Abstract
Serum insulin-like growth factor 1 (IGF-1) and insulin-like growth factor binding protein 3 (IGFBP-3) levels are regulated by growth hormone, nutritional status, and other endocrine factors, and are implicated in aging, metabolic disorders, and malignancies. Establishing appropriate reference ranges for these biomarkers is essential for safety monitoring and pharmacodynamic assessments in clinical trials. Although various detection methods—including enzyme-linked immunosorbent assay, chemiluminescence immunoassay, radioimmunoassay, and immunoradiometric assay (IRMA)—are available, defining reliable reference intervals remains challenging due to variability linked to age, sex, season, and assay methodology. IRMA manufacturers typically provide generalized reference values and recommend that laboratories establish context-specific ranges; however, data from healthy adult populations are limited. We analyzed data from 255 self-reported healthy Korean men aged 19–40 years using the IRMA method and R software. Age-specific reference ranges and standardized centile curves for serum IGF-1 and IGFBP-3 were established. Both biomarkers demonstrated significant seasonal variation, with notable differences observed among spring, autumn, and winter groups. This study presents a method for establishing institution-specific reference ranges and highlights the importance of considering seasonal variations to help reduce unnecessary dropouts during clinical trials.
Keywords: IGF-1, IGFBP-3, Seasonal Variation, Reference Ranges, Koreans
INTRODUCTION
Insulin like growth factor 1 (IGF-1) is a polypeptide hormone structurally similar to insulin. By binding to the IGF-1 receptor, it activates the mitogen-activated protein kinase and phosphoinositide 3-kinase signaling pathways in target tissues. IGF-1 is produced primarily by hepatocytes in response to growth hormone (GH) stimulation but is expressed in nearly all tissues, where it plays critical roles in cell proliferation, differentiation, and survival. Circulating GH/IGF-1 levels decline markedly with age, suggesting that reduced IGF-1 activity contributes to age-related physiological changes [1].
Insulin-like growth binding protein 3 (IGFBP-3) is the most abundant IGF-binding protein in circulation, with high affinity for both IGF-I and IGF-II. By binding IGFs more tightly than the IGF receptor, IGFBP-3 stabilizes circulating IGF levels. As a multifunctional protein, it prolongs the half-life of IGF-I, limits receptor engagement, inhibits cell proliferation, and promotes apoptosis. IGFBP-3 also forms complexes with receptors or cytokines in the cytoplasm and nucleus. Its biological activity is regulated by GH, nutrition, and age, rendering it a focus of intensive research [2]. Recent studies suggest that IGF-1 and IGFBP-3 display distinct expression patterns and play pivotal roles in the development and progression of various cancers [3].
Measuring the serum levels of IGF-1 and IGFBP-3 is essential for diagnosing and monitoring diverse clinical conditions, including growth disorders, metabolic diseases, and certain cancers. In biopharmaceutics, these biomarkers are equally important for pharmacodynamic assessments and safety evaluations. However, variations arising from season, age, and sex, as well as inter-assay variability, make the determination of precise reference ranges a challenging [4,5,6]. Techniques for measuring IGF-1 and IGFBP-3 include immunoassays and mass-spectrometry-based methods. At our center, we employ an immunoradiometric assay (IRMA) to measure IGF-1 and IGFBP-3; this approach offers simplicity, sensitivity, and high throughput. Nevertheless, IRMA shows poor agreement across immunoassay platforms and notable inter-assay variability, underscoring the need for each laboratory to establish its own reference range [4,7,8]. The lack of reference-range studies for healthy adults can lead to difficulties in data interpretation, increased dropout rates, or misinterpretation of results in clinical trials. Here, we present standardized centile curves and reference ranges for IGF-1 and IGFBP-3 derived from data on self-declared healthy Korean male adults aged 19–40 years, analyzed using R version 4.4.3 (R Foundation for Statistical Computing, Vienna, Austria), with the goal of assisting laboratories in establishing appropriate reference ranges. We also examine seasonal variation in serum IGF-1 and IGFBP-3 levels among the participants.
METHODS
Study subjects
Data on serum IGF-1, IGFBP-3, and age were collected from 255 self-identified healthy male adults aged 19–40 years who participated in the screening phase of three clinical trials conducted at a single center between May 2017 and April 2025. Daily minimum, maximum, and mean temperatures corresponding to each participant’s blood sampling date were obtained from the Korea Meteorological Administration website (https://www.weather.go.kr/w/observation/land/past-obs/obs-by-day.do?stn=108&yy=2025&mm=4&obs=1).
Exclusion criteria comprised a medical history of liver disease, renal disease, diabetes mellitus, pituitary or thyroid disorders, malignancies, or any other major conditions affecting the cardiovascular, respiratory, or gastrointestinal systems. Additional exclusions applied to individuals with a body mass index > 28 kg/m2; those who had taken prescription medications, traditional herbal medicines, or dietary supplements within 14 days prior to the screening visit; and those who had used over-the-counter medications within 7 days prior to screening. Participants reporting heavy smoking (> 10 cigarettes/day) or heavy alcohol consumption (> 8 units/week) within 30 days prior to screening were also excluded.
