whole blood samples (~ 2 mL) were collected from each participant into ethylenediaminetetraacetic acid-anticoagulant tubes and processed within 24 h. Briefly, 100 µL of anticoagulated whole blood was dispensed into pre-labeled tubes using the reverse-pipetting method, gently mixed with pre-mixed antibody cocktails, and incubated in the dark at room temperature (20–25 °C) for 15 min. Subsequently, 1 mL of erythrocyte lysis solution was added to each tube, followed by thorough mixing and an additional 15 min incubation under light-protected conditions. After lysis, 2 mL of phosphate-buffered saline (PBS) was added to the suspension, which was then vortexed and centrifuged at 400 × g for 5 min. The supernatant was carefully aspirated and the cell-containing pellet was resuspended in 300 µL of PBS for further analysis. Flow cytometry was performed using a 13-color DxFlex flow cytometer (Beckman Coulter), and the data were analyzed using Kaluza software version 2.1 (Beckman-Coulter).

The antibody panel included the following reagents (all from Beckman-Coulter): CD19-fluorescein isothiocyanate (FITC; clone J4.119), CD123-phycoerythrin (PE; clone SSDCLY107D2), HLA-DR-phycoerythrin-Texas (ECD; clone Immu-357), CD56-phycoerythrin-cyanin 5.5 (PC5.5; clone N901), CD11c-phycoerythrin-cyanin 7 (PC7; clone BU15), CD14-allophycocyanin (APC; clone RMO52), CD3-APC-alexa fluor 750 (AA750; clone UCHT1), CD16-pacific blue (PB; clone 3G8), and CD45-krome orange (KO; clone J.33).

The gating strategy was adapted from that used in recent related publications [15, 16]. NK cells were categorized into CD56high, CD56+CD16+, and CD56+CD16− subsets. Monocyte populations were classified as classical monocytes (Mo1; CD14⁺CD16⁻), intermediate monocytes (Mo2; CD14⁺CD16⁺), and non-classical monocytes (Mo3; CD14dimCD16⁺). DCs were differentiated into plasmacytoid DCs (pDCs; CD123+) and myeloid DCs (mDCs; CD11c+), as is further detailed in Additional File 1.

Flow cytometry was used to quantify the relative proportions of immune cell subsets, while XN-9000 hematology analyzers (SYSMEX, Japan) were used to determine the total white blood cell and lymphocyte counts. The absolute counts for each subset were derived by multiplying these two measurements.

Establishment and validation of reference intervals

A cohort of 320 healthy volunteers, recruited between April 2023 and August 2024, was enrolled to establish the reference intervals. The analytical process involved three key steps. First, potential outliers were identified and excluded through systematic screening. Second, distribution-based statistical methods were applied. For normally distributed parameters, the 95% reference range was defined as x̄ ± 1.96 × standard deviation, while nonparametric percentile methods (P2.5–P97.5) were used for any skewed distributions.

The established reference intervals were validated using data from 40 additional healthy volunteers who were recruited later, between September to October of 2024. Each age group for this cohort consisted of 20 participants (10 men, 10 women). The validation criterion required that the proportion of results that fell outside of the reference intervals should not exceed 5% in any subgroup.

Statistical analysis

All analyses were performed using R version 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria). Outlier detection followed the 1/3 Rule: data points were sorted in ascending order, and potential outliers were identified by calculating the difference ratios between adjacent values (D) and the total data range (R). If any raw data values for the proportions or absolute counts of peripheral blood NK cells, monocytes, or DC subsets exceeded the D/R × 1/3 threshold, that participant’s entire dataset was excluded as an outlier.

Normality was assessed using both skewness and kurtosis tests. Inter-group comparisons were conducted using the Z-test for normally distributed data, whereas the Kruskal–Wallis test was used for non-normally distributed data. We evaluated the necessity of establishing age-stratified reference intervals based on the outcomes of our inter-group comparisons. Sex-stratified linear regression models were used to investigate the associations among age and sex and the proportions and absolute counts of peripheral blood NK cell, monocyte, and DC subsets. All tests used α = 0.05 as the threshold for statistical significance.

