Establishing reference interval of serum CA-125 for the healthy Han population in China.

After applying the predefined clinical and laboratory exclusion criteria and removing 170 statistical outliers, 9,871 serum CA-125 results from apparently healthy Han adults (7,688 females and 2,183 males) were included in the analysis. The age and sex distributions of the study population are shown in Fig.  1 . The D’Agostino–Pearson test indicated that the serum CA-125 levels were not normally distributed. Fig. 1 Histograms of age and CA-125 concentration (U/mL) distributions among subjects Histograms of age and CA-125 concentration (U/mL) distributions among subjects Using the robust method recommended by CLSI C28-A3, we established sex-specific reference intervals. The serum CA-125 values differed significantly between males and females ( P  < 0.0001), and the median CA-125 level in males was lower than that in females (Fig.  2 ). The overall reference intervals for serum CA-125 were 0–33.1 U/mL for females and 0–19.7 U/mL for adult males (Table  1 ). Fig. 2 CA-125 (U/mL.) levels between females and males CA-125 (U/mL.) levels between females and males Table 1 Reference intervals of cancer antigen 125 (CA-125) for the different age and sex groups Group Age group CA-125(U/mL) N Median Upper limit (90% Cl) Female 18 ≤ age ≤ 30 1655 11.9 33.1 (32.0-34.3) 31 ≤ age ≤ 70 5852 10.1 28.1 (27.6–28.8) 71 ≤ age ≤ 90 181 8.9 21.8 (19.9–24.1) Male 18 ≤ age ≤ 87 2183 8.20 19.7 (19.2–20.3) 97.5% Reference interval, Right-sided. Robust method (CLSI C28-A3) CI Confidence interval Reference intervals of cancer antigen 125 (CA-125) for the different age and sex groups 33.1 (32.0-34.3) 28.1 (27.6–28.8) 21.8 (19.9–24.1) 19.7 (19.2–20.3) 97.5% Reference interval, Right-sided. Robust method (CLSI C28-A3) CI Confidence interval Age-related patterns in female CA-125 levels are shown in Fig.  3 . The data support the partitioning of women into three age groups (18–30, 31–70 and 71–90 years). The upper reference limits were 33.1 U/mL for women aged 18–30 years, 28.1 U/mL for those aged 31–70 years and 21.8 U/mL for those aged 71–90 years (Table  1 ). In women, CA-125 showed a modest but significant negative correlation with age (Spearman r = − 0.2488, P  < 0.001), whereas no significant correlation with age was observed in men. Fig. 3 Age-related reference intervals for CA-125 (U/mL). The upper blue lines represent the 95.0th percentiles, and the dashed lines are 90% confidence intervals for the upper limits of the reference intervals. ( A : Female; B : Male) Age-related reference intervals for CA-125 (U/mL). The upper blue lines represent the 95.0th percentiles, and the dashed lines are 90% confidence intervals for the upper limits of the reference intervals. ( A : Female; B : Male)