This study was approved by the Institutional Review Board (IRB) of Asan Medical Center (IRB No. 2025-0428).
Measurements
Blood samples were collected in the morning after an overnight fast. Serum IGF-1 and IGFBP-3 were measured using IRMAs (IGF-1: IRMA IGF-1 kit, Beckman Coulter Inc., Brea, CA, USA; IGFBP-3: RIAKEY IGFBP-3 IRMA Tube II kit; SHINJIN MEDICS Inc., Goyang, Korea). Samples were stored at 2–8°C, or frozen at −20°C for storage exceeding 24 hours.
Both assays use a 2-site “sandwich” format with one monoclonal antibody immobilized on the tube and a second 125I-labeled antibody serving as tracer. For IGF-I, 50 µL of sample was incubated with dissociation buffer at 37°C for 30 minutes, then with 350 µL tracer at room temperature (18–25°C) for 2 hours with agitation. For IGFBP-3, 50 µL of sample was incubated with 100 µL tracer in coated tubes for 180 minutes at room temperature (15–25°C) on a shaker. Bound radioactivity was measured with a gamma counter, and concentrations were calculated from 6-point standard curves. The IGF-I assay demonstrated a sensitivity was < 6.25 ng/mL (dynamic range up to 1,500 ng/mL), with intra-and inter-assay CVs ≤ 8.3%. IGFBP-3 showed comparable precision.
Statistical analysis
Standardized centile curves and reference ranges for serum IGF-1 and IGFBP-3 were generated using the gamlss (Generalized Additive Models for Location, Scale, and Shape) R package [9]. Specifically, the gamlss::lms function was used to fit Lambda–Mu–Sigma (LMS) curves for centile estimation, and the gamlss::centiles function was used to plot centile curves for the GAMLSS objects.
One-way analysis of variance (ANOVA) was applied to compare IGF-1, IGFBP-3, and the IGF-1/IGFBP-3 molar ratio across seasonal groups, with Bonferroni correction for post hoc pairwise comparisons. To examine the relationship between ambient temperature on the day of sampling and hormone levels, daily minimum, maximum, and mean temperatures were obtained from the Korea Meteorological Administration and matched to each participant’s collection date. Pearson correlation analysis was performed to assess associations between temperature variables and hormone concentrations.
All analyses were conducted in R version 4.4.3 (R Foundation for Statistical Computing), as detailed in the Results section, and p-values < 0.05 were considered statistically significant.
RESULTS
Table 1 summarizes the number of subjects in each age group. The standardized centile curve for IGF-1 showed a progressive decline with age in healthy male adults aged 19–40 years (Fig. 1). In contrast, the standardized centile curve for IGFBP-3 (Fig. 2) did not display a clear age-related trend within the same range. Accordingly, age-specific reference ranges for IGF-1 were calculated at 1-, 2-, 5-, and 10-year intervals, using the 2.5th, 5th, 10th, 25th, 75th, 90th, 95th, and 97.5th percentiles (Table 2). Reference ranges for IGFBP-3 were derived in the same manner (Table 3). However, because no significant age-related differences emerged for IGFBP-3 between 19 and 40 years, a single, age-independent reference range was deemed sufficient. The 95% reference interval for IGFBP-3 ranged from 1,199.75 (2.5th percentile) to 3,412.6 (97.5th percentile).
Table 1. Number of subjects by 2-year age interval (n = 255).
| 2-yr age interval | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| [19,22) | [22,24) | [24,26) | [26,28) | [28,30) | [30,32) | [32,34) | [34,36) | [36,38) | [38,41) | |
| No. of subjects | 28 | 33 | 46 | 48 | 34 | 26 | 20 | 7 | 9 | 4 |
The symbol “[“ (left square bracket) indicates inclusion of the endpoint, whereas “)” (right parenthesis) indicates exclusion of the endpoint.
Figure 1. Standardized centile curve for serum IGF-1 levels in healthy male adults aged 19–40 years. The curve demonstrates a gradual decline in IGF-1 levels with increasing age, calculated using the LMS method.
IGF-1, insulin-like growth factor 1; LMS, Lambda–Mu–Sigma.
Figure 2. Standardized centile curve for serum IGFBP-3 levels in healthy male adults aged 19–40 years. Unlike IGF-1 levels, IGFBP-3 levels do not show a clear age-related trend within this age range.
IGFBP-3, insulin-like growth factor binding protein 3; IGF-1, insulin-like growth factor 1.
