Proposed reference intervals of serum insulin-like growth factor-1 (IGF-1) levels in older Japanese populations

Abstract

Measurements of serum insulin-like growth factor 1 (IGF-1) levels are useful surrogate markers for the diagnosis and management of patients with growth hormone-related disorders. We have previously published normative data of serum IGF-1 levels for the Japanese population aged 0–77 years by combining and analyzing previously reported references, which were separately and independently constructed, to properly reflect data in the transition period. Although the reference is widely used in both clinical and research settings, the reference did not include data for those aged >77 years, raising the question of how we would evaluate patients over those ages. In this study, we extended the age- and sex-specific reference ranges of serum IGF-1 levels to the age of 80 years by reanalyzing combined data on serum IGF-1 levels from previously published references. Based on our results, we proposed that individuals aged >80 years can be evaluated using the references set at the age of 80 years. However, our proposal was based on a very limited number of participants. Therefore, physicians should exercise caution when interpreting IGF-1 standard deviation scores for those aged >80 years because they are not exactly correct but acceptable.

SERUM LEVELS of insulin-like growth factor 1 (IGF-1) are considered to reflect the endogenous growth hormone (GH) status. IGF-1 measurements are currently used for the diagnosis and management of patients with GH-related disorders [1-3]. Given that serum IGF-1 levels vary with age, sex, pubertal stage, physiological condition, and ethnicity [4-6], normative IGF-1 reference values were established in various countries [1, 5-7], including Japan [8].

We have previously published normative data on serum IGF-1 levels for Japanese aged 0–77 years [8] by combining and analyzing previously reported references [9, 10], which were separately and independently constructed, to reflect data in the transition period, when serum IGF-1 levels rapidly decline. At present, this reference is widely used both in clinical and research settings. However, as the reference did not include data for those aged >77 years, many physicians requested proposals for reference intervals for these individuals.

In this study, we reanalyzed serum IGF-1 levels by combining data previously analyzed in 2012 [8] and reported in 1996 [11] to propose normative data for the Japanese population from birth to senescence.

Materials and Methods

In a study published in 1996 [11], participants aged ≥70 years were extracted to extend the present references published in 2012 [8] to older age groups because the study published in 2012 included only a small number of participants aged ≥70 years [8]. As a result, 60 participants (12 males and 48 females) from the 1996 cohort [11] were added for the present analyses. In total, 1,745 data of serum IGF-1 levels (857 males and 888 females) were reanalyzed to determine the reference intervals. In previous studies [8-11], serum IGF-1 levels were determined using a commercially available immunoradiometric assay (IRMA) kit (IGF-I assay “Daiichi” TFB, Inc., Japan) [12].

The Ethical Committee of the Omi Medical Center approved this study using secondary data from previous publications [8-11] (Approval no. 2023-002).

Statistical Analyses

The centile curves were constructed using the lambda-mu-sigma (LMS) Chart Maker Pro version 2.3 (The Institute of Child Health, London), which fit smooth centile curves to the reference data of serum IGF-1 with the LMS method [13]. The analyses did not include outliers because we had already excluded them in the previous studies [8-11]. In this method, percentiles at each age were based on the power of the age-specific Box-Cox power transformations that are used for normalization. The final curves of the percentiles were established by three curves representing skewness (L curve), median (M curve), and coefficient of variation (S curve). Cubic spline curves were used for smoothing the L, M, and S values for each age and sex. The standard deviation (SD) score (Z score) of serum IGF-1 measurement (y value) was calculated from the L, M, and S curves, using values appropriate for the age and sex, by the following equation: Z = [(y/M)^L – 1]/(L × S), and if L = 0, Z = ln(y/M)/S

Results

Table 1 shows the numbers and sex of participants in each group aged ≥60 years analyzed in this study. Fig. 1a and 1b present scattered plots and smoothing curves (black circles and solid lines, respectively) of +2SD, +1SD, median, –1SD, and –2SD of serum IGF-1 levels for males and females aged 60–80 years, respectively. Since cases of individuals aged >80 years were very limited, developing the percentiles in those age groups were difficult. On the contrary, data in those age groups were distributed within median ±2SD at the age of 80 years in both males and females, except for one individual, when distributions within ±2SD at the age of 80 years were extended to older ages as depicted in Fig. 1a and 1b, respectively (dotted lines). Tables 2 and 3 show the IGF-1 reference ranges, and L, M, and S parameters for IGF-1 for males and females over age 60, respectively.

