Abstract
Objectives
A large number of people undergo annual health checkup but accurate laboratory criterion for evaluating their health status is limited. The present study determined annual biological variation (BV) and derived parameters of common laboratory analytes in order to accurately evaluate the test results of the annual healthcare population.
Methods
A total of 43 healthy individuals who had regular healthcare once a year for six consecutive years, were enrolled using physical, electrocardiogram, ultrasonography and laboratory. The annual BV data and derived parameters, such as reference change value (RCV) and index of individuality (II) were calculated and compared with weekly data. We used annual BV and homeostatic set point to calculate personalized reference intervals (RIper) which were compared with population-based reference intervals (RIpop).
Results
We have established the annual within-subject BV (CVI), RCV, II, RIper of 24 commonly used clinical chemistry and hematology analytes for healthy individuals. Among the 18 comparable measurands, CVI estimates of annual data for 11 measurands were significantly higher than the weekly data. Approximately 50% measurands of II were <0.6, the utility of their RIpop were limited. The distribution range of RIper for most measurands only copied small part of RIpop with reference range index for 8 measurands <0.5.
Conclusions
Compared with weekly BV, for annual healthcare individuals, annual BV and related parameters can provide more accurate evaluation of laboratory results. RIper based on long-term BV data is very valuable for “personalized” diagnosis on annual health assessments.
Introduction
The number of people subscribing to annual healthcare is increasing year by year and nearly 435 million in China in 2018 [1]. Of this population, approximately 33% have a good health status [2]. One of the main methods of disease detection for healthcare population is using biomarkers including clinical chemistry and hematology analytes. These biomarkers play central roles in diagnosing and monitoring common diseases such as liver and kidney diseases, diabetes mellitus and cardiovascular diseases [3], [4], [5]. Therefore, it is very important to accurately evaluate the results of these measurands for monitoring health status.
To assess the results of these measurands, population-based reference intervals (RIpop) is the most widely used method. But many measurands show marked individuality, meaning that a real fluctuation range of an individual only occupies a small part of RIpop. To address this issue, biological variation (BV) and associated parameters provide more objective evaluation criteria [6, 7]. Based on within-subject biological variation (CVI), reference change value (RCV) can be calculated to estimate whether changes in a series of tests for an individual are clinically significant. Personalized reference interval (RIper) can be used to determine whether the actual result is indicative of disease or not for a specific individual [8, 9]. Compared with RIpop, RIper may provide more accurate interpretation of laboratory results, which would help the development of personalized medicine in the future.
The current authoritative BV database is found online at the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM), where most studies cited in the database were medium-term biological variations (blood sampling once/week, several weeks) [10]. For annual healthcare population, medium-term data may not be appropriate for the interpretation of laboratory results. A few long-term BV studies have been published, such as monthly BV for hematology analytes [11] and annual BV for bone turnover markers [12]. However, annual BV for routine measurands is lacking.
To address these issues, we investigated healthy individuals who had continuous regular healthcare for six years, and got the annual BV as well as derived parameters of commonly used clinical chemistry and hematology analytes. The RIper calculated based on annual BV will be very valuable for the “personalized” assessment of long-term health status of individuals.
Materials and methods
Annual biological variation
Study population
This study was approved by the ethics committee of the First Hospital of China Medical University. In total, 1,046 individuals who had regular physical examinations in the healthcare center of the First Hospital of China Medical University for six consecutive years were selected. A total of 43 healthy individuals (21 males and 22 females) were enrolled. The procedures for selection of the study individuals were shown in Supplementary Figure 1.
Screening criteria
Physical examinations were performed by physicians. The exclusion criteria each year were as followings: hypertension (systolic pressure ≥140 mmHg and/or diastolic pressure ≥90 mmHg), Obese individuals (body mass index [BMI] ≥30 kg/m2) [13]. Serious arrhythmia and heart disease (ventricular fibrillation, atrial fibrillation, ventricular flutter, or premature beat) by electrocardiogram (ECG). Fatty liver, liver fibrosis, liver cirrhosis, occupying lesions (liver, pancreas, renal), acute or chronic pyelonephritis, acute or chronic pancreatitis by ultrasonography. Pregnant or lactating.
Laboratory results each year meet the following requirements: Fasting glucose (Glu) <7.0 mmol/L; alanine aminotransferase (ALT) <80 U/L; γ-glutamyl transferase (GGT) <150 U/L; serum creatinine (Crea): <106 mmol/L (males), <98 mmol/L (females); albumin (ALB) >35 g/L; white blood cell (WBC) <10 × 109/L; hemoglobin (Hb): ≥120 g/L (males), and ≥110 g/L (females).
Sample collection and measurement
Fasting blood samples were collected in the morning from cubital vein. Blood samples were dispensed into 5 mL SST gel tubes (Becton Dickinson, USA) and centrifuged at 1,200g for 10 min within 2 h after collection to separate serum for clinical chemistry analytes. For hematology analytes, samples were dispensed into 2 mL K2EDTA tubes (Becton Dickinson, USA). Samples were analyzed only once within 2 h of collection. Roche automatic biochemical analyzer (Roche, Germany) and Sysmex hematology analyzers (Sysmex, Japan) were used for analysis. These instruments remained unchanged during the previous six years. measurands for annual BV included liver function tests (ALT, aspartate aminotransferase [AST], alkaline phosphatase [ALP], GGT, total protein [TP], ALB, total bilirubin [TBIL], and total bile acids [TBA]), renal function tests (urea, Crea, and uric acid [UA]), serum lipids (high density lipoproteins-cholesterol [HDL-C], low density lipoproteins-cholesterol [LDL-C], total cholesterol [TC], and triglycerides [TG]), Glu, complete blood count (CBC) (WBC, red blood cell [RBC], Hb, hematocrit [HCT], mean cellular volume [MCV], mean cellular hemoglobin [MCH], mean cellular hemoglobin concentration [MCHC], platelets [PLT]) (Supplementary Table 1). The analytical methods were shown in Supplementary Table 2. These analytical measurements passed governmental external quality assessment (EQA) every year.
Weekly biological variation
The individual selection and sample processing for our weekly BV study were previously published [14]. In brief, 25 healthy volunteers (8 males and 17 females) were selected. After blood collection once a week, serum was separated and stored at −80 °C for six consecutive weeks. After sampling was completed, we analyzed samples in duplicate in the same run. All measurands above were detected in weekly BV study except for TBA, TG, HDL-C, LDL-C, UA and Glu (Supplementary Table 1). The first result from the two parallel measurements of weekly variation was selected to calculate by the same way as those used for the annual variation. All instruments were the same as for annual BV assessments.