Table 2. Reference ranges for IGF-1(ng/mL) by 1-, 2-, 5- and 10-year age intervals.
| Percentile | 2.5th | 5th | 10th | 25th | 50th | 75th | 90th | 95th | 97.5th | |
|---|---|---|---|---|---|---|---|---|---|---|
| Age interval: 1 yr | ||||||||||
| [19,20) | 101.4 | 135.7 | 173 | 225.8 | 268.2 | 311.9 | 370 | 414.5 | 459.4 | |
| [20,21) | 101 | 130.3 | 162.9 | 211.7 | 254.8 | 299.9 | 356.7 | 398.8 | 440.7 | |
| [21,22) | 102.2 | 126.8 | 154.9 | 199.5 | 242.7 | 289 | 344.2 | 384.2 | 423.4 | |
| [22,23) | 103.2 | 123.8 | 148.2 | 188.9 | 231.8 | 278.8 | 332.7 | 370.7 | 407.5 | |
| [23,24) | 103.6 | 121.1 | 142.2 | 179.4 | 221.8 | 269.4 | 321.9 | 358.2 | 392.8 | |
| [24,25) | 103.5 | 118.4 | 136.9 | 170.9 | 212.6 | 260.6 | 311.9 | 346.6 | 379.3 | |
| [25,26) | 102.8 | 115.7 | 132.1 | 163.3 | 204.2 | 252.4 | 302.5 | 335.8 | 366.9 | |
| [26,27) | 101.9 | 113.1 | 127.6 | 156.3 | 196.4 | 244.7 | 293.8 | 325.8 | 355.4 | |
| [27,28) | 100.7 | 110.6 | 123.6 | 150.1 | 189.1 | 237.4 | 285.6 | 316.6 | 344.8 | |
| [28,29) | 99.4 | 108.1 | 119.8 | 144.3 | 182.4 | 230.6 | 278 | 308 | 335.1 | |
| [29,30) | 98 | 105.7 | 116.2 | 139.1 | 176.1 | 224.2 | 270.9 | 300 | 326.1 | |
| [30,31) | 96.4 | 103.4 | 112.9 | 134.2 | 170.3 | 218.1 | 264.2 | 292.6 | 317.9 | |
| [31,32) | 94.9 | 101.1 | 109.8 | 129.7 | 164.8 | 212.3 | 258 | 285.8 | 310.3 | |
| [32,33) | 93.3 | 98.9 | 106.8 | 125.6 | 159.7 | 206.9 | 252.2 | 279.4 | 303.3 | |
| [33,34) | 91.6 | 96.8 | 104.1 | 121.7 | 154.9 | 201.7 | 246.7 | 273.6 | 296.8 | |
| [34,35) | 90 | 94.7 | 101.4 | 118.1 | 150.3 | 196.8 | 241.6 | 268.1 | 290.9 | |
| [35,36) | 88.4 | 92.7 | 98.9 | 114.7 | 146 | 192.2 | 236.8 | 263.1 | 285.5 | |
| [36,37) | 86.8 | 90.7 | 96.5 | 111.5 | 142 | 187.7 | 232.4 | 258.4 | 280.6 | |
| [37,38) | 85.3 | 88.9 | 94.3 | 108.5 | 138.1 | 183.5 | 228.2 | 254.1 | 276.1 | |
| [38,39) | 83.7 | 87.1 | 92.1 | 105.7 | 134.5 | 179.5 | 224.2 | 250.2 | 272.1 | |
| [39,40) | 82.2 | 85.3 | 90.1 | 103 | 131 | 175.6 | 220.6 | 246.6 | 268.4 | |
| [40,41) | 80.7 | 83.6 | 88 | 100.4 | 127.6 | 171.8 | 217 | 243.1 | 265 | |
| Age interval: 2 yr | ||||||||||
| [19,22) | 102 | 131.1 | 163.5 | 212.1 | 255.1 | 300.3 | 357.6 | 400.3 | 442.9 | |
| [22,24) | 105.1 | 123.6 | 145.8 | 184.3 | 226.9 | 274.5 | 328.4 | 366.4 | 403.1 | |
| [24,26) | 104.6 | 118.1 | 135.1 | 167.2 | 208.4 | 256.9 | 308.5 | 343.4 | 376.3 | |
| [26,28) | 102.5 | 112.7 | 126.1 | 153.3 | 192.7 | 241.4 | 291 | 323.4 | 353.2 | |
| [28,30) | 99.5 | 107.5 | 118.3 | 141.7 | 179.2 | 227.6 | 275.5 | 305.7 | 333 | |
| [30,32) | 96 | 102.5 | 111.5 | 131.9 | 167.4 | 215.3 | 261.6 | 290 | 315.1 | |
| [32,34) | 92.4 | 97.7 | 105.3 | 123.4 | 157.1 | 204.1 | 249.1 | 275.9 | 299.1 | |
| [34,36) | 88.8 | 93.3 | 99.8 | 116.1 | 148 | 194 | 237.9 | 263.4 | 285.1 | |
| [36,38) | 85.4 | 89.2 | 94.9 | 109.6 | 139.8 | 184.9 | 227.9 | 252.3 | 272.8 | |
| [38,41) | 81.3 | 84.5 | 89.3 | 102.5 | 130.8 | 174.5 | 216.8 | 240.3 | 259.5 | |
| Age interval: 5 yr | ||||||||||
| [19,25) | 96.6 | 118.1 | 143.8 | 187.3 | 233.9 | 285.3 | 343.8 | 385.1 | 424.8 | |
| [25,30) | 104.9 | 114.9 | 127.8 | 153.5 | 190 | 234.3 | 278.6 | 307.3 | 333.5 | |
| [30,35) | 94.2 | 100.6 | 109.3 | 128.9 | 162.7 | 208.1 | 252.4 | 279.9 | 304.5 | |
| [35,41) | 83.4 | 87.6 | 93.8 | 109.4 | 140.6 | 188.2 | 236.7 | 266.7 | 293.3 | |
| Age interval: 10 yr | ||||||||||
| [19,30) | 99 | 114.2 | 133 | 166.8 | 206.8 | 253.6 | 308.5 | 348.3 | 387.6 | |
| [30,41) | 89.1 | 94.6 | 102.5 | 121.3 | 156.5 | 206.5 | 255.2 | 284.7 | 310.7 | |
The symbol “[“ (left square bracket) indicates inclusion of the endpoint, whereas “)” (right parenthesis) indicates exclusion of the endpoint.