Shuyang Hospital provides routine health examinations to adults from the surrounding region. For this retrospective study, we searched the laboratory information system (LIS) for all Han Chinese adults aged 18–90 years who underwent health checks between October 2018 and December 2023 and whose serum CA-125 levels were measured. This search identified 10,076 individuals. At the time of the health check, examinees complete standard registration forms and a structured questionnaire and undergo a physical examination. Age, sex and self-reported ethnicity are recorded on these forms and stored in the LIS. We extracted these demographic variables together with laboratory results for the present analysis. Reference interval samples were selected according to the CLSI C28-A3 guidelines and our previous work [ 11 – 13 ]. To assemble a reference population, we applied predefined clinical and laboratory criteria. The exclusion criteria included self-reported or clinically documented chronic diseases (such as cardiovascular, hepatic, renal, autoimmune or neoplastic disorders), acute infection, pregnancy and abnormal findings on routine hematological or biochemical testing. Routine tests included complete blood count (white blood cell count, haemoglobin and platelet count), liver function tests (alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, γ-glutamyl transferase, total bilirubin and albumin), renal function tests (serum creatinine and urea), and fasting plasma glucose and C-reactive protein. Abnormality was defined as a value outside the locally established reference interval for that test. These criteria are routinely used in our health-check programme to define apparently healthy examinees. On this basis, 35 individuals were excluded, leaving 10,041 Han adults whose serum CA-125 results were available. Two laboratory medicine physicians independently reviewed the LIS and health-check records to apply the clinical and laboratory criteria, and discrepancies were resolved by consensus. After this review, we screened the remaining 10,041 CA-125 results for statistical outliers (see below). After all the exclusions, 9,871 apparently healthy Han adults (7,688 women and 2,183 men) were included in the final reference interval analysis. The selection process is summarised in Supplementary Fig. 1. Because this was a retrospective LIS-based study, some determinants of CA-125 could not be evaluated. In particular, body mass index, use of oral contraceptives, menstrual cycle phase, parity and benign gynecological conditions were not consistently recorded in the health-check database and therefore could not be incorporated into the inclusion criteria or statistical analyses. Venous blood samples were collected in the morning after an overnight fast as part of the routine health check before any diagnostic or therapeutic intervention. Serum CA-125 was measured via a DxI 800 chemiluminescent immunoassay analyser (Beckman Coulter, USA) [ 12 ]. In accordance with the manufacturer’s instructions, the assay has a limit of detection of 0.6 U/mL and a measurement range of 0.6–5,129 U/mL. Regular internal quality control procedures and participation in an external quality assessment scheme were used throughout the study period to monitor the analytical performance of CA-125 measurements. The study protocol was reviewed and approved by the Ethics Committee of Shuyang Hospital (SYYYLL202424). The committee classified the project as minimal-risk research on the basis of retrospective analysis of deidentified health-check data and banked serum samples and granted a waiver of written informed consent. Health-check clients at our institution are routinely informed that anonymised data and residual samples may be used for scientific research, with the option to decline; no such requests were recorded for the study period. Statistical analyses followed the CLSI C28-A3 recommendations [ 13 ]. The distribution of CA-125 values was assessed via the D’Agostino–Pearson test, which revealed that the data were not normally distributed. To exclude outliers, we first removed results with clear analytical or biological problems. Statistical outliers were then identified within each sex and age partition via Tukey’s method: values below the first quartile (Q1) minus 1.5 × interquartile range (IQR) or above the third quartile (Q3) plus 1.5 × IQR were considered outliers [ 10 ]. For values close to these cut-off values, the Dixon D/R ratio was also examined. Overall, 170 CA-125 values were classified as outliers and removed before the reference intervals were calculated. Sex-specific reference intervals were established for males and females. For females, age-related changes in CA-125 were evaluated, and the data were divided into three age groups (18–30, 31–70 and 71–90 years) on the basis of the observed distribution and statistical criteria. Reference intervals were calculated via the robust method to obtain right-sided 95% upper reference limits and their 90% confidence intervals, in line with CLSI guidance [ 13 ]. The correlation between age and CA-125 level was assessed with Spearman’s rank correlation coefficient. All the statistical analyses were performed via EXCEL 2019 (Microsoft Corporation, Beijing, China) and MedCalc 18.2.1 (MedCalc Software, Ostend, Belgium). A two-sided P value < 0.05 was considered statistically significant.

Establishing appropriate reference intervals is essential for the correct interpretation of laboratory tests [ 14 ]. CA-125 is one of the most frequently requested tumour markers in clinical practice, especially in gynaecology and oncology [ 2 ]. However, the upper reference limit of 35 U/mL recommended by many manufacturers is largely based on historical data from Western populations, and there is increasing recognition that reference intervals may vary by population, age and assay [ 5 – 9 ]. Using CLSI-based methods and a large health-check cohort, we defined sex-specific and age-partitioned reference intervals for serum CA-125 in Han Chinese adults measured on the Beckman DxI 800 platform. Our data confirmed that CA-125 concentrations are higher in women than in men. The overall upper reference limit for men was 19.7 U/mL, whereas women had higher values and a clear age pattern. In women, CA-125 showed a modest but significant negative correlation with age, and the upper reference limits were 33.1 U/mL for those aged 18–30 years, 28.1 U/mL for those aged 31–70 years and 21.8 U/mL for those aged 71–90 years. These findings support the use of sex-specific and age-partitioned reference intervals rather than a single universal cut-off. Compared with previous reports, our female upper reference limits are generally lower than the widely used threshold of 35 U/mL, particularly in older age groups [ 5 – 9 ]. Large European studies have reported female CA-125 reference limits close to 35 U/mL, whereas in our Han cohort, the upper limit in women aged 71–90 years was only 21.8 U/mL [ 9 ]. Studies from other East Asian populations have also described lower upper reference limits and age-related decreases in CA-125 [ 5 – 8 ]. Taken together, these observations suggest that population, ethnicity and age all influence CA-125 distributions and that a single “one-size-fits-all” cut-off may not be optimal for every setting. Moreover, differences in study design, inclusion criteria and assay calibration mean that cross-study comparisons should be interpreted with caution. Although CA-125 is best known as a marker for ovarian cancer, it can be affected by several physiological and benign conditions. In women, CA-125 varies across the menstrual cycle and increases during pregnancy [ 2 ]. Benign gynecological diseases such as uterine fibroids, adenomyosis and endometriosis can also increase CA-125 levels in the absence of malignancy. In men and women, CA-125 may increase in heart failure, peritoneal involvement and some inflammatory or autoimmune diseases. In our study, we tried to reduce the impact of overt disease by excluding individuals with documented chronic conditions or abnormal results of routine laboratory tests. However, our retrospective design did not allow us to fully account for body mass index, hormonal contraceptive use, menstrual phase, parity or benign gynecological diagnoses, as these data were not consistently recorded in the health-check database. In particular, undiagnosed or asymptomatic endometriosis cannot be ruled out and may have contributed to mildly elevated CA-125 in a small fraction of women. Several limitations should be considered when applying our reference intervals. First, this was a single-centre study, and all measurements were performed on one immunoassay platform. The intervals we report are therefore most directly applicable to Han adults undergoing health checks at our institution and to laboratories using the same analyser and reagents. Other centers should verify these limits locally, in line with CLSI recommendations [ 13 ]. Second, several important determinants of CA-125, including body mass index, hormonal contraceptive use and detailed gynecological history, were incompletely captured in the retrospective database. Third, although the overall sample size was large, the oldest female age group contained fewer individuals than would be ideal, and more finely stratified age partitions would require larger cohorts. Finally, we did not perform external validation in an independent cohort or in other ethnic groups. Despite these limitations, the study has several strengths. We analysed a large, real-world health-check population, applied predefined clinical and laboratory criteria to select apparently healthy individuals, and followed the CLSI C28-A3 guidance for reference interval estimation. We also provide a flow diagram of participant selection and a transparent description of outlier handling, which should help other laboratories understand and, if needed, replicate our approach. In summary, we established sex-specific and age-specific reference intervals for serum CA-125 in a large cohort of Han Chinese adults measured via the Beckman DxI 800 system. In this population, CA-125 concentrations are higher in women than in men and tend to decrease with age in women but remain relatively stable in men. These locally derived intervals may improve the interpretation of CA-125 results in Han adults, but they should be verified in other centers and refined in future prospective studies that collect detailed information on body mass index, hormonal status and benign gynecologic conditions.