Statistical analysis
The same statistical method was used for annual and weekly BV assessments. Before data analysis, Cochran’s and Reed’s tests were used to exclude outliers in two steps [6]: (a) outliers among within-subject variances. Cochran’s test was performed to examine the ratio of the maximum variance to the sum of variances. If an outlier was identified, we excluded all data of that subject; (b) Reed’s method considered the difference between the extreme value and the next highest (or lowest) value and rejected the extreme value if this difference exceeded one-third of the range of all values. CVI and between-subject BV (CVG) estimates were calculated using the method by Fraser and Harris [7]. Briefly, total within-subject variances were calculated and included both biological and analytical components S2
(I+A). The within-subject variance (S2
I) was obtained by S2
(I+A) minus the analytical variance (S2
A) using the formula: S2
I = S2
(I+A) – S2
A. Between-subject variance (S2
G) was calculated using following formula: S2
G = S2
T – S2
I – S2
A (S2
T was the variance derived from the complete data matrix, all data from all subjects) [15]. Standard Deviation (SD) was transformed into the corresponding CV according to the mean value. The CVA was calculated as follows: for hematology analytes, three different lots of quality control (QC) materials were selected and each lot of QC material accumulated for at least two months. We calculated the average imprecision of each level of the three lot numbers, and the CV of QC material with similar average concentration to this study population was taken as its CVA. For clinical chemistry analytes, the imprecision of QC materials with similar average concentration of 12 months was taken as its CVA. The 95% confidence intervals (CIs) for BV estimates were calculated as described by Burdick [16]. The significant differences in subgroups were assessed by the overlap of the 95% CI. The calculation formula for index of individuality (II) and reference change value (RCV) were: II = CVI/CVG and RCV = Z*21/2*(CVA
2 + CVI
2)1/2, where Z is 1.96 for a p-value<0.05. The number of specimens required to determine the homeostatic set point for an individual was calculated using the formula: n = [Z × (CVA
2 + CVI
2)1/2/D]2, where n = number of specimens, and D = the percentage closeness to the homeostatic set point. RIper was calculated according to Coskun’s method [9]: RIper =
RIpop came from the National Guide to Clinical Laboratory Procedures (4th edition), which outlined currently recommended reference intervals in China [17]. To investigate the relationship between RIper from this study and RIpop, reference range index (RRi) was calculated based on the formula [9]: RRi =
Results
Participant characteristics
In total, 1,046 individuals were screened and 43 healthy individuals (21 males and 22 females) were selected into annual variation analysis, and 25 healthy individuals (8 males and 17 females) were enrolled in weekly variation analysis. Numbers of excluded subjects by Cochran’s and Reed’s tests were shown in Supplementary Table 3. The means of age for the two groups were 37 (range, 23–52), and 28 (range, 23–49) years, respectively. The means of BMI of the two groups were 22.1 (range, 18.0–26.8), and 21.4 (range, 18.0–25.2) kg/m2, respectively.
BV data for measurands
Annual BV and weekly BV data for each measurand were shown (Table 1). Significant gender differences in CVI estimates were found for ALT, ALP, urea, UA in the annual group and for RBC, Hb, and Hct in the weekly group. No gender differences in CVG estimates were found in both groups except for MCH in weekly group.
Biological variation estimates for all parameters for the annual and weekly groups.
| Parameters | Group | CVA, % | Annual group | Weekly group | EFLM databasea [10] | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n | Mean value (95% CI) | CVI (95% CI), % | CVG (95% CI), % | n | Mean value (95% CI) | CVI (95% CI), % | CVG (95% CI), % | CVI, % | CVG, % | |||
| ALT, U/L | All | 6.3 | 41 | 18.2 (17.2–19.4) | 33.2 (30.3–36.7) | 35.5 (28.0–46.8) | 22 | 13.9 (12.9–14.8) | 16.0 (14.0–18.4) | 38.9 (29.6–56.0) | 10.1 (9.3–15.6) | 29.3 (28.0–38.4) |
| Male | 21 | 20.6 (19.0–22.2) | 28.1 (24.7–32.5) | 35.0 (25.6–52.0) | 8 | 18.0 (16.3–19.7) | 13.9 (11.4–17.8) | 32.8 (21.1–67.8) | ||||
| Female | 20 | 15.7 (14.4–17.1) | 40.2 (35.3–46.8) | 27.6 (17.7–44.0) | 14 | 11.6 (10.5–12.6) | 17.5 (15.0–21.1) | 29.1 (20.4–47.6) | ||||
| AST, U/L | All | 5.7 | 42 | 19.6 (19.0–20.3) | 17.7 (16.2–19.6) | 21.8 (17.4–28.6) | 21 | 19.9 (19.1–20.8) | 10.8 (9.5–12.5) | 19.2 (14.3–28.2) | 9.6 (9.5–13.5) | 20.8 (20.3–23.8) |
| Male | 21 | 20.3 (19.4–21.2) | 15.3 (13.5–17.7) | 20.2 (14.8–29.9) | 8 | 21.9 (20.9–23.0) | 8.9 (7.4–11.5) | 22.6 (14.6–46.8) | ||||
| Female | 21 | 18.9 (17.8–20.0) | 20.1 (17.6–23.2) | 23.1 (16.8–34.5) | 13 | 18.6 (17.8–19.4) | 12.1 (9.3–13.2) | 11.4 (12.5–30.4) | ||||
| ALP, U/L | All | 2.0 | 42 | 60.7 (58.9–62.5) | 11.8 (10.8–13.1) | 20.7 (16.8–26.8) | 24 | 54.2 (52.2–56.3) | 5.7 (5.0–6.5) | 21.3 (16.5–29.9) | 5.3 (4.5–6.0) | 24.3 (19.9–26.5) |