IGF-1, insulin-like growth factor 1.
Table 3. Reference ranges for IGFBP-3 (ng/mL) by 1-, 2-, 5- and 10-year age intervals.
| Percentile | 2.5th | 5th | 10th | 25th | 50th | 75th | 90th | 95th | 97.5th | |
|---|---|---|---|---|---|---|---|---|---|---|
| Age interval: 1 yr | ||||||||||
| [19,20) | 1,334.5 | 1,396.2 | 1,477.2 | 1,643.3 | 1,896.2 | 2,276.9 | 2,844.1 | 3,397.0 | 4,141.2 | |
| [20,21) | 1,259.0 | 1,319.3 | 1,398.4 | 1,560.4 | 1,805.5 | 2,170.8 | 2,704.8 | 3,213.7 | 3,883.4 | |
| [21,22) | 1,200.4 | 1,260.0 | 1,338.1 | 1,497.5 | 1,737.5 | 2,091.2 | 2,598.4 | 3,070.3 | 3,675.7 | |
| [22,23) | 1,168.8 | 1,228.9 | 1,307.5 | 1,467.8 | 1,707.6 | 2,057.1 | 2,548.6 | 2,994.9 | 3,552.4 | |
| [23,24) | 1,162.3 | 1,224.2 | 1,305.1 | 1,469.5 | 1,714.2 | 2,067.0 | 2,553.6 | 2,985.0 | 3,510.0 | |
| [24,25) | 1,166.8 | 1,231.2 | 1,315.2 | 1,485.6 | 1,737.7 | 2,097.3 | 2,584.3 | 3,006.3 | 3,507.1 | |
| [25,26) | 1,170.8 | 1,237.7 | 1,325.0 | 1,501.4 | 1,761.2 | 2,127.9 | 2,615.8 | 3,029.8 | 3,509.8 | |
| [26,27) | 1,165.9 | 1,235.1 | 1,325.0 | 1,506.6 | 1,772.5 | 2,144.1 | 2,630.5 | 3,035.2 | 3,494.8 | |
| [27,28) | 1,150.9 | 1,221.7 | 1,313.7 | 1,499.1 | 1,769.2 | 2,143.2 | 2,625.4 | 3,019.5 | 3,458.8 | |
| [28,29) | 1,129.9 | 1,202.1 | 1,295.9 | 1,484.4 | 1,757.8 | 2,133.1 | 2,610.1 | 2,993.9 | 3,414.7 | |
| [29,30) | 1,108.8 | 1,182.5 | 1,278.2 | 1,470.1 | 1,747.3 | 2,124.6 | 2,598.1 | 2,973.4 | 3,379.0 | |
| [30,31) | 1,093.0 | 1,168.6 | 1,266.7 | 1,463.3 | 1,745.9 | 2,127.7 | 2,601.1 | 2,971.2 | 3,366.1 | |
| [31,32) | 1,085.2 | 1,163.4 | 1,264.9 | 1,467.9 | 1,758.5 | 2,148.2 | 2,625.7 | 2,994.5 | 3,383.1 | |
| [32,33) | 1,085.0 | 1,166.6 | 1,272.4 | 1,483.6 | 1,784.8 | 2,185.6 | 2,671.3 | 3,041.8 | 3,428.0 | |
| [33,34) | 1,090.4 | 1,176.0 | 1,286.8 | 1,507.7 | 1,821.5 | 2,235.9 | 2,732.5 | 3,107.1 | 3,493.4 | |
| [34,35) | 1,099.1 | 1,189.2 | 1,305.7 | 1,537.6 | 1,865.5 | 2,295.3 | 2,804.7 | 3,184.7 | 3,572.8 | |
| [35,36) | 1,109.7 | 1,204.6 | 1,327.4 | 1,571.2 | 1,914.4 | 2,360.9 | 2,884.4 | 3,270.7 | 3,661.5 | |
| [36,37) | 1,120.7 | 1,220.8 | 1,350.2 | 1,606.6 | 1,966.0 | 2,429.9 | 2,968.1 | 3,361.2 | 3,755.3 | |
| [37,38) | 1,131.1 | 1,236.7 | 1,373.1 | 1,642.8 | 2,018.9 | 2,500.7 | 3,053.8 | 3,453.8 | 3,851.4 | |