Table 1

Age and sex distribution of the study population

Age (years)	Male	Female
60–64	51	50
65–69	51	51
70–74	21	50
75–79	11	11
80–84	4	8
85–89	0	2
≥90	0	2
Total	138	174

Fig. 1

Scatter plots and smoothing curves of +2SD, +SD, median, –SD, –2SD of serum IGF-1 values in Japanese male (a) and female (b) individuals aged ≥60 years.

Solid lines represent smoothing curves established in this study, whereas dotted lines indicate ranges proposed for individuals aged ≥80 years. Discussions should be referred.

Table 2

Serum IGF-1 reference ranges and L, M, and S parameters for IGF-1 for males aged ≥60 years

Age (years)	L	M	S	–2SD	–1SD	0SD	+1SD	+2SD
60	0.300	141	0.268	79	107	141	182	232
61	0.312	140	0.270	77	105	140	181	230
62	0.327	138	0.272	76	104	138	180	228
63	0.344	137	0.274	75	103	137	178	226
64	0.363	135	0.276	73	101	135	176	224
65	0.383	134	0.278	72	100	134	174	221
66	0.405	132	0.280	70	98	132	172	219
67	0.428	130	0.282	68	96	130	170	216
68	0.453	128	0.284	66	95	128	168	213
69	0.479	126	0.287	65	93	126	165	209
70	0.506	124	0.289	63	91	124	162	206
71	0.533	122	0.291	61	89	122	160	202
72	0.562	119	0.294	58	87	119	157	198
73	0.591	117	0.296	56	84	117	153	194
74	0.620	114	0.299	54	82	114	150	190
75	0.650	112	0.301	52	80	112	147	185
76	0.680	109	0.303	50	78	109	144	181
77	0.710	106	0.306	48	75	106	140	177
78	0.740	104	0.307	45	73	104	137	172
79	0.770	101	0.310	43	71	101	133	167
≥80	0.801	98	0.313	41	68	98	130	163

IGF-1: insulin growth factor-1

SD: standard deviation

Table 3

Serum IGF-1 reference ranges and L, M, and S parameters for IGF-I for females aged ≥60 years

Age (years)	L	M	S	–2SD	–1SD	0SD	+1SD	+2SD
60	0.172	121	0.263	70	93	121	157	201
61	0.170	120	0.264	69	91	120	155	198
62	0.168	118	0.266	68	90	118	153	196
63	0.167	116	0.267	66	88	116	151	194
64	0.168	114	0.269	65	87	114	149	191
65	0.169	112	0.270	64	85	112	146	188
66	0.170	110	0.272	62	84	110	144	186
67	0.173	109	0.273	61	82	109	142	183
68	0.177	107	0.275	60	80	107	139	180
69	0.181	105	0.276	59	79	105	137	177
70	0.185	103	0.278	57	77	103	135	175
71	0.189	101	0.279	56	76	101	133	172
72	0.194	100	0.281	55	75	100	131	170
73	0.199	98	0.282	54	73	98	129	167
74	0.203	96	0.284	53	72	96	127	165
75	0.208	95	0.286	52	71	95	125	163
76	0.212	93	0.287	50	69	93	123	160
77	0.217	92	0.289	49	68	92	121	158
78	0.221	90	0.290	48	67	90	119	155
79	0.235	86	0.310	44	62	86	115	152
≥80	0.243	84	0.310	43	61	84	113	149

IGF-1: insulin growth factor-1

SD: standard deviation

Discussion

In this study, we extended the age- and sex-specific ranges for serum IGF-1 levels to the age of 80 years by reanalyzing data of 1,745 Japanese individuals enrolled in the previous studies [8-11]. However, we could not determine the correct distributions of individuals aged >80 years because of the limited number of cases. Studies have reported the progressive decrease in serum IGF-1 concentrations before reaching a plateau in older ages [1, 7, 14-16], although the exact age for reaching a plateau is unknown. For example, Bidlingmaier et al. reported the median serum IGF-1 concentrations of males and females aged 81–85 and 86–90 years, i.e., 86.1 and 85.0 ng/mL, and 89.1 and 90.3 ng/mL, respectively, presenting that their progressive decline reached a plateau at approximately the age of 80 years [1]. On the contrary, Stojanovic et al. reported that a plateau of the decline in IGF-1 levels was observed at the age of 46–70 years [14]. In our cohort, nearly all individuals aged >80 years were distributed within the median ±2SD at the age of 80 years in both males and females. In addition, the median serum IGF-1 concentrations of males and females at the age of 79 and 80 years were 101 and 98 ng/mL, and 86 and 84 ng/mL, respectively, which may suggest that the decline after the age of 80 years may be negligible. Thus, the near plateau in our cohort may occur at approximately the age of 80 years, although there is the possibility that serum IGF-1 concentrations decrease very gradually after the age of 80 years. Therefore, we propose that the distributions within the median ±2SD at the age of 80 years can be extrapolated for calculating the SD scores for individuals aged ≥80 years.