| Male | 21 | 62.8 (60.6–65.2) | 8.9 (7.8–10.2) | 20.0 (15.1–29.1) | 7 | 64.8 (61.0–68.5) | 5.0 (4.1–6.5) | 14.9 (9.4–32.8) | ||||
| Female | 21 | 58.5 (55.9–61.3) | 14.5 (12.8–16.8) | 21.0 (15.6–30.9) | 17 | 49.8 (48.0–51.7) | 6.0 (5.2–7.0) | 18.5 (13.6–28.3) | ||||
| GGT, U/L | All | 3.2 | 41 | 22.3 (20.8–23.6) | 24.7 (22.6–27.3) | 47.9 (38.9–61.9) | 25 | 14.7 (13.8–15.7) | 8.2 (7.3–9.3) | 34.8 (27.1–48.4) | 9.1 (7.3–12.0) | 44.5 (40.2–47.0) |
| Male | 21 | 27.7 (25.7–29.8) | 24.6 (21.7–28.4) | 41.1 (31.0–60.7) | 8 | 20.1 (19.1–21.3) | 8.5 (7.0–10.9) | 20.1 (13.0–41.6) | ||||
| Female | 20 | 16.6 (15.3–18.0) | 22.0 (19.3–25.5) | 34.0 (25.1–50.5) | 17 | 12.1 (11.5–12.8) | 7.0 (6.1–8.3) | 26.5 (19.6–40.6) | ||||
| TP, g/L | All | 1.5 | 43 | 70.2 (69.8–70.7) | 3.1 (2.9–3.5) | 3.3 (2.6–4.3) | 25 | 70.3 (69.6–70.9) | 2.8 (2.5–3.2) | 4.6 (3.5–6.5) | 2.6 (2.3–2.7) | 4.6 (2.8–5.7) |
| Male | 21 | 70.6 (70.1–71.2) | 3.2 (2.8–3.6) | 3.5 (2.5–5.2) | 8 | 71.4 (70.5–72.3) | 2.0 (1.6–2.6) | 2.7 (1.6–5.7) | ||||
| Female | 22 | 69.9 (69.4–70.5) | 3.1 (2.8–3.6) | 3.0 (2.1–4.5) | 17 | 69.8 (69.0–70.9) | 3.1 (2.7–3.6) | 5.2 (3.7–8.0) | ||||
| ALB, g/L | All | 1.1 | 43 | 44.7 (44.3–45.0) | 4.4 (4.0–4.8) | 3.0 (2.2–4.0) | 25 | 44.5 (44.2–45.0) | 3.0 (2.6–3.4) | 4.5 (3.4–6.3) | 2.6 (2.2–3.9) | 5.1 (2.2–6.3) |
| Male | 21 | 45.3 (44.9–45.6) | 3.9 (3.4–4.5) | 1.9 (0.9–3.2) | 8 | 45.7 (45.0–46.1) | 2.6 (2.1–3.3) | 2.6 (1.4–5.6) | ||||
| Female | 22 | 44.1 (43.6–44.6) | 4.8 (4.2–5.5) | 3.3 (2.2–5.2) | 17 | 44.0 (43.5–44.6) | 3.1 (2.7–3.7) | 4.7 (3.4–7.4) | ||||
| TBA, µmol/L | All | 4.1 | 41 | 2.6 (2.4–2.9) | 54.4 (49.7–60.1) | 54.2 (42.6–71.7) | ||||||
| Male | 20 | 2.8 (2.5–3.2) | 50.8 (44.6–59.1) | 63.4 (46.1–95.0) | ||||||||
| Female | 21 | 2.5 (2.1–2.7) | 58.4 (51.4–67.4) | 38.9 (25.0–61.4) | ||||||||
| TBIL, µmol/L | All | 7.2 | 41 | 13.3 (12.7–13.8) | 19.6 (17.9–21.7) | 26.9 (21.6–35.1) | 25 | 12.4 (11.7–13.1) | 18.6 (16.6–21.2) | 24.6 (18.5–35.0) | ||
| Male | 21 | 14.9 (14.2–15.6) | 18.6 (16.4–21.5) | 23.5 (17.1–34.9) | 8 | 15.2 (13.7–16.5) | 17.1 (14.1–21.9) | 18.3 (10.5–39.6) | ||||
| Female | 20 | 11.5 (10.9–12.1) | 20.9 (18.3–24.3) | 24.1 (17.3–36.4) | 17 | 11.1 (10.4–11.7) | 19.5 (16.9–22.9) | 19.6 (13.4–31.4) | ||||
| Urea, mmol/L | All | 1.7 | 43 | 4.7 (4.5–4.9) | 17.8 (16.2–19.6) | 20.0 (15.9–26.0) | 25 | 4.7 (4.5–4.9) | 12.6 (11.2–14.4) | 21.1 (16.1–29.7) | 13.9 (9.5–14.4) | 21.0 (13.4–22.5) |
| Male | 21 | 5.2 (5.1–5.5) | 14.9 (13.1–17.2) | 16.5 (11.9–24.6) | 8 | 5.2 (4.9–5.5) | 9.9 (8.1–12.7) | 14.7 (9.2–30.9) | ||||
| Female | 22 | 4.2 (4.0–4.4) | 21.1 (18.7–24.4) | 16.4 (11.3–25.1) | 17 | 4.5 (4.3–4.7) | 14.1 (12.2–16.6) | 22.4 (16.2–34.8) | ||||
| Crea, µmol/L | All | 1.4 | 43 | 61.8 (60.0–63.6) | 6.3 (5.7–7.0) | 20.0 (16.5–25.6) | 25 | 63.2 (61.0–65.0) | 4.3 (3.8–4.8) | 19.1 (14.9–26.6) | 4.5 (4.4–5.7) | 14.3 (7.0–17.4) |
| Male | 21 | 72.5 (70.9–74.6) | 5.7 (5.1–6.6) | 10.9 (8.2–15.9) | 8 | 78.1 (76.0–80.6) | 3.9 (3.2–5.0) | 8.0 (5.1–16.7) | ||||
| Female | 22 | 51.5 (50.6–52.6) | 7.0 (6.2–8.1) | 9.7 (7.2–14.2) | 17 | 56.1 (54.8–57.8) | 4.4 (3.8–5.2) | 11.4 (8.4–17.5) | ||||
| UA, µmol/L | All | 1.3 | 43 | 288 (279–297) | 10.0 (9.1–11.0) | 21.8 (17.8–27.9) | ||||||
| Male | 21 | 331 (321–341) | 7.9 (7.0–9.2) | 14.3 (10.7–20.9) | ||||||||
| Female | 22 | 247 (238–257) | 12.5 (11.0–14.4) | 18.6 (13.9–27.0) | ||||||||
| TG, mmol/L | All | 1.6 | 42 | 0.97 (0.91–1.04) | 31.5 (28.8–34.8) | 38.7 (30.9–50.6) | 20.0 (18.9–21.9) | 37.1 (23.4–40.3) | ||||
| Male | 20 | 1.17 (1.08–1.27) | 32.3 (28.3–37.5) | 36.7 (26.4–55.3) | ||||||||
| Female | 22 | 0.79 (0.75–0.85) | 28.1 (24.9–32.5) | 24.7 (17.4–37.1) | ||||||||
| TC, mmol/L | All | 1.3 | 43 | 4.7 (4.6–4.8) | 10.2 (9.3–11.3) | 11.9 (9.5–15.5) | 25 | 4.4 (4.3–4.5) | 5.4 (4.8–6.1) | 13.2 (10.2–18.5) | 5.3 (5.1–6.4) | 16.3 (14.2–17.4) |
| Male | 21 | 4.7 (4.5–4.8) | 9.2 (8.1–10.6) | 12.9 (9.6–19.1) | 8 | 4.4 (4.3–4.6) | 5.1 (4.2–6.5) | 10.2 (6.5–21.1) | ||||
| Female | 22 | 4.7 (4.6–4.8) | 11.1 (9.8–12.8) | 10.9 (7.9–16.3) | 17 | 4.4 (4.3–4.6) | 5.5 (4.8–6.5) | 14.5 (10.7–22.3) | ||||
| HDL-C, mmol/L | All | 4.8 | 43 | 1.50 (1.44–1.54) | 9.5 (8.6–10.5) | 23.9 (19.6–30.6) | 5.8 (5.7–8.8) | 24.5 (18.4–51.4) | ||||
| Male | 21 | 1.29 (1.22–1.35) | 8.7 (7.7–10.1) | 23.1 (17.5–33.7) | ||||||||
| Female | 22 | 1.68 (1.61–1.75) | 9.7 (8.6–11.2) | 17.6 (13.2–25.6) | ||||||||
| LDL-C, mmol/L | All | 2.3 | 43 | 2.8 (2.8–2.9) | 12.2 (11.2–13.5) | 18.8 (15.2–24.2) | 8.3 (6.8–10.3) | 26.1 (18.5–37.7) | ||||
| Male | 21 | 3.0 (2.8–3.1) | 11.8 (10.4–13.6) | 20.8 (15.6–30.5) | ||||||||
| Female | 22 | 2.7 (2.6–2.8) | 12.7 (11.2–14.6) | 15.4 (11.3–22.6) | ||||||||
| GLU, mmol/L | All | 1.3 | 43 | 5.1 (5.0–5.2) | 5.5 (5.0–6.1) | 6.0 (4.7–7.8) | 5.0 (4.1–12.0) | 8.1 (2.7–10.8) | ||||
| Male | 21 | 5.1 (5.0–5.2) | 5.9 (5.2–6.8) | 6.2 (4.5–9.3) | ||||||||
| Female | 22 | 5.1 (5.0–5.2) | 5.1 (4.5–5.9) | 5.8 (4.2–8.5) | ||||||||
| WBC, ×109/L | All | 1.6 | 43 | 5.7 (5.6–5.9) | 14.0 (12.7–15.4) | 17.3 (13.9–22.4) | 25 | 5.3 (5.0–5.5) | 11.1 (9.9–12.7) | 17.7 (13.5–24.9) | 10.8 (8.9–15.9) | 16.4 (15.0–23.7) |