Cancer antigen 125 (CA-125) is a glycoprotein biomarker that has gained considerable attention in oncology because of its role in the diagnosis, treatment and monitoring of several malignancies [ 1 ]. CA-125 was first identified in patients with epithelial ovarian cancer and is therefore predominantly associated with female reproductive system tumours. Elevated serum CA-125 is also observed in other malignancies, including cancers of the endometrium, fallopian tube and colon, as well as in a range of benign conditions [ 2 , 3 ]. In clinical practice, CA-125 is most often measured in serum. It is widely used as a marker for the detection and follow-up of epithelial ovarian cancer and for monitoring the response to chemotherapy [ 1 ]. Decreasing CA-125 levels during treatment usually indicate a favourable response, whereas increasing levels may signal treatment resistance or disease recurrence and prompt further investigation. However, CA-125 is not specific for malignancy and can be increased in physiological states, such as menstruation and pregnancy, or in benign gynecological and inflammatory conditions [ 2 ]. Interpretation of CA-125 therefore requires an appropriate reference interval. At present, the upper reference limit of 35 U/mL recommended by many manufacturers is largely based on early studies conducted in American and European populations [ 4 ]. Data for Chinese populations remain limited, and there is increasing evidence that reference intervals for tumour markers may vary by ethnicity, age and analytical platform used [ 5 – 14 ]. Establishing locally derived reference intervals in well-defined healthy populations is essential for improving test interpretation and avoiding both overdiagnosis and missed disease. The aim of this study was to establish sex- and age-specific reference intervals for serum CA-125 in apparently healthy Han adults via real-world health-check data from Shuyang Hospital in accordance with the CLSI C28-A3 guidelines [ 10 ].

Supplementary Material 1: Supplementary Figure 1. Flow diagram of participant selection for establishing serum CA-125 reference intervalsA total of 10,076 Han adults whose serum CA-125 levels were measured during routine health checks were identified from the laboratory information system. Thirty-five individuals were excluded on the basis of predefined clinical and routine laboratory criteria, leaving 10,041 eligible Han adults. After 170 statistical outliers were removed according to Tukey’s rule and the Dixon test, 9,871 apparently healthy Han adults (7,688 females and 2,183 males) were included in the final reference interval analysis. Supplementary Material 1: Supplementary Figure 1. Flow diagram of participant selection for establishing serum CA-125 reference intervalsA total of 10,076 Han adults whose serum CA-125 levels were measured during routine health checks were identified from the laboratory information system. Thirty-five individuals were excluded on the basis of predefined clinical and routine laboratory criteria, leaving 10,041 eligible Han adults. After 170 statistical outliers were removed according to Tukey’s rule and the Dixon test, 9,871 apparently healthy Han adults (7,688 females and 2,183 males) were included in the final reference interval analysis. Supplementary Material 2. Supplementary Material 2.