This study has three main limitations. First, serum IGF-1 levels were determined by the old IRMA kit (IGF-I assay “Daiichi”) [12], which is not available these days. However, the old assay for IGF-1 had been reported to show an excellent correlation with the newly introduced electro-chemiluminescent immunoassay (ECLIA) IGF-1 (Eclusys IGF-1; Roche Diagnostics, Japan) [17]. Therefore, we think that we can establish the reference using previous datasets [8-11], and the IGF-1 levels determined by the ECLIA IGF-1 assay can be extrapolated as the same numbers by the old assay in clinical settings. Second, there is the possibility that recent IGF-1 levels in individuals aged ≥70 years may have changed from those in 1996 and 2012 due to equivalent changes in nutritional or physical conditions. According to the national nutrition survey, average body mass indices (SD values) of males and females aged ≥70 years in 1996, 2012 and 2019 were 22.3 (3.2) and 22.9 (3.6) kg/m², 23.3 (3.1) and 22.9 (3.5) kg/m², and 23.4 (3.3) and 22.9 (3.5) kg/m², respectively [18]. The survey also reported that average number of steps (SD values) of males and females aged ≥70 years in 1996, 2012, and 2019 were 5,078 (3,487) and 4,202 (3,287) steps, 5,216 (378) and 4,283 (3,168) steps, and 5,016 (3,682) and 4,225 (3,213) steps, respectively [18]. These data suggest that nutritional or physical status of Japanese individuals aged ≥70 years did not change substantially between 1996 and 2019, although it is not well known how these small changes with times affect serum IGF-1 levels in elderly individuals. Third, we could not determine the correct distributions for serum IGF-1 levels of individuals aged >80 years because of the very limited number of cases. We discussed that the gradual decrease in serum IGF-1 concentrations had seemed to reach a near plateau in our cohort approximately at the age of 80 years, but there is the possibility that further minimal decrease may occur after the age of 80 years. In that case, we will slightly underestimate the SD scores for individuals aged ≥80 years, when our proposal is adopted. Ideally, the correct distributions should be determined with additional sufficient numbers of healthy individuals aged >80 years. However, it was not easy to recruit healthy volunteers aged >80 years. Despite these limitations, we believe that our proposal is valid and practical due to the above-discussed reasons. However, physicians should exercise caution when interpreting IGF-1 SD scores for those aged >80 years because they are not exactly correct but acceptable.

In conclusion, this study established age- and sex-specific normative data of IGF-1 for Japanese aged 0–80 years and proposed that individuals aged >80 years can be evaluated using the reference set at the age of 80 years. These reference intervals of serum IGF-1 levels could be widely used in clinical practice and would assist in calculating the SD scores of patients.

Acknowledgments

We thank Dr. Susumu Yokoya, Dr. Hiroshi Nishioka, and Dr. Yutaka Takahashi for their valuable discussions and critical comments on the manuscript. This study was supported in part by a Grant-in-Aid for Scientific Research from the Foundation for Growth Science in Japan.

Disclosure

None of the authors have any potential conflicts of interest associated with this research.