| [38,39) | 1,140.4 | 1,251.7 | 1,395.4 | 1,678.9 | 2,072.2 | 2,572.1 | 3,140.1 | 3,546.9 | 3,948.0 | |
| [39,40) | 1,148.3 | 1,265.6 | 1,416.9 | 1,714.5 | 2,125.4 | 2,643.5 | 3,226.3 | 3,639.7 | 4,044.2 | |
| [40,41) | 1,154.5 | 1,278.0 | 1,437.2 | 1,749.6 | 2,178.6 | 2,715.0 | 3,312.2 | 3,731.9 | 4,139.6 | |
| Age interval: 2 yr | ||||||||||
| [19,22) | 1,207.9 | 1,269.1 | 1,349.4 | 1,513.9 | 1,763.0 | 2,132.9 | 2,669.3 | 3,174.6 | 3,830.8 | |
| [22,24) | 1,185.6 | 1,249.1 | 1,332.1 | 1,501.1 | 1,753.6 | 2,119.4 | 2,627.9 | 3,082.7 | 3,641.1 | |
| [24,26) | 1,168.8 | 1,234.7 | 1,320.5 | 1,494.3 | 1,750.9 | 2,115.1 | 2,604.0 | 3,023.2 | 3,514.7 | |
| [26,28) | 1,151.3 | 1,220.1 | 1,309.5 | 1,489.8 | 1,753.0 | 2,119.0 | 2,594.3 | 2,986.5 | 3,427.8 | |
| [28,30) | 1,129.8 | 1,202.6 | 1,297.0 | 1,486.5 | 1,760.5 | 2,134.7 | 2,606.9 | 2,983.8 | 3,394.1 | |
| [30,32) | 1,105.1 | 1,183.3 | 1,284.8 | 1,487.8 | 1,778.9 | 2,170.4 | 2,652.2 | 3,026.1 | 3,422.1 | |
| [32,34) | 1,080.5 | 1,166.1 | 1,277.2 | 1,498.7 | 1,813.9 | 2,231.1 | 2,732.0 | 3,110.4 | 3,501.2 | |
| [34,36) | 1,055.9 | 1,151.1 | 1,274.4 | 1,519.6 | 1,864.9 | 2,313.2 | 2,836.8 | 3,221.3 | 3,608.6 | |
| [36,38) | 1,026.7 | 1,134.0 | 1,272.8 | 1,547.2 | 1,928.4 | 2,412.2 | 2,960.0 | 3,350.7 | 3,734.5 | |
| [38,41) | 972.6 | 1,099.8 | 1,263.8 | 1,584.7 | 2,020.6 | 2,555.1 | 3,135.3 | 3,533.7 | 3,913.6 | |
| Age interval: 5 yr | ||||||||||
| [19,25) | 1,184.3 | 1,243.7 | 1,321.2 | 1,478.8 | 1,713.9 | 2,055.1 | 2,532.5 | 2,963.9 | 3,500.6 | |
| [25,30) | 1,152.6 | 1,223.2 | 1,315.3 | 1,501.7 | 1,775.5 | 2,159.6 | 2,664.0 | 3,084.7 | 3,563.0 | |
| [30,35) | 1,098.3 | 1,184.0 | 1,295.4 | 1,518.7 | 1,839.1 | 2,268.3 | 2,791.7 | 3,193.2 | 3,613.4 | |
| [35,41) | 986.9 | 1,099.1 | 1,243.9 | 1,528.1 | 1,917.9 | 2,403.8 | 2,942.0 | 3,318.2 | 3,682.1 | |
| Age interval: 10 yr | ||||||||||
| [19,30) | 1,211.0 | 1,279.1 | 1,359.6 | 1,507.6 | 1,719.1 | 2,051.3 | 2,589.7 | 3,121.4 | 3,789.9 | |
| [30,41) | 1,025.5 | 1,120.6 | 1,242.5 | 1,482.9 | 1,822.3 | 2,273.5 | 2,826.3 | 3,255.6 | 3,711.6 | |
The symbol “[“ (left square bracket) indicates inclusion of the endpoint, whereas “)” (right parenthesis) indicates exclusion of the endpoint.
IGFBP-3, insulin-like growth factor binding protein 3.