Results

Establishment of reference intervals

After excluding all outliers, 316 healthy volunteers were enrolled to establish the reference intervals, comprising 155 participants in the 18–40 year age range (79 men and 76 women) and 161 in the 41–65 year group (77 men and 84 women). Skewness and kurtosis tests revealed that all cellular subset percentages and absolute counts followed a non-normal distribution. Therefore, the reference intervals were determined using the P2.5–P97.5 approach. The reference ranges for the percentages and absolute counts of NK cell, monocyte, and DC subsets in the different age groups are presented in Table 1.

Table 1 Reference intervals for the proportions and absolute counts of NK cell, monocyte, and DC subsets in different age groups
Full size table

Age validation of reference intervals

Measurements from the 40 healthy volunteers used to perform the reference interval validation were compared with the established reference intervals. The raw data for this validation analysis are presented in Additional File 2. The pre-analytical and analytical factors were consistent with the reference intervals we established. Based on the results of our inter-group comparisons, we determined that the proportions of NK cells, as well as the proportions and absolute counts of CD56high NKs, Mo1s, Mo3s, DCs, pDCs, and mDCs should be stratified by age as different grouped reference intervals. All of the results achieved ≥ 95% conformity rates (Additional File 3).

Sex-stratified linear regression analysis

Figure 1 and Additional File 4 demonstrate the sex-stratified linear regression analysis of age-related variations in both proportions and absolute counts of NK cell, monocyte, and DC subsets. After adjusting for sex, age exhibited a ​significant positive association with the proportions of NK cells (β = 0.167, p = 0.029) and the MO3 subset (both proportion and absolute count, p < 0.05), while showing a ​significant negative association with the absolute counts of pDC and mDC cells, as well as the proportions and absolute counts of the MO1 and DC cells (p < 0.05). All age-by-sex interaction terms demonstrated insignificant moderation effects (p > 0.05; Additional File 4), suggesting that ​sex does not significantly modulate age-related changes in the dynamics of NK cell, monocyte, and DC subsets.

Fig. 1
Fig. 1
Full size image

Sex-stratified linear regression analysis. a–k Proportions of natural killer cell, monocyte, and dendritic cell subsets. l–v Absolute counts of natural killer cell, monocyte, and dendritic cell subsets NK, natural killer; Mo1, classical monocytes; Mo2, intermediate monocytes; Mo3, non-classical monocytes; DC, dendritic cells; pDC, plasmacytoid dendritic cells; mDC, myeloid dendritic cells

Discussion

Reliable reference intervals for immune cell subsets serve as essential benchmarks for evaluating immune function, diagnosing immune-related disorders, and monitoring treatment responses. Region-specific reference ranges significantly enhance clinical applicability and diagnostic accuracy in this regard. Flow cytometry, recognized as the “gold standard” for immunophenotyping, owing to its high sensitivity and multiplex analysis capabilities [17], was used in this study. This study established the first age-stratified reference intervals for peripheral blood monocytes, NK cells, and DC subsets (including both percentages and absolute counts) in healthy adults from Zhejiang Province, China—which were then validated in an independent cohort with a 5% exclusion threshold. These findings provide critical evidence for the development of regional precision medicine strategies and optimization of immune evaluation frameworks.

