Serum free IGF-I, total IGF-I, IGFBP-1 and IGFBP-3 levels in an elderly population: relation to age and sex steroid levels

Clinical Endocrinology (1998) 48, 471–478

471q 1998 Blackwell Science Ltd

Serum free IGF-I, total IGF-I, IGFBP-1 and IGFBP-3levels in an elderly population: relation to ageand sex steroid levels

J. A. M. J. L. Janssen*, R. P. Stolk†‡,H. A. P. Pols*†, D. E. Grobbee†‡, F. H. de Jong*and S. W. J. Lamberts*Departments of *Internal Medicine III, and †Epidemiology& Biostatistics, Erasmus University, Rotterdam, and‡Julius Centre for Patient Oriented Research, UtrechtUniversity, Utrecht, The Netherlands

(Received 10 February 1997; returned for revision 14 March1997; finally revised 21 April 1997; accepted 1 July 1997)

Summary

BACKGROUND Most previous studies concerningthe relationship between IGF-I and age used assaysmeasuring total IGF-I. Although free IGF-I is con-sidered of greater biological relevance, little isknown about its relationship with sex steroids levelsin elderly healthy subjects.MEASUREMENTS In a cross-sectional study of 218healthy people (103 men, 115 women) aged 55–80years we measured serum total and free IGF-I,IGFBP-1 and IGFBP3 levels and sex steroids. Freeandrogen index and free oestradiol index wereused as an indicator for free oestradiol and freetestosterone levels, respectively.RESULTS Free IGF-I levels did not decline with age inthe whole study population. Free IGF-I levels evenincreased in individuals above 70 years of age incomparison to those aged between 55 and 70 years(mean 6 SE 0.106 6 0.007 nmol/l vs. 0 .086 6 0.004nmol/l, P ¼ 0.009). Total IGF-I and IGFBP-3 decreasedwith age ( r ¼ ¹0.20, P ¼ 0.005 and r ¼ ¹0.24,P ¼ 0.001, respectively). Total IGF-I levels were posi-tively related with free oestrogen index in both sexes.Free IGF-I did not relate to free oestrogen or androgenindex. In women only, free IGF-I was related positivelywith DHEAS while IGFBP-1 was inversely correlatedwith DHEAS.CONCLUSIONS Free IGF-I levels do not decrease with

age and are even higher in individuals above 70 years.There was no relationship between free IGF-I andfree androgen or oestrogen index in either gender. Wehypothesize that higher free IGF-I levels in olderpersons may be the consequence of selective survivalin the cohort: subjects with high free IGF-I levels maylive longer. The absence of a relationship betweenfree IGF-I levels and free androgen and oestrogenindices suggests that there is no direct interactionbetween the biological activity of circulating IGF-Ilevels and sex hormone production in a healthyageing population.

The changes in total circulating insulin-like growth factor-I(IGF-I) with age parallel those of growth hormone (GH) (Ho &Hoffman, 1993). After a peak during early adulthood, thereappears to be a gradual decrease in the serum concentrations oftotal IGF-I and insulin-like growth factor binding protein-3(IGFBP-3), which is progressive with increasing age (Rudmanet al., 1990; Copelandet al., 1990; Cohenet al., 1992).However, IGF-I measured in serum is the total extractableIGF-I, which offers only a crude estimate of biologically activeIGF-I due to the wide interindividual variation in circulatingIGFBP (Orskovet al., 1996). Free IGF-I probably has greaterphysiological and clinical relevance than total IGF-I in analogywith sex and adrenal steroids and thyroid hormones, andaccounts for approximately 1% of total IGF-I (Frystyket al.,1994). The amount of total IGF-I is dependent on the free IGF-Ilevel and on the concentrations of the specific insulin-likegrowth factor binding proteins (IGFBPs). IGFBP-3 is quantita-tively by far the major binding protein and it is thought tofunction as an intravascular reservoir and buffer for IGF-I(Bang et al., 1994). IGFBP-1 has been proposed as an acuteregulator of IGF-I bioactivity and might simultaneously bothinhibit and potentiate IGF-I action at different sites (Hallet al.,1993).