For IGF-1, age-specific reference ranges remain appropriate and can be generated with the gamlss::lms and gamlss::centiles functions, facilitating centile-curve exploration of laboratory data. In contrast, given the absence of a meaningful age effect, a single reference range for IGFBP-3 is preferable for individuals aged 19–40 years in our center.
During the study period, the screening-failure rate attributable to IGFBP-3 concentrations was higher in winter (December to February) than in the other seasons. Therefore, participants were classified by the season in which their samples were obtained—spring (March to May), summer (June to August), autumn (September to November), and winter (December to February)—based on the standard Korean definition. A one-way ANOVA compared IGF-1, IGFBP-3, and the IGF-1/IGFBP-3 molar ratio across seasons. Ultimately, subjects were allocated to spring (n = 90), autumn (n = 47), and winter (n = 118); no samples were collected in summer.
Age distribution did not differ significantly among the three seasonal groups (p = 0.326; Table 4). In contrast, IGF-1, IGFBP-3, and the IGF-1/IGFBP-3 molar ratio varied significantly by season (all p < 0.001; Fig. 3). Bonferroni post hoc analysis showed that IGF-1 concentrations differed between spring and winter (p < 0.001) and between autumn and winter (p < 0.001), whereas spring and autumn did not differ (p = 0.636). For IGFBP-3, all pairwise comparisons—spring vs. winter, autumn vs. winter, and spring vs. autumn—were significant (p < 0.001; Tables 5 and 6).
Table 4. Comparison of age across seasonal groups.
| Seasonal group comparison of age (p = 0.326) | ||||
|---|---|---|---|---|
| Spring | Autumn | Winter | Combined | |
| No. of subjects | 90 | 47 | 118 | 255 |
| Mean ± SD | 26.74 ± 4.52 | 27.74 ± 5.23 | 26.61 ± 4.13 | 26.86 ± 4.49 |
SD, standard deviation.
Figure 3. Comparison of serum IGF-1, IGFBP-3, and IGF-1/IGFBP-3 molar ratio across seasonal groups. Statistically significant differences were observed among the spring, autumn, and winter groups for all 3 parameters (p < 0.001 for all comparisons).
IGF-1, insulin-like growth factor 1; IGFBP-3, insulin-like growth factor binding protein 3.
Table 5. Parameters of serum IGF-1, IGFBP-3, and the IGF-1/IGFBP-3 ratio among seasonal groups.
| Parameter (ng/mL) | Spring (n = 90) | Autumn (n = 47) | Winter (n = 118) | Combined (n = 225) |
|---|---|---|---|---|
| IGF-1 | 179.15 ± 59.38 | 168.46 ± 49.60 | 237.38 ± 74.78 | 204.12 ± 72.29 |
| IGFBP-3 | 1,863.28 ± 447.41 | 2,901.04 ± 499.61 | 1,562.50 ± 205.38 | 1,915.36 ± 611.00 |
| IGF-1/IGFBP-3 molar ratio | 0.09 ± 0.03 | 0.05 ± 0.01 | 0.15 ± 0.04 | 0.11 ± 0.05 |
Values are presented as mean ± standard deviation.
IGF-1, insulin-like growth factor 1; IGFBP-3, insulin-like growth factor binding protein 3.
Table 6. Statistical comparison of serum IGF-1, IGFBP-3, and the IGF-1/IGFBP-3 ratio among seasonal groups.
| Parameter | Bonferroni-adjusted p-values | |||
|---|---|---|---|---|
| Spring | Autumn | Winter | ||
| IGF-1 | ||||
| Spring | - | > 0.99 | < 0.001 | |
| Autumn | > 0.99 | - | < 0.001 | |
| Winter | < 0.001 | < 0.001 | - | |
| IGFBP-3 | ||||
| Spring | - | < 0.001 | < 0.001 | |
| Autumn | < 0.001 | - | < 0.001 | |
| Winter | < 0.001 | < 0.001 | - | |
| IGF-1/IGFBP-3 molar ratio | ||||
| Spring | - | < 0.001 | < 0.001 | |
| Autumn | < 0.001 | - | < 0.001 | |
| Winter | < 0.001 | < 0.001 | - | |
IGF-1, insulin-like growth factor 1; IGFBP-3, insulin-like growth factor binding protein 3.
Although data were unavailable for the summer (June to August), seasonal values for this period could not be extrapolated from the time-series plot (Fig. 4). Nevertheless, visual inspection of mean values across spring and autumn suggested that IGF-1 and the IGF-1/IGFBP-3 molar ratio rose in winter relative to spring and autumn, whereas IGFBP-3 increased in autumn and declined in winter.
Figure 4. Seasonal variation in mean IGF-1, IGFBP-3, and IGF-1/IGFBP-3 molar ratio. The mean IGF-1 and IGF-1/IGFBP-3 molar ratio tended to increase during winter, whereas IGFBP-3 levels peaked in autumn and declined in winter. No data were available for the summer season.
IGF-1, insulin-like growth factor 1; IGFBP-3, insulin-like growth factor binding protein 3.