| Male | 21 | 5.7 (5.5–6.0) | 14.7 (12.0–16.9) | 20.0 (14.8–29.6) | 8 | 5.4 (5.0–5.8) | 11.6 (9.5–14.8) | 19.5 (12.3–40.6) | ||||
| Female | 22 | 5.7 (5.5–5.9) | 13.3 (11.7–15.3) | 14.3 (10.4–21.2) | 17 | 5.2 (5.0–5.4) | 10.8 (9.4–12.7) | 16.8 (12.1–26.1) | ||||
| RBC, ×1012/L | All | 1.0 | 43 | 4.8 (4.8–4.9) | 3.5 (3.2–3.9) | 10.0 (8.2–12.8) | 25 | 4.6 (4.5–4.7) | 2.8 (2.5–3.2) | 8.7 (6.8–12.1) | 2.6 (1.4–4.0) | 6.5 (6.1–7.4) |
| Male | 21 | 5.2 (5.2–5.3) | 3.4 (3.0–3.9) | 5.9 (4.4–8.6) | 8 | 5.1 (5.0–5.1) | 1.9 (1.6–2.5) | 3.3 (2.0–6.8) | ||||
| Female | 22 | 4.4 (4.4–4.5) | 3.6 (3.2–4.2) | 5.6 (4.2–8.1) | 17 | 4.4 (4.3–4.4) | 3.2 (2.8–3.8) | 5.4 (3.9–8.4) | ||||
| Hb, g/L | All | 0.9 | 43 | 145 (143–148) | 3.9 (3.5–4.3) | 9.9 (8.1–12.6) | 25 | 138 (136–140) | 2.7 (2.4–3.1) | 10.5 (8.2–14.6) | 2.7 (1.7–3.6) | 5.9 (3.2–7.0) |
| Male | 21 | 158 (157–161) | 3.5 (3.1–4.0) | 5.1 (3.8–7.5) | 8 | 157 (155–160) | 1.8 (1.4–2.2) | 4.3 (2.8–9.0) | ||||
| Female | 22 | 133 (132–134) | 4.3 (3.8–5.0) | 3.4 (2.3–5.2) | 17 | 129 (128–131) | 3.2 (2.8–3.8) | 4.5 (3.2–7.0) | ||||
| HCT, L/L | All | 1.7 | 43 | 0.43 (0.43–0.44) | 3.5 (3.2–3.9) | 8.8 (7.2–11.2) | 25 | 0.41 (0.41–0.42) | 2.7 (2.4–3.1) | 8.2 (6.3–11.4) | 2.8 (2.2–2.8) | 5.5 (5.2–6.5) |
| Male | 21 | 0.47 (0.46–0.47) | 3.4 (3.0–3.9) | 4.8 (3.5–7.1) | 8 | 0.46 (0.45–0.46) | 1.9 (1.6–2.5) | 3.8 (2.4–7.9) | ||||
| Female | 22 | 0.40 (0.40–0.41) | 3.6 (3.2–4.2) | 3.9 (2.8–5.8) | 17 | 0.39 (0.39–0.40) | 3.2 (2.8–3.8) | 3.5 (2.4–5.7) | ||||
| MCV, fl | All | 1.0 | 43 | 90.0 (89.6–90.4) | 1.3 (1.2–1.5) | 3.1 (2.5–4.0) | 25 | 90.0 (89.4–90.6) | 0.6 (0.5–0.7) | 4.2 (3.2–5.8) | 0.8 (0.7–1.7) | 3.7 (3.4–4.7) |
| Male | 21 | 89.5 (88.8–90.1) | 1.2 (1.1–1.4) | 3.2 (2.4–4.6) | 8 | 89.9 (89.2–90.6) | 0.5 (0.4–0.7) | 2.2 (1.4–4.5) | ||||
| Female | 22 | 90.6 (90.0–91.1) | 1.4 (1.3–1.7) | 3.0 (2.2–4.3) | 17 | 90.1 (89.3–90.9) | 0.6 (0.5–0.7) | 4.8 (3.6–7.4) | ||||
| MCH, pg | All | 0.7 | 42 | 30.3 (30.1–30.4) | 1.9 (1.7–2.0) | 3.3 (2.7–4.2) | 25 | 30.1 (29.8–30.3) | 0.7 (0.6–0.8) | 5.3 (4.1–7.3) | 0.8 (0.3–1.6) | 4.4 (4.1–5.6) |
| Male | 21 | 30.4 (30.2–30.6) | 1.6 (1.4–1.9) | 3.1 (2.3–4.6) | 8 | 31.0 (30.8–31.2) | 0.7 (0.6–0.9) | 1.8 (1.1–3.7) | ||||
| Female | 21 | 30.1 (29.9–30.3) | 2.0 (1.8–2.4) | 3.4 (2.5–5.0) | 17 | 29.7 (29.3–30.1) | 0.7 (0.6–0.9) | 5.9 (4.4–8.9) | ||||
| MCHC, g/L | All | 1.3 | 43 | 336 (335–337) | 1.8 (1.6–2.0) | 1.8 (1.4–2.4) | 25 | 335 (333–336) | 1.0 (0.8–1.1) | 2.7 (2.1–3.9) | 1.0 (0.6–1.8) | 1.4 (0.9–2.3) |
| Male | 21 | 340 (338–342) | 1.7 (1.5–1.9) | 1.6 (1.0–2.4) | 8 | 345 (342–347) | 0.7 (0.6–0.9) | 1.9 (1.2–4.0) | ||||
| Female | 22 | 332 (330–333) | 1.9 (1.7–2.2) | 1.0 (0.4–1.7) | 17 | 330 (328–331) | 1.1 (0.9–1.3) | 1.7 (1.2–2.7) | ||||
| PLT, ×109/L | All | 2.4 | 43 | 223 (218–228) | 7.9 (7.2–8.7) | 18.8 (15.4–24.0) | 25 | 203 (197–208) | 7.3 (6.5–8.4) | 13.3 (10.2–18.6) | 5.6 (4.5–10) | 19.4 (16.4–27.3) |
| Male | 21 | 208 (200–216) | 8.3 (7.3–9.5) | 18.4 (13.9–26.8) | 8 | 197 (189–205) | 7.7 (6.3–9.8) | 11.6 (7.2–24.4) | ||||
| Female | 22 | 237 (229–246) | 7.6 (6.7–8.8) | 17.0 (12.9–24.5) | 17 | 205 (197–212) | 7.2 (6.3–8.5) | 13.8 (10.1–21.4) | ||||
-
ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; GGT, γ-glutamyl transferase; TP, total protein; ALB, albumin; TBIL, total bilirubin; TBA, total bile acids; Crea, creatinine; UA, uric acid; HDL-C, high density lipoproteins-cholesterol; LDL-C, low density lipoproteins-cholesterol; TC, total cholesterol; TG, triglycerides; Glu, glucose; WBC, white blood cell; RBC, red blood cell; Hb, hemoglobin; HCT, hematocrit; MCV, mean cellular volume; MCH, mean cellular hemoglobin; MCHC, mean cellular hemoglobin concentration; PLT, platelets. aData obtained from online European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) biological variation database [10]. The results in bold were recommended by the judgment of 95% CI overlap. N was the number of subjects used for biological variation estimation.
We compared our weekly CVI with EFLM data and found that all measurands showed a constant overlap of 95% CIs. The weekly CVI estimates for all measurands were significantly lower than the annual CVI estimates except for TP, TBIL, WBC, RBC (female), Hb (female), HCT (female) and PLT (Table 1). The CVI estimates of ALT, ALP, GGT, MCV, MCH and MCHC showed large difference (>50%) between weekly group and annual group (Figure 1). The CVG estimates for all measurands showed a constant overlap of 95% CIs between annual group and weekly group.
Percentage bias of CVI between annual and weekly group.
Indices derived from BV
RCV, II in the two groups were shown in Table 2. The annual RCVs were all higher than the weekly RCVs, where GGT and MCH showed large difference (>50%). We observed that for the annual group, only II of ALB was >1.4, and the II of ALP, GGT, Crea, HDL-C, UA, RBC, Hb, HCT, MCV, MCH, and PLT were <0.6, while the remaining were between 0.6 and 1.4. For the weekly group, the II values for all measurands were <1.0, with ALT, ALP, GGT, AST, Crea, TC, RBC, Hb, HCT, MCV, MCH, MCHC and PLT being <0.6.