Consequently, it seems desirable to distinguish bound andunbound components of IGF-I when studying the IGF/IGFBPsystem (Orskovet al., 1996). Recently a method has beendeveloped and validated to measure free IGF-I levels (Leeetal., 1994; Juulet al., 1996). The increase in serum total IGF-Iand IGFBP-3 levels during puberty suggest the influence of

Correspondence: J. A. M. J. L. Janssen, Department of InternalMedicine III, Room D438, University Hospital Dijkzigt, Dr.Molewaterplein 40, 3015 GD Rotterdam, The Netherlands. Fax:þ3110 463 3268.

oestrogens and/or androgens on circulating IGF-I and IGFBP-3levels (Daughaday & Rotwein, 1989; Blum, 1996). This effectmight be at least partially mediated through stimulated growthhormone (GH) secretion (Blum, 1996).

Little is known about free IGF-I levels or the relationshipbetween the circulating IGF-I/IGFBP system and circulatingsex steroids at older age. We investigated the free IGF-I, totalIGF-I, IGFBP-1 and IGFBP-3 and plasma sex steroid levels in ahealthy older population.

Subjects and methods

Study population

For the present study, a sample of participants from theRotterdam Study was invited for an additional examination.The Rotterdam Study is a population-based cohort study ofthe determinants of chronic disabling diseases in the elderly.All of the approximately 10 000 inhabitants of a suburb ofRotterdam, aged 55 years and over, were invited to participateas described elsewhere (Hofmanet al., 1991). Overall, 7983participants were examined in the Rotterdam Study (responserate 78%).

The population for the present study included 218 peopleaged 55–80 years, who had completed the baseline visit of theRotterdam Study not more than 6 months earlier. Subjects withacute, psychiatric or endocrine diseases, including diabetesmellitus treated with medication, were not invited. Compared tothe other participants of the Rotterdam Study of the same agewithout known diabetes mellitus, there were no differencesin age and gender distribution. Only 4 women in this studypopulation were on hormone replacement therapy. These wereexcluded from analysis. The study population was stratified byage into three groups of about equal size: group 1 (55–64 yearsold, 73 subjects), group 2 (65–70 years old, 72 subjects) andgroup 3 (>70 years old, 69 subjects). Informed consent wasobtained from all subjects and the study was approved by theMedical Ethics Committee of Erasmus University MedicalSchool.

Measurements

Participants were examined in the morning after an overnightfast. Fasting blood samples were taken by venepuncturebetween 0800 and 0900 h and allowed to coagulate for 30minutes. Subsequently serum was separated by centrifugationand frozen quickly in liquid nitrogen. Free IGF-I wasmeasured with a commercially available two-site immuno-radiometric assay (Diagnostic System Laboratories Inc.,

Webster, Texas, USA; intra-assay and interassay CV: 10.3%and 10.7%, respectively) (Leeet al., 1994; Juulet al., 1996).Estimation of free IGF-I in other assay systems usuallyinvolves size-exclusion extraction followed by a displacementimmunoassay. Sample matrix alterations with consequentdisturbance of the free/bound equilibrium is an inevitableconsequence. The free IGF-I assay used in this study does notneed an initial sample extraction as part of the standardprocedure to measure IGF-I. Samples were added directly totubes coated with IGF-I antibodies, washed, incubated with125I-labelled antibody directed to a second epitope on IGF-I,washed and counted. Assay standards are rhIGF-I: 0.04–2.6 nmol/l, the minimal detection limit is 4 pmol/l. There is nocross-reactivity with IGF-II and no residual IGFBP-1 orIGFBP-3 is detectable after the first wash. The free IGF-Ifraction measured in this assay is a combination of truefree IGF-I and the fraction of IGF-I which can be readilydissociated from IGFBPs under the specific assay conditions(Lee et al., 1994). Total IGF-I was determined by acommercially available radioimmunoassay (Medgenix Diag-nostics, Brussels, Belgium, intra-assay and interassay coeffi-cients of variation (CV): 6.1%; 9.9%). Commercially availableimmunoradiometric assays were also used for the measurementof IGFBP-1 and IGFBP-3 (Diagnostic System LaboratoriesInc., intra-assay and interassay CV for IGFBP-1: 4.0% and6.0% respectively; and for IGFBP-3: 1.8% and 1.9%,respectively). Oestradiol and oestrone were assayed with aradioimmunoassay (Diagnostic Products Corporation, LosAngeles, USA; oestradiol: intra-assay and interassay CV:7.0% and 8.1%; oestrone: intra-assay and interassay CV:9.4% and 11.1%, respectively). SHBG was assayed with acommercially available immunoradiometric assay (DiagnosticProducts Corporation; intra-assay and interassay CV: 3.6% and6.9%, respectively). The ratio of oestradiol to SHBG was usedas an index of free oestradiol (Khawet al., 1988). Testosteronewas measured with a non-commercial radioimmunoassay(intra-assay and interassay CV: 5.6%; 9.0%) (Verjanset al.,1973). The ratio of total testosterone to SHBG was used as anindex of free testosterone (Boltonet al., 1989; Coxet al., 1992).DHEAS and androstenedione were assayed by radioimmu-noassay (Diagnostic Products Corporation; DHEAS: intra-assay and interassay CV: 5.3% and 7.0%; androstenedioneintra-assay and interassay CV: 8.3% and 9.2%). Insulin wasdetermined by a commercially available radioimmunoassay(Medgenix Diagnostics, intra-assay and interassay CV: 8.0%;13.7%). Serum glucose levels were determined using a standardglucose hexokinase method. Height and weight of the subjectswere measured while wearing indoor clothes and without shoes.Body mass index was defined as weight divided by the squareof height (kg/m2), body fat distribution was estimated using theratio of waist and hip circumferences.