Given these seasonal trends, we further explored the potential correlation between ambient temperature on the day of blood sampling and circulating hormone levels. Daily temperature data (minimum, maximum, and mean) were obtained from the Korea Meteorological Administration and matched to each participant’s blood collection date. Pearson correlation analysis suggested that IGF-1 tended to be moderately and inversely associated with minimum temperature (r = −0.446), with comparable negative tendencies observed with maximum and mean temperatures (Fig. 5). In contrast, IGFBP-3 showed a moderate tendency toward a positive association with minimum temperature (r = 0.569), as well as with maximum and mean temperatures.
Figure 5. Correlation between ambient temperature on the day of blood sampling and serum hormone levels. Pearson correlation analysis revealed a moderate negative correlation between IGF-1 levels and minimum temperature, and a moderate positive correlation between IGFBP-3 levels and minimum temperature.
IGF-1, insulin-like growth factor 1; IGFBP-3, insulin-like growth factor binding protein 3.
DISCUSSION
Serum IGF-1 levels vary with age, sex, pubertal stage, physiological status, and ethnicity [6,10]. Avichai et al. recently analyzed millions of hormone test results and demonstrated that human hormone concentrations exhibit seasonal variability of several percent. In their study, most pituitary hormones peaked in late summer, whereas effector hormones secreted by peripheral organs reached maximum levels in winter or spring [11]. Consistent with these findings, we observed that IGF-1 concentrations were relatively higher during the colder seasons (spring and winter) and lower in the fall.
Several analytical methods—including enzyme-linked immunosorbent assay, chemiluminescent immunoassay, radioimmunoassay, and IRMA—are used to measure serum IGF-1 and IGFBP-3. Because of inter-assay variability, individual laboratories are advised to establish method-specific reference ranges. Notably, manufacturers of IRMA kits provide only broad reference limits and recommend institution-specific values.
Using the LMS method, Tsuyoshi et al. [12] reported population-based percentile curves and reference ranges for serum IGF-1 in 1,685 healthy Japanese individuals (845 males and 840 females) aged 0–83 years. They employed LMSchartmaker Pro version 2.3 (Institute of Child Health, London, UK) to generate smooth percentile curves. Similarly, Stojanovic et al. [13] reported age-specific reference values for serum IGF-1 in 1,200 healthy Serbian adults aged 21–80 years, using LMSchartmaker Light version 2.54 (Huiqi Pan, Tim Cole, Medical Research Council, Swindon, UK).
However, both the Pro and Light versions of LMSchartmaker are only supported up to Windows 7 and are not functional on Windows 10 or later, which represents a technical limitation. Additionally, Jeong et al. [14] established decade-based reference ranges for serum IGF-1 and IGFBP-3 in 354 healthy Korean adults (180 males and 174 females) aged 21–70 years. They found a significant negative correlation between age and IGF-1 in both males (r = −0.34, p < 0.001) and females (r = −0.50, p < 0.001). Although IGFBP-3 also tended to decrease with age, statistical significance was reached only in females (males: r = −0.049, p = 0.5225; females: r = −0.2293, p = 0.0029) [14].
These findings align with our results. Among healthy adult males aged 19–40 years, IGF-1 decreased with age, whereas IGFBP-3 showed no clear age-related pattern. Given the limited number of reports describing IGF-1 and IGFBP-3 reference ranges in specific populations, this single-center study contributes valuable data from 255 self-reported healthy young men. By minimizing inter-assay variability, we enhanced data homogeneity and strengthened the reliability of our results.
Furthermore, a major strength of this study is the inclusion of a seasonal variation analysis. Participants were classified according to the season in which blood samples were collected, and hormone levels were compared among the seasonal groups. Additionally, ambient- temperature data on the day of blood sampling were obtained from the Korea Meteorological Administration and matched to each participant, allowing exploration of the relationship between temperature and hormone levels. This approach yields new insights into the seasonal variability of IGF-1 and IGFBP-3 levels.
The reference ranges derived from this study are expected to serve as important benchmarks for clinical and research settings. Moreover, the gamlss::lms and gamlss::centiles functions in R allow flexible recalculation and application of age-specific reference ranges. The methodology and results presented here can, therefore, serve as foundational data for future clinical trials and research involving IGF-1 and IGFBP-3 measurements. In clinical studies, applying flexible, institution-specific reference ranges derived using LMS, along with consideration of assay methods, can help minimize unnecessary participant dropout and facilitate accurate interpretation of results.
Nevertheless, several limitations must be acknowledged. First, serial measurements across different seasons were not available for individual participants. Second, data collection during the summer season was limited, restricting comprehensive seasonal comparisons. Third, the periods of data collection for each seasonal group were not evenly distributed. Despite these limitations, the establishment of reference ranges and the seasonal analysis in a relatively narrow age range of healthy males provide meaningful insights and valuable reference data for future studies.