The results of RCVs, II, TVset and n for the annual and weekly groups.
| Parameters | Annual group | Weekly group | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Group | RCV, % | II | TVset, % | n (D=10%) | RCV, % | II | TVset, % | n (D=10%) | |
| ALT, U/L | All | 93.7 | 0.94 | 71.5 | 44 | 47.7 | 0.41 | 36.4 | 12 |
| Male | 79.8 | 0.80 | 61.0 | 32 | 42.3 | 0.42 | 32.3 | 9 | |
| Female | 112.8 | 1.46 | 86.1 | 64 | 51.6 | 0.60 | 39.4 | 14 | |
| AST, U/L | All | 51.5 | 0.81 | 39.4 | 14 | 33.8 | 0.56 | 25.9 | 6 |
| Male | 45.3 | 0.76 | 34.6 | 11 | 29.3 | 0.39 | 22.4 | 5 | |
| Female | 57.9 | 0.87 | 44.2 | 17 | 37.1 | 1.06 | 28.3 | 7 | |
| ALP, U/L | All | 33.3 | 0.57 | 25.4 | 6 | 16.7 | 0.27 | 12.8 | 2 |
| Male | 25.3 | 0.45 | 19.3 | 4 | 14.9 | 0.34 | 11.4 | 2 | |
| Female | 40.6 | 0.69 | 31.0 | 9 | 17.5 | 0.32 | 13.4 | 2 | |
| GGT, U/L | All | 69.0 | 0.52 | 52.7 | 24 | 24.4 | 0.24 | 18.6 | 3 |
| Male | 68.8 | 0.60 | 52.5 | 24 | 25.2 | 0.42 | 19.2 | 4 | |
| Female | 61.6 | 0.65 | 47.1 | 19 | 21.3 | 0.26 | 16.3 | 3 | |
| TP, g/L | All | 9.5 | 0.94 | 7.3 | 1 | 8.8 | 0.61 | 6.7 | 1 |
| Male | 9.8 | 0.91 | 7.5 | 1 | 6.9 | 0.74 | 5.3 | 1 | |
| Female | 9.5 | 1.03 | 7.3 | 1 | 9.5 | 0.60 | 7.3 | 1 | |
| ALB, g/L | All | 12.6 | 1.47 | 9.6 | 1 | 8.9 | 0.67 | 6.8 | 1 |
| Male | 11.2 | 2.05 | 8.6 | 1 | 7.8 | 1.00 | 6.0 | 1 | |
| Female | 13.7 | 1.45 | 10.4 | 1 | 9.1 | 0.66 | 7.0 | 1 | |
| TBA, µmol/L | All | 151.2 | 1.00 | 115.5 | 115 | ||||
| Male | 141.2 | 0.80 | 107.9 | 100 | |||||
| Female | 162.2 | 1.50 | 123.9 | 132 | |||||
| TBIL, µmol/L | All | 57.9 | 0.73 | 44.2 | 17 | 55.3 | 0.76 | 42.2 | 16 |
| Male | 55.3 | 0.79 | 42.2 | 16 | 51.4 | 0.93 | 39.3 | 14 | |
| Female | 61.3 | 0.87 | 46.8 | 19 | 57.6 | 0.99 | 44.0 | 17 | |
| Urea, mmol/L | All | 49.6 | 0.89 | 37.9 | 13 | 35.2 | 0.60 | 26.9 | 7 |
| Male | 41.6 | 0.90 | 31.7 | 9 | 27.8 | 0.67 | 21.3 | 4 | |
| Female | 58.7 | 1.29 | 44.8 | 18 | 39.4 | 0.63 | 30.1 | 8 | |
| Crea, µmol/L | All | 17.9 | 0.32 | 13.7 | 2 | 12.5 | 0.23 | 9.6 | 1 |
| Male | 16.3 | 0.52 | 12.4 | 2 | 11.5 | 0.49 | 8.8 | 1 | |
| Female | 19.8 | 0.72 | 15.1 | 2 | 12.8 | 0.39 | 9.8 | 1 | |
| UA, µmol/L | All | 28.0 | 0.46 | 21.4 | 4 | ||||
| Male | 22.2 | 0.55 | 17.0 | 3 | |||||
| Female | 34.8 | 0.67 | 26.6 | 7 | |||||
| TG, mmol/L | All | 87.4 | 0.81 | 66.8 | 39 | ||||
| Male | 89.6 | 0.88 | 68.5 | 41 | |||||
| Female | 78.0 | 1.14 | 59.6 | 31 | |||||
| TC, mmol/L | All | 28.5 | 0.86 | 21.8 | 5 | 15.4 | 0.41 | 11.7 | 2 |
| Male | 25.7 | 0.71 | 19.7 | 4 | 14.6 | 0.50 | 11.1 | 2 | |
| Female | 31.0 | 1.02 | 23.7 | 5 | 15.6 | 0.38 | 12.0 | 2 | |
| HDL-C, mmol/L | All | 29.5 | 0.40 | 22.5 | 5 | ||||
| Male | 27.5 | 0.38 | 21.0 | 4 | |||||
| Female | 30.0 | 0.55 | 22.9 | 5 | |||||
| LDL-C, mmol/L | All | 34.4 | 0.65 | 26.3 | 6 | ||||
| Male | 33.3 | 0.57 | 25.5 | 6 | |||||
| Female | 35.8 | 0.82 | 27.3 | 7 | |||||
| GLU, mmol/L | All | 15.7 | 0.92 | 12.0 | 2 | ||||
| Male | 16.7 | 0.95 | 12.8 | 2 | |||||
| Female | 14.6 | 0.88 | 11.1 | 2 | |||||
| WBC, ×109/L | All | 39.1 | 0.81 | 29.8 | 8 | 31.1 | 0.63 | 23.7 | 5 |
| Male | 41.0 | 0.74 | 31.3 | 9 | 32.5 | 0.59 | 24.8 | 6 | |
| Female | 37.1 | 0.93 | 28.4 | 7 | 30.3 | 0.64 | 23.1 | 5 | |
| RBC, ×1012/L | All | 10.1 | 0.35 | 7.7 | 1 | 8.2 | 0.32 | 6.3 | 1 |
| Male | 9.8 | 0.58 | 7.5 | 1 | 6.0 | 0.58 | 4.5 | 1 | |
| Female | 10.4 | 0.64 | 7.9 | 1 | 9.3 | 0.59 | 7.1 | 1 | |
| Hb, g/L | All | 11.1 | 0.39 | 8.5 | 1 | 7.9 | 0.26 | 6.0 | 1 |
| Male | 10.0 | 0.69 | 7.7 | 1 | 5.6 | 0.42 | 4.3 | 1 | |
| Female | 12.2 | 1.26 | 9.3 | 1 | 9.2 | 0.71 | 7.0 | 1 | |
| HCT, L/L | All | 10.8 | 0.40 | 8.2 | 1 | 8.8 | 0.33 | 6.8 | 1 |
| Male | 10.5 | 0.71 | 8.0 | 1 | 7.1 | 0.50 | 5.4 | 1 | |
| Female | 11.0 | 0.92 | 8.4 | 1 | 10.0 | 0.91 | 7.7 | 1 | |
| MCV, fl | All | 4.5 | 0.42 | 3.5 | 1 | 3.2 | 0.14 | 2.5 | 1 |
| Male | 4.3 | 0.38 | 3.3 | 1 | 3.1 | 0.23 | 2.4 | 1 | |
| Female | 4.8 | 0.47 | 3.6 | 1 | 3.2 | 0.13 | 2.5 | 1 | |
| MCH, pg | All | 5.6 | 0.58 | 4.3 | 1 | 2.7 | 0.13 | 2.1 | 1 |
| Male | 4.8 | 0.52 | 3.7 | 1 | 2.7 | 0.39 | 2.1 | 1 | |
| Female | 5.9 | 0.59 | 4.5 | 1 | 2.7 | 0.12 | 2.1 | 1 | |
| MCHC, g/L | All | 6.2 | 1.00 | 4.7 | 1 | 4.5 | 0.37 | 3.5 | 1 |
| Male | 5.9 | 1.06 | 4.5 | 1 | 4.1 | 0.37 | 3.1 | 1 | |
| Female | 6.4 | 1.90 | 4.9 | 1 | 4.7 | 0.65 | 3.6 | 1 | |
| PLT, ×109/L | All | 22.9 | 0.42 | 17.5 | 3 | 21.3 | 0.55 | 16.3 | 3 |
| Male | 23.9 | 0.45 | 18.3 | 3 | 22.4 | 0.66 | 17.1 | 3 | |
| Female | 22.1 | 0.45 | 16.9 | 3 | 21.0 | 0.52 | 16.1 | 3 | |
-
RCV, reference change value, Z is 1.96 for a p-value<0.05. TVset, the total variation around the homeostatic set point for a particular individual. n, number of specimens required to determine the homeostatic set point (D=10%).