472 J. A. M. J. L. Janssen et al.

q 1998 Blackwell Science Ltd,Clinical Endocrinology, 48, 471–478

Statistical analysis

The clinical characteristics of the study population stratifiedby gender are expressed as the mean and standard error (SE).Mean total IGF-I, free IGF-I, IGFBP-1 and IGFBP-3 levelswith SE were also calculated after stratification for age. Linearregression analyses were used to calculate differences in thebaseline characteristics between men and women with adjust-ment for age. Pearson’s correlation coefficients were calculatedto assess associations between the variables. One-way ANOVAwas used to test for difference in the free IGF-I levels acrossage groups and multiple linear regression analysis wasperformed to adjust for confounding variables. A two-sidedP-value of<0.05 was considered significant. Analyses in whichthe values were logarithmically transformed yielded resultssimilar to those with untransformed data: the non-transformedresults are presented. All statistical analyses were performedwith the Stata statistical package (Computing Resource Center,Santa Monica, Ca, USA).

Results

The general characteristics of the study population arepresented in Table 1. Men were slightly older than women.

BMI, and mean (free or total) IGF-I and IGFBP-1 levels (age-adjusted) did not differ between the sexes. Mean free over totalIGF-I ratio and IGFBP-3 levels were significantly lower in menthan in women. Mean insulin and glucose levels did not differbetween the sexes. As expected, serum SHBG levels weresignificantly lower in men. Although the differences in meanserum oestradiol levels did not reach statistical significance,mean free oestradiol index and oestrone levels were signifi-cantly higher in men than in women as were total and freetestosterone. Mean DHEAS and androstenedione levels werealso significantly higher in men.

Total IGF-I levels decreased significantly with age in thewhole study population (r ¼ ¹0.20, P ¼ 0.005) (Fig. 1). FreeIGF-I levels did not significantly decline with age (r ¼ 0.12,P ¼ 0.08), but showed a significant increase in individualsabove 70 years when compared with those between 55 and 70years (P ¼ 0.009) (Fig. 1). This difference remained inmultivariate analysis after adjustments for BMI and insulin.IGFBP-1 levels showed no relation with age and IGFBP-3levels decreased with age (r ¼ ¹0.24, P ¼ 0.001) (Fig. 1).

Serum total IGF-I was positively related with free IGF-I,IGFBP-3, and inversely related with IGFBP-1 (Table 2). Serumfree IGF-I levels were positively correlated with IGFBP-3, andinversely with IGFBP-1 (Table 2). IGFBP-1 was negatively

IGF-1 in an elderly population 473


Table 1 General characteristics of the studypopulation Men Women

n ¼ 103 n ¼ 111mean (SE) mean (SE) P value*

Age (years) 67.6 (0.56) 65.9 (0.59) 0.05BMI (kg/m2) 26.4 (0.29) 26.6 (0.41) 0.70Total IGF-I (nmol/l) 19.3 (0.83) 17.9 (0.70) 0.11Free IGF-I (nmol/l) 0.088 (0.005) 0.095 (0.005) 0.22Free IGF-I/Total IGF-I (%) 0.51 (0.03) 0.59 (0.04) 0.05IGFBP-1 (nmol/l) 0.71 (0.07) 0.84 (0.10) 0.21IGFBP-3 (nmol/l) 103.5 (2.5) 116.6 (2.5) 0.002Insulin (IU/ml) 14.2 (0.8) 13.0 (0.8) 0.47Glucose (nmol/l) 6.0 (0.1) 5.8 (0.1) 0.24SHBG (nmol/l) 50.9 (2.0) 60.2 (2.7) 0.006Oestradiol (pmol/l) 106.4 (2.0) 82.1 (4.8) 0.10Free oestradiol index 2.26 (0.25) 1.65 (0.13) 0.009