References

• 1   Bidlingmaier  M,  Friedrich  N,  Emeny  RT,  Spranger  J,  Wolthers  OD, et al. (2014) Reference intervals for insulin-like growth factor-1 (igf-i) from birth to senescence: results from a multicenter study using a new automated chemiluminescence IGF-I immunoassay conforming to recent international recommendations. J Clin Endocrinol Metab 99: 1712–1721.
• 2   Friedrich  N,  Wolthers  OD,  Arafat  AM,  Emeny  RT,  Spranger  J, et al. (2014) 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 99: 1675–1686.
• 3   Johannsson  G,  Bidlingmaier  M,  Biller  BMK,  Boguszewski  M,  Casanueva  FF, et al. (2018) Growth Hormone Research Society perspective on biomarkers of GH action in children and adults. Endocr Connect 7: R126–R134.
• 4   Cruickshank  JK,  Heald  AH,  Anderson  S,  Cade  JE,  Sampayo  J, et al. (2001) Epidemiology of the insulin-like growth factor system in three ethnic groups. Am J Epidemiol 154: 504–513.
• 5   Juul  A,  Bang  P,  Hertel  NT,  Main  K,  Dalgaard  P, et al. (1994) Serum insulin-like growth factor-I in 1030 healthy children, adolescents, and adults: relation to age, sex, stage of puberty, testicular size, and body mass index. J Clin Endocrinol Metab 78: 744–752.
• 6   Löfqvist  C,  Andersson  E,  Gelander  L,  Rosberg  S,  Blum  WF, et al. (2001) Reference values for IGF-I throughout childhood and adolescence: a model that accounts simultaneously for the effect of gender, age, and puberty. J Clin Endocrinol Metab 86: 5870–5876.
• 7   Liao  ZH,  Yin  QQ,  Wan  JX,  He  W,  Ji  W, et al. (2016) Serum Insulin-like growth factor-1 levels of healthy adults in southern China. Endocr J 63: 1081–1086.
• 8   Isojima  T,  Shimatsu  A,  Yokoya  S,  Chihara  K,  Tanaka  T, et al. (2012) 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 59: 771–780.
• 9   Fujieda  K,  Shimatsu  A,  Hanew  K,  Tanaka  T,  Yokoya  S, et al. (1996) Clinical evaluation of serum IGF-I, IGF-II and IGFBP-3 measured by IRMA kits in childhood. Horumon to Rinsho 44: 1229–1239 (In Japanese).
• 10   Shimatsu  A,  Chihara  K,  Hizuka  N,  Teramoto  A,  Tanaka  T, et al. (2007) Reference intervals of serum insulin-like growth factor-I in adult Japanese population: calculated with LMS methods. Horumon to Rinsho 55: 393–399 (In Japanese).
• 11   Shimatsu  A,  Fujieda  K,  Hanew  K,  Tanaka  T,  Yokoya  S, et al. (1996) Clinical evaluation of serum IGF-I, IGF-II and IGFBP-3 measured by IRMA kits in adulthood. Horumon to Rinsho 44: 1129–1138 (In Japanese).
• 12   Takasu  S,  Tsuchiya  M,  Mori  K,  Iwamoto  K,  Kasahara  H, et al. (1996) Development and fundamental studies of non-extract IGF-I/II IRMA. Horumon to Rinsho 44: 383–391 (In Japanese).
• 13   Cole  TJ,  Green  PJ (1992) Smoothing reference centile curves: the LMS method and penalized likelihood. Stat Med 11: 1305–1319.
• 14   Stojanovic  M,  Popevic  M,  Pekic  S,  Doknic  M,  Miljic  D, et al. (2021) Serum Insulin-like Growth Factor-1 (IGF-1) age-specific reference values for healthy adult population of Serbia. Acta Endocrinol (Buchar) 17: 462–471.
• 15   Granada  ML,  Ulied  A,  Casanueva  FF,  Pico  A,  Lucas  T, et al. (2008) Serum IGF-I measured by four different immunoassays in patients with adult GH deficiency or acromegaly and in a control population. Clin Endocrinol (Oxf) 68: 942–950.
• 16   Berrigan  D,  Potischman  N,  Dodd  KW,  Hursting  SD,  Lavigne  J, et al. (2009) Race/ethnic variation in serum levels of IGF-I and IGFBP-3 in US adults. Growth Horm IGF Res 19: 146–155.
• 17   Shimatsu  A,  Ogata  S,  Yoshida  M (2019) Performance evaluation of the insulin-like growth factor-I assay with ECLIA reagent Eclusys IGF-1. Igaku to Yakugaku 76: 541–547 (In Japanese).
• 18  National Institute of Health and Nutrition, Japan. Annual changes in primary health indicatiors (physical status questionnaire) from the national nutrition survey. https://view.officeapps.live.com/op/view.aspx?src=https%3A%2F%2Fwww.nibiohn.go.jp%2Feiken%2Fkenkounippon21%2Fdownload_files%2Fkeinen_henka%2Fshintai%2F05.xlsx&wdOrigin=BROWSELINK accessed on July 17, 2023 (In Japanese).