Footnotes
Conflict of Interest: - Authors: Nothing to declare
- Reviewers: Nothing to declare
- Editors: Nothing to declare
Reviewer: This article was reviewed by peer experts who are not TCP editors.
- Conceptualization: Bae KS.
- Formal analysis: Bae KS.
- Investigation: Lee SH.
- Methodology: Bae KS, Lee H.
- Writing - original draft: Jung SM.
- Writing - review & editing: Bae KS, Kim EN.
References
- 1.Al-Samerria S, Radovick S. The role of insulin-like growth factor-1 (IGF-1) in the control of neuroendocrine regulation of growth. Cells. 2021;10:2664. doi: 10.3390/cells10102664. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Wang Y, Zhang H, Zhang X, Mu P, Zhao L, Qi R, et al. The role of IGFBP-3 in tumor development and progression: enlightenment for diagnosis and treatment. Med Oncol. 2024;41:141. doi: 10.1007/s12032-024-02373-x. [DOI] [PubMed] [Google Scholar]
- 3.Liu Q, Jiang J, Zhang X, Zhang M, Fu Y. Comprehensive analysis of IGFBPs as biomarkers in gastric cancer. Front Oncol. 2021;11:723131. doi: 10.3389/fonc.2021.723131. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Huang R, Shi J, Wei R, Li J. Challenges of insulin-like growth factor-1 testing. Crit Rev Clin Lab Sci. 2024;61:388–403. doi: 10.1080/10408363.2024.2306804. [DOI] [PubMed] [Google Scholar]
- 5.Gelander L, Bjarnason R, Carlsson LM, Albertsson-Wikland K. Growth hormone-binding protein levels over one year in healthy prepubertal children: intraindividual variation and correlation with height velocity. Pediatr Res. 1998;43:256–261. doi: 10.1203/00006450-199802000-00017. [DOI] [PubMed] [Google Scholar]
- 6.Friedrich N, Wolthers OD, Arafat AM, Emeny RT, Spranger J, Roswall J, et al. Age- and sex-specific reference intervals across life span for insulin-like growth factor binding protein 3 (IGFBP-3) and the IGF-I to IGFBP-3 ratio measured by new automated chemiluminescence assays. J Clin Endocrinol Metab. 2014;99:1675–1686. doi: 10.1210/jc.2013-3060. [DOI] [PubMed] [Google Scholar]
- 7.Seo Y, Park J, Kim M, Sung C, Kwon OS, Lee HJ, et al. Optimization, validation, and comparison of a rapid method for the quantification of insulin-like growth factor 1 in serum using liquid chromatography-high-resolution mass spectrometry. Drug Test Anal. 2021;13:451–459. doi: 10.1002/dta.2944. [DOI] [PubMed] [Google Scholar]
- 8.Frystyk J, Freda P, Clemmons DR. The current status of IGF-I assays--a 2009 update. Growth Horm IGF Res. 2010;20:8–18. doi: 10.1016/j.ghir.2009.09.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Stasinopoulos M. Functions for fitting the Generalized Additive Models for Location Scale and Shape. R package version 5.4-22 (2024) [Internet] [Accessed January18, 2025]. https://CRAN.R-project.org/package=gamlss .
- 10.Cruickshank JK, Heald AH, Anderson S, Cade JE, Sampayo J, Riste LK, et al. Epidemiology of the insulin-like growth factor system in three ethnic groups. Am J Epidemiol. 2001;154:504–513. doi: 10.1093/aje/154.6.504. [DOI] [PubMed] [Google Scholar]
- 11.Tendler A, Bar A, Mendelsohn-Cohen N, Karin O, Korem Kohanim Y, Maimon L, et al. Hormone seasonality in medical records suggests circannual endocrine circuits. Proc Natl Acad Sci U S A. 2021;118:e2003926118. doi: 10.1073/pnas.2003926118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Isojima T, Shimatsu A, Yokoya S, Chihara K, Tanaka T, Hizuka N, et al. Standardized centile curves and reference intervals of serum insulin-like growth factor-I (IGF-I) levels in a normal Japanese population using the LMS method. Endocr J. 2012;59:771–780. doi: 10.1507/endocrj.ej12-0110. [DOI] [PubMed] [Google Scholar]
- 13.Stojanovic M, Popevic M, Pekic S, Doknic M, Miljic D, Medic-Stojanoska M, et al. Serum insulin-like growth factor-1 (IGF-1) age-specific reference values for healthy adult population of Serbia. Acta Endocrinol (Bucur) 2021;17:462–471. doi: 10.4183/aeb.2021.462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Jeong IK, Byun JK, Noh J, Kim SW, Chung YS, Park TS, et al. Reference ranges of serum insulin-like growth factor-I and insulin-like growth factor binding protein-3: results from a multicenter study in healthy Korean adults. Endocrinol Metab (Seoul) 2020;35:954–959. doi: 10.3803/EnM.2020.785. [DOI] [PMC free article] [PubMed] [Google Scholar]