Personalized reference intervals
TVset of both groups were shown in Table 2 and we calculated RIper based on it. Take TP for example, the homeostatic set point for TP of a specific individual in the annual group is 70.0 g/L, with TVset being 7.3%, and thus RIper for TP was 70.0 ± (70.0 × 7.3%) g/L; that is, the lower limit of the RIper was 64.9 g/L and the upper limit was 75.1 g/L. We also calculated the number of tests that around 10% of the homeostatic set point for the individual in the two groups (Table 2). For most measurands (67% of the annual BV and 83% of the weekly BV), n was no more than 6.
Linear regression was performed to evaluate whether results of each year had changed during the six years. There were 10 measurands that had significant changes with age, of which the p value of ALB, TC, LDL-C, MCHC were <0.01, and the p value of ALT, TBA, TG, HCT, MCH, PLT were <0.05. TC and ALB were shown as examples in Figure 2A. This means that the changing with age for these measurands had an impact on the estimate of homeostatic set point. To examine whether age influenced the estimation of CVI and CVG, we performed trend correction based on regression equation. There is no significant difference between the previous and corrected BV data (95% CI overlap). The corrected data, the comparison of original and corrected data, and the correction method were shown in Supplementary Table 4. The remaining 14 measurands kept stable during this period, such as HDL-C and WBC (Figure 2B).
The trend changing with age of analytes during the six years.
(A) There was significant difference within six years, for example TC and ALB; (B), there was no significant difference within six years, for example HDL-C and WBC.
We compared RIper with RIpop using RRi. For gender-dependence of the reference intervals (ALT, ALP, GGT, urea, TBIL, AST, Crea, UA, RBC, Hb, and HCT), we took the male subgroup for comparison. Except for HDL-C, the RRis for the other measurands were <1.0, and the RRis for ALP, TBIL, Crea, TC, LDL-C, MCV, MCHC, and PLT were even <0.5 (Table 3). In order to compare RIper and RIpop more intuitively, we then normalized the RIpop of all measurands to [0, 1] interval. The relative distribution range of the RIper using mean value of annual group as the homeostatic set point was shown as an example (Figure 3). We observed that most measurands only accounted for part of the RIpop. If test results were only judged according to the RIpop, it may cause the deviation of results evaluation.
The range of RIper and the ratio of RIper to RIpop (RRi) for annual group.
| Parameters | Meana | RIper b | RIpop c | RRi |
|---|---|---|---|---|
| ALT, U/L | 20.6 | 8–33 | 9–50 | 0.61 |
| AST, U/L | 20.3 | 13–27 | 15–40 | 0.56 |
| ALP, U/L | 62.8 | 51–75 | 45–125 | 0.30 |
| GGT, U/L | 27.7 | 13–42 | 10–60 | 0.58 |
| TP, g/L | 70.2 | 65–75 | 65–85 | 0.50 |
| ALB, g/L | 44.7 | 40–49 | 40–55 | 0.60 |
| TBA, µmol/L | 2.6 | 0–6 | 0–10 | 0.60 |
| TBIL, µmol/L | 14.9 | 9–21 | 0–26 | 0.46 |
| Urea, mmol/L | 5.2 | 3.5–6.9 | 3.1–8.0 | 0.69 |
| Crea, µmol/L | 72.5 | 64–82 | 57–97 | 0.45 |
| UA, µmol/L | 331.0 | 275–387 | 208–428 | 0.51 |
| TG, mmol/L | 0.97 | 0.32–1.62 | 0–1.7 | 0.76 |
| TC, mmol/L | 4.7 | 3.68–5.72 | 0–5.17 | 0.39 |
| HDL-C, mmol/L | 1.50 | 1.16–1.84 | >1.03 | NA |
| LDL-C, mmol/L | 2.8 | 2.06–3.54 | 0–3.33 | 0.44 |
| GLU, mmol/L | 5.1 | 4.5–5.7 | 3.9–6.1 | 0.55 |
| WBC, ×109/L | 5.7 | 4.0–7.4 | 3.5–9.5 | 0.57 |
| RBC, ×1012/L | 5.2 | 4.8–5.6 | 4.3–5.8 | 0.53 |
| Hb, g/L | 158 | 145–170 | 130–175 | 0.56 |
| HCT, L/L | 0.47 | 0.43–0.50 | 0.40–0.50 | 0.70 |
| MCV, fl | 90.0 | 87–93 | 82–100 | 0.33 |
| MCH, pg | 30.3 | 29–32 | 27–34 | 0.43 |
| MCHC, g/L | 336 | 320–351 | 316–354 | 0.82 |
| PLT, ×109/L | 223 | 184–262 | 125–350 | 0.35 |
-
aMean of annual healthy individuals. For gender-dependence of the reference intervals, mean of male subgroup (bold measurands) was selected to list. bRIper of annual healthy individuals (taking mean value as the homeostatic set point). cRIpop data came from the National Guide to Clinical Laboratory Procedures (4th edition) [17]. For gender-dependence of the reference intervals, male subgroup (bold measurands) was selected to list.
The relative distribution range of the RIper at the mean level of annual healthy group after normalization of the RIpop.
Discussion
BV and the indices derived from BV can be used to interpret the utility of population-based reference intervals, estimate the significance changes of two serial results and define analytical performance specifications for each analyte [6]. This study executed strict entry criteria for the individuals to ensure the accuracy of BV data. Individuals were screened by physical, electrocardiogram, laboratory, and ultrasound examinations (liver, pancreas, renal). Through screening, the potential diseases which might affect laboratory results during six years were eliminated.
For clinical chemistry analytes, the CVI estimates of female were significantly higher than those of male for ALT, ALP, urea, and UA in annual group. The weekly CVI estimates for these biomarkers of female were also higher than those of male in this study without significant difference. Aarsand [18] found weekly CVI estimates of female were higher than those of male for urea and UA, which was consistent with our result. The change of hormone level during perimenopause could increase bone turnover which could lead to significant changes in ALP, and thus might be the possible reason for CVI gender difference [19]. For hematology analytes, similar to Buoro’s study [20], CVI estimates of RBC, Hb, and HCT of female were significantly higher than those of male in weekly BV. It was probably because of the effects of menstrual cycles, which could reflect a loss of red blood cells in female, and thus could lead to higher CVI of erythrocyte related parameters than male. The duration in this study (six years) had covered multiples of the turnover periods for erythrocytes (about four months) [21]. Therefore, we observed that the CVI estimates of male in annual group were significantly larger than those of weekly group and tended to be consistent with those of female.
Although most measurands cited in EFLM database were analyzed in duplicate to calculate CVA and our CVA was obtained by daily quality control in our laboratory, weekly CVI in this study were almost identical to the EFLM database. Our results showed that the annual CVI estimates for 11 measurands were significantly higher than the weekly CVI and the differences for ALT, ALP, GGT, MCV, MCH and MCHC were >50% (Figure 1). The reason for such difference could be different research duration. This inference was supported by Li et al. [11] who conducted a monthly CVI study on CBC, and found that the monthly CVI estimates of RBC, Hb and HCT were higher than the weekly CVI estimates. We compared annual, monthly and weekly BV of CBC, for RBC, Hb, and HCT, annual CVI was higher than monthly CVI, and monthly CVI was higher than weekly CVI [11].