(pmol/nmol)Oestrone (pmol/l) 194.2 (7.1) 126.6 (5.4) <0.001Testosterone (nmol/l) 20.4 (0.5) 1.4 (0.06) <0.001Free androgen index 0.45 (0.02) 0.03 (0.002) <0.001

(nmol/nmol)DHEAS (mmol/l) 4.1 (0.2) 2.6 (0.2) <0.001Androstenedione (nmol/l) 6.6 (0.3) 4.6 (0.2) <0.001

* Difference between men and women, adjusted for age. BMI, body mass index; IGF-I: insulin-likegrowth factor I; IGFBP: insulin-like growth factor binding protein; SHBG: sex-hormone bindingglobulin; DHEAS: dehydroepiandrosterone sulphate.

related to IGFBP-3 and insulin (Table 2). After adjustment forage and gender these relations remained unchanged.

Free IGF-I levels were inversely related to BMI in men (r ¼

¹0.23, P ¼ 0.03), but not in women (r ¼ 0.004, P ¼ 0.97).No relationship was observed between total IGF-I and BMI(men:r ¼ ¹0.03, P ¼ 0.77; women:r ¼ 0.09, P ¼ 0.35).

IGFBP-1 levels were inversely related with BMI in women(r ¼ ¹0.20,P ¼ 0.004) but not in men (r ¼ ¹0.16,P ¼ 0.12),

while no relationship was observed between IGFBP-3 levels

and BMI in either sex (men: r ¼ 0.12, P ¼ 0.24; women:r ¼ 0.09, P ¼ 0.38).

Both in men and women serum total IGF-I was positivelyrelated to the free oestradiol index, and inversely to SHBG(Table 3). In women, serum total IGF-I was also positivelyrelated to the free androgen index while serum free IGF-I waspositively related to DHEAS (Table 3). Serum IGFBP-1 waspositively related to SHBG and inversely to the free androgenindex both in men and women. In women (only), IGFBP-1 was



Fre

e I

GF

-1 (

nm

ol/

l)

55–64

(a)

0.00

Age (years)

0.05

0.10

0.15

65–70 >70

To

tal

IGF

-1 (

nm

ol/

l)

55–64

(b)

0

30

65–70 >70

25

20

15

10

5

IGFB

P-1

(n

mo

l/l)

55–64

(c)

0.00

0.60

0.80

1.00

65–70 >70

IGFB

P-3

(n

mo

l/l)

55–64

(d)

0

150

65–70 >70

75

50

25

0.40

0.20

125

100

//

Fig. 1 Free IGF-1 (a), total IGF-I (b), IGFBP-1 (c) and IGFBP-3 (d) levels according to three groups of increasing age (see text). Note thedifferent scales on the vertical axes, the increase in free IGF-I levels in individuals after 70 years in comparison to those between 55 and 70 years(#: ANOVA, P¼ 0.009), and the comparable decreases in total IGF-I and IGFBP-3 levels with increasing age (see text).

also inversely related to DHEAS. IGFBP-3 was positivelyrelated to the free androgen index and negatively to SHBG inboth sexes, and positively related to the free oestradiol index inwomen only (Table 3). Free oestradiol index (age-adjusted) wasinversely related to SHBG in women (r ¼ ¹0.53, P < 0.001)while free androgen index was inversely related to SHBG inboth sexes (men:r ¼ ¹0.68, P < 0.001; women:r ¼ ¹0.52,P < 0.001).

Discussion

In this population-based study in the elderly, total IGF-I andIGFBP-3 levels decreased with age while free IGF-I levels didnot, even tending to be higher in subjects over 70 years of age(Fig. 1). A close relationship was found between total IGF-I andIGFBP-3 levels with free androgen and free oestrogen levelsboth in elderly men and women. No relationship was found

between free IGF-I levels and free sex steroid indices in eithersex.

Before these findings can be accepted some issues need tobe addressed. An important, at present insufficiently answered,question in this regard remains whether the free IGF-I assayused in our study represents a better, more biologically activerepresentation of the IGF-I/IGFBP system than thecurrently used total IGF-I, IGFBP-1 and IGFBP-3 assays. Thefree IGF-I assay used in our study measures in unextractedserum the unbound and readily dissociable IGF-I (the socalled easily dissociable free fraction) (Bereketet al