Compared to well-established reference intervals for T/B/NK lymphocyte subsets, research into reference ranges for monocyte, NK cell, and DC subsets remains in its early stages. Regarding NK cells, Campos et al. conducted a small-scale study that focused on age-related changes in NK cell subpopulations in a population of Spanish adults aged > 70 years old [18]. Although Phan et al. established NK cell reference ranges in healthy individuals from South Korea, their study lacked an age-stratified analysis [19]. Regarding monocyte subsets, Seidler et al. observed an age-dependent increase in non-classical monocytes among German adults, which was consistent with our findings [20]. For DC subsets, Heinze et al. only established reference ranges in a German pediatric population, meaning that adult counterparts for the region remain lacking [12]. An Italian longitudinal study of 69 subjects revealed significant age-related declines in absolute pDC and mDC counts [21]. Gu et al. established reference intervals for monocyte, NK cell, and DC subsets that primarily focused on the northern Han population of Chinese adults [22]. When these are compared with those for our southeastern Han cohort, several notable differences emerge: (1) Our study produced a higher upper limit for non-classical monocytes; (2) We had broader reference ranges for total NK cell percentages and absolute counts; and (3) Our subjects had significantly higher pDC and lower mDC proportions in the DC subsets. These inter-study discrepancies likely result from multifactorial influences, including geographical and methodological heterogeneity from different detection platforms, antibody panels, reagent sources, and gating strategies. The stepwise gating strategy implemented in our working panel significantly enhanced its analytical accuracy regarding DC populations, by prioritizing the exclusion of highly abundant and confounding cell types (e.g., NK cells and monocytes).

The establishment of reference intervals for monocyte, NK cell, and DC subsets facilitates the systematic investigation of immune dysregulation-disease associations in specific populations. For instance, patients with controlled h​​uman ​​i​​mmunodeficiency ​​v​​irus (HIV) infection demonstrate significantly elevated CD56+CD16− NK cell subsets vs. healthy controls [23], suggesting their critical role in HIV suppression through cytolytic activity. In hematological diagnostics, monocyte subset redistribution serves as a sensitive biomarker: a > 94.0% fraction of classical monocytes within the total monocyte population exhibits high specificity for diagnosing chronic myelomonocytic leukemia, thus facilitating prompt differentiation from overlapping myeloid disorders [24]. Patients with progressive vitiligo exhibit DC subset al.terations that correlate strongly with inflammatory cytokine cascades [25]. These reference intervals offer multidimensional clinical value by facilitating both the identification of early biomarkers for infections, malignancies, autoimmune diseases, and metabolic disorders through immune subset dynamics quantification, as well as concurrently clarifying the state of tissue microenvironment remodeling via migratory immune cell patterns such as activated vs. resident phenotypes. However, the distribution patterns between peripheral blood and tissue-infiltrating immune cells may exhibit incomplete synchronization [26, 27]. Further investigations into this notion might consider leveraging single-cell sequencing to characterize immune cell transcriptomes alongside dynamic tissue-residency marker expression (e.g., CD69, CD49a, CD103), thereby precisely elucidating the mechanisms through which peripheral and tissue immunity interact.

This study was subject to several limitations that warrant emphasizing. First, although the dual-platform methodology we used for establishing absolute counts of cellular subsets offers advantages in terms of cost-efficiency and reducing technical dependency, it has inferior inter-laboratory consistency compared to single-platform approaches [28]. Therefore, optimizing the dual-platform approach can help minimize inter-laboratory variability [29]. Second, the sample size of the present study constrained its statistical power in terms of age-stratified subgroup analyses, precluding a comprehensive exploration of the reference intervals across additional demographic strata while maintaining subgroup robustness. Furthermore, although rigorous inclusion and exclusion criteria were applied to recruit ostensibly healthy participants, the adjustment solely for sex—without accounting for covariates such as body mass index and lifestyle factors—may have introduced some degree of residual confounding bias concerning immune phenotype characterization. Finally, given the geographical influence on immunophenotypic parameters, region-specific validation is strongly recommended before our reference intervals can be applied to different populations. Future investigations should consider prioritizing cohort expansion to facilitate more refined subgroup stratification, incorporate multivariable regression frameworks to adjust for key confounders, and validate our findings in prospective multicenter cohorts.

Conclusion

This study established reference intervals for peripheral blood NK cell, monocyte, and DC subsets in healthy adults from China’s Zhejiang Province, and their reliability and clinical applicability were confirmed through an independent validation cohort study. Our findings are essential for advancing precision diagnostics and therapeutic strategies in clinical practice.