The CVI estimates of WBC and PLT in our weekly group were very close to those of the weekly data of Buoro’s study [22, 23]. For CVI of WBC, annual (this study), monthly [11], weekly [22], and daily data [22] were 14.0, 11.6, 11.1, and 8.7, respectively, which showed a decreasing and not significant trend by 95% CIs overlap (except for daily vs. monthly and annual data). For CVI of PLT, annual (this study), monthly [11], weekly [23], and daily data [23] were 7.9, 6.7, 7.2, and 3.3, respectively. The daily CVI of PLT was significantly lower than that of the data from other duration. These aspects were probably attributable to the turnover time of WBC (11–15 days) [24] and PLT (7–10 days) [25]. The medium-term and long-term studies covered several multiples of the turnover periods for WBC and PLT while short duration study covered a time period shorter than the required for these turnover periods, which may cause the difference of BV data.
The II, as proposed by Harris [26], was used to assess the utility of RIpop. When II>1.4, RIpop was useful for test interpretations. When II<0.6, RIpop was not sensitive enough for disease monitoring and follow-up. For the annual group, we noted that except for ALB, the II of the other measurands was <1.4, with approximately 50% measurands being <0.6. For the weekly group, the II was <0.6 for about 78% measurands, suggesting that RIpop of most measurands were not suitable for test result interpretation. Except for TBIL and PLT, the II of the annual BV was higher than the weekly BV, suggesting the application of reference interval could be improved with increased study duration. RCV application could provide an appropriate interpretation strategy of monitoring of longitudinal time changes of test results [6, 27]. The weekly RCV was lower than the annual RCV, indicating that the weekly RCV was oversensitive to the judgment of a series changes in the annual test result.
RIper is suitable to evaluate whether a measurand result is normal for a particular individual based on his own homeostatic set point. Currently, only one study from Coskun [9] has described the RIper calculation method. According to Coskun, we calculated the RIper for each measurand based on annual TVset. Figure 3 showed comparisons between RIper and RIpop, and the width of RIper for most measurands only covered small part of RIpop. The RRi of 33% of the measurands was <50%, suggesting the normal fluctuation range of these measurands was obviously narrower than RIpop. If an individual’s homeostatic set point was leaning to one side of RIpop, the overlap between RIper and RIpop would be significantly reduced, which lead to the worsened utility of RIpop. In Coskun’s study [9], TVset was calculated based on weekly CVI estimates (such as Crea and WBC), which were smaller than our annual results. In this study, the RIper based on annual data would be valuable to individuals who keep annual physical examinations. It is worth noting that CVA and the accuracy may change significantly according to different methods, such as Creatinine [28] (Jaffe or Enzymatic method). Thus, the differences of detection methods should be considered when using RIper.
When D was 10% closeness to the homeostatic set point, n of 67% measurands was no more than six. If D was 20%, n of 88% measurands was no more than six. In principle, for the homeostatic set point, the greater the n, the more accurate it would be [6]. However, serum biomarker concentrations changed with age [29], e.g. serum LDL-C concentrations increased with age [30], while ALB levels decreased [19, 31]. The time span in our study was six years, and the levels of 10 measurands exhibited the trend changing with age. Therefore, taking the mean value of six consecutive results as homeostatic set points might bring bias to current RIper calculations. For these measurands, it does not mean that a greater n is equivalent to a better result. Coskun proposed that when n≥3, RIper was robust [9]. It may be more appropriate for the above measurands to use the most recent results (for example, n=3) to calculate RIper which is conducive to the accurate judgment of the individual’s current situation. In addition, we note that when D was 10%, the n of three measurands was far more than 6 (>100 for TBA, >40 for ALT and >30 for TG). This may lead to unreliable results of RIper for these measurands. We further evaluated the effect of age on BV through trend correction, and found that there was no significant difference between the previous and corrected data. Because of the correlation between these measurands and age was weak (r was small), we showed the BV and derived parameters based on original data.
Study limitations
The study period was six years. It was difficult to identify the participants with consistent stable health status in advance, which made it impossible to perform the study according to international recommendation BV studies [32, 33], i.e., analyzed in duplicate and in the same analytical run. To overcome these limitations, we strictly controlled study entry criteria for the enrolled population, and performed robust research proposal to ensure data accuracy and reliability. In addition, the status of instrument detection among 6 years and the trend of change with age may influence the calculation of BV and derived parameters.
Conclusions
For the first time, we investigated long-term CVI, II, RCV, and TVset of commonly used clinical chemistry and hematology measurands in healthy individuals. The annual CVI estimates were higher than weekly CVI estimates for all measurands with 11 measurands showing significant difference. Most measurands showed marked individuality which meant the applicability of RIpop was limited. RIper provided the “personalized” criterion for a particular individual, and annual RIper is very valuable for long-term health assessments and diseases detection.
Funding source: National Key R&D Program of China provided by Ministry of Science and Technology of the People's Republic of China http://dx.doi.org/10.13039/501100002855
Award Identifier / Grant number: 2019YFC0840701
Acknowledgments
We are grateful to all the staff members for taking part in this study. We thank the participants for their cooperation and sample contributions.
-
Research funding: This study was supported by National Key R&D Program of China (2019YFC0840701) provided by the Ministry of Science and Technology of the People’s Republic of China. http://dx.doi.org/10.13039/501100002855.
-
Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.
-
Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.
-
Informed consent: Not applicable.
-
Ethical approval: This study was approved by the ethics committee of the First Hospital of China Medical University.
References
1. Liuxin, W. Annual report on development of health management and health industry in China No. 2. Beijing: Social Sciences Academic Press; 2019.Search in Google Scholar
2. Liuxin, W. Report on development of health management and health industry in China – new subject and new format No. 1. Beijing: Social Sciences Academic Press; 2018.Search in Google Scholar
3. Jonker, N, Aslan, B, Boned, B, Marques-Garcia, F, Ricos, C, Alvarez, V, et al.. Critical appraisal and meta-analysis of biological variation estimates for kidney related analytes. Clin Chem Lab Med 2022;60:469–78. https://doi.org/10.1515/cclm-2020-1168.Search in Google Scholar
4. Gonzalez-Lao, E, Corte, Z, Simon, M, Ricos, C, Coskun, A, Braga, F, et al.. Systematic review of the biological variation data for diabetes related analytes. Clin Chim Acta 2019;488:61–7. https://doi.org/10.1016/j.cca.2018.10.031.Search in Google Scholar
5. Diaz-Garzon, J, Fernandez-Calle, P, Minchinela, J, Aarsand, AK, Bartlett, WA, Aslan, B, et al.. Biological variation data for lipid cardiovascular risk assessment biomarkers. A systematic review applying the biological variation data critical appraisal checklist (BIVAC). Clin Chim Acta 2019;495:467–75. https://doi.org/10.1016/j.cca.2019.05.013.Search in Google Scholar
6. Braga, F, Panteghini, M. Generation of data on within-subject biological variation in laboratory medicine: an update. Crit Rev Clin Lab Sci 2016;53:313–25. https://doi.org/10.3109/10408363.2016.1150252.Search in Google Scholar
7. Fraser, CG, Harris, EK. Generation and application of data on biological variation in clinical chemistry. Crit Rev Clin Lab Sci 1989;27:409–37. https://doi.org/10.3109/10408368909106595.Search in Google Scholar
8. Siest, G, Henny, J, Grasbeck, R, Wilding, P, Petitclerc, C, Queralto, JM, et al.. The theory of reference values: an unfinished symphony. Clin Chem Lab Med 2013;51:47–64. https://doi.org/10.1515/cclm-2012-0682.Search in Google Scholar
9. Coskun, A, Sandberg, S, Unsal, I, Cavusoglu, C, Serteser, M, Kilercik, M, et al.. Personalized reference intervals in laboratory medicine: a new model based on within-subject biological variation. Clin Chem 2021;67:374–84. https://doi.org/10.1093/clinchem/hvaa233.Search in Google Scholar
10. EFLM. Available from: https://biologicalvariation.eu/. 2020.Search in Google Scholar
11. Li, C, Peng, M, Wu, J, Du, Z, Lu, H, Zhou, W. Long-term biological variation estimates of 13 hematological parameters in healthy Chinese subjects. Clin Chem Lab Med 2020;58:1282–90. https://doi.org/10.1515/cclm-2019-1141.Search in Google Scholar
12. Alvarez, L, Ricos, C, Peris, P, GuaNabens, N, Monegal, A, Pons, F, et al.. Components of biological variation of biochemical markers of bone turnover in Paget’s bone disease. Bone 2000;26:571–6. https://doi.org/10.1016/s8756-3282(00)00279-9.Search in Google Scholar
13. Flegal, KM, Kit, BK, Orpana, H, Graubard, BI. Association of all-cause mortality with overweight and obesity using standard body mass index categories A systematic review and meta-analysis. JAMA-J Am Med Assoc 2013;309:71–82. https://doi.org/10.1001/jama.2012.113905.Search in Google Scholar PubMed PubMed Central
14. Wang, S, Mu, R, Zhang, X, Yun, K, Shang, H, Zhao, M. Biological variation in serum bone turnover markers. Ann Clin Biochem 2020;57:144–50. https://doi.org/10.1177/0004563219899119.Search in Google Scholar PubMed
15. Pineda-Tenor, D, Laserna-Mendieta, EJ, Timon-Zapata, J, Rodelgo-Jimenez, L, Ramos-Corral, R, Recio-Montealegre, A, et al.. Biological variation and reference change values of common clinical chemistry and haematologic laboratory analytes in the elderly population. Clin Chem Lab Med 2013;51:851–62. https://doi.org/10.1515/cclm-2012-0701.Search in Google Scholar PubMed
16. Burdick, RK, Graybill, FA. Confidence intervals on variance components. New York: Marcel Dekker Inc.; 1992.10.1201/9781482277142Search in Google Scholar
17. Shang, H, Wang, Y, Shen, Z. National guide to clinical laboratory procedures, 4th ed Beijing: People’s Medical Publishing House; 2015.Search in Google Scholar
18. Aarsand, AK, Diaz-Garzon, J, Fernandez-Calle, P, Guerra, E, Locatelli, M, Bartlett, WA, et al.. The EuBIVAS: within- and between-subject biological variation data for electrolytes, lipids, urea, uric acid, total protein, total bilirubin, direct bilirubin, and glucose. Clin Chem 2018;64:1380–93. https://doi.org/10.1373/clinchem.2018.288415.Search in Google Scholar PubMed
19. Mu, R, Chen, W, Pan, B, Wang, L, Hao, X, Huang, X, et al.. First definition of reference intervals of liver function tests in China: a large-population-based multi-center study about healthy adults. PLoS One 2013;8:e72916. https://doi.org/10.1371/journal.pone.0072916.Search in Google Scholar PubMed PubMed Central
20. Buoro, S, Carobene, A, Seghezzi, M, Manenti, B, Dominoni, P, Pacioni, A, et al.. Short- and medium-term biological variation estimates of red blood cell and reticulocyte parameters in healthy subjects. Clin Chem Lab Med 2018;56:954–63. https://doi.org/10.1515/cclm-2017-0902.Search in Google Scholar PubMed
21. Coskun, A, Braga, F, Carobene, A, Tejedor Ganduxe, X, Aarsand, AK, Fernandez-Calle, P, et al.. Systematic review and meta-analysis of within-subject and between-subject biological variation estimates of 20 haematological parameters. Clin Chem Lab Med 2019;58:25–32. https://doi.org/10.1515/cclm-2019-0658.Search in Google Scholar PubMed
22. Buoro, S, Carobene, A, Seghezzi, M, Manenti, B, Pacioni, A, Ceriotti, F, et al.. Short- and medium-term biological variation estimates of leukocytes extended to differential count and morphology-structural parameters (cell population data) in blood samples obtained from healthy people. Clin Chim Acta 2017;473:147–56. https://doi.org/10.1016/j.cca.2017.07.009.Search in Google Scholar PubMed
23. Buoro, S, Seghezzi, M, Manenti, B, Pacioni, A, Carobene, A, Ceriotti, F, et al.. Biological variation of platelet parameters determined by the Sysmex XN hematology analyzer. Clin Chim Acta 2017;470:125–32. https://doi.org/10.1016/j.cca.2017.05.004.Search in Google Scholar PubMed
24. Cerny, J, Rosmarin, AG. Why does my patient have leukocytosis? Hematol Oncol Clin N Am 2012;26:303–19. https://doi.org/10.1016/j.hoc.2012.01.001.Search in Google Scholar PubMed
25. Lu, SJ, Li, F, Yin, H, Feng, Q, Kimbrel, EA, Hahm, E, et al.. Platelets generated from human embryonic stem cells are functional in vitro and in the microcirculation of living mice. Cell Res 2011;21:530–45. https://doi.org/10.1038/cr.2011.8.Search in Google Scholar PubMed PubMed Central
26. Harris, EK. Effects of intra- and interindividual variation on the appropriate use of normal ranges. Clin Chem 1974;20:1535–42. https://doi.org/10.1093/clinchem/20.12.1535.Search in Google Scholar
27. Harris, EK, Yasaka, T. On the calculation of a “reference change” for comparing two consecutive measurements. Clin Chem 1983;29:25–30. https://doi.org/10.1093/clinchem/29.1.25.Search in Google Scholar
28. Carobene, A, Ceriotti, F, Infusino, I, Frusciante, E, Panteghini, M. Evaluation of the impact of standardization process on the quality of serum creatinine determination in Italian laboratories. Clin Chim Acta 2014;427:100–6. https://doi.org/10.1016/j.cca.2013.10.001.Search in Google Scholar PubMed
29. Ma, C, Xia, L, Chen, X, Wu, J, Yin, Y, Hou, L, et al.. Establishment of variation source and age-related reference interval models for 22 common biochemical analytes in older people using real-world big data mining. Age Ageing 2020;49:1062–70. https://doi.org/10.1093/ageing/afaa096.Search in Google Scholar PubMed
30. Rustad, P, Felding, P, Franzson, L, Kairisto, V, Lahti, A, Martensson, A, et al.. The Nordic Reference Interval Project 2000: recommended reference intervals for 25 common biochemical properties. Scand J Clin Lab Invest 2004;64:271–84. https://doi.org/10.1080/00365510410006324.Search in Google Scholar PubMed
31. Veering, BT, Burm, AG, Souverijn, JH, Serree, JM, Spierdijk, J. The effect of age on serum concentrations of albumin and alpha 1-acid glycoprotein. Br J Clin Pharmacol 1990;29:201–6. https://doi.org/10.1111/j.1365-2125.1990.tb03620.x.Search in Google Scholar PubMed PubMed Central
32. Bartlett, WA, Braga, F, Carobene, A, Coskun, A, Prusa, R, Fernandez-Calle, P, et al.. A checklist for critical appraisal of studies of biological variation. Clin Chem Lab Med 2015;53:879–85. https://doi.org/10.1515/cclm-2014-1127.Search in Google Scholar PubMed
33. Aarsand, AK, Roraas, T, Fernandez-Calle, P, Ricos, C, Diaz-Garzon, J, Jonker, N, et al.. The biological variation data critical appraisal checklist: a standard for evaluating studies on biological variation. Clin Chem 2018;64:501–14. https://doi.org/10.1373/clinchem.2017.281808.Search in Google Scholar PubMed
Supplementary Material
The online version of this article offers supplementary material (https://doi.org/10.1515/cclm-2021-0479).
© 2021 Walter de Gruyter GmbH, Berlin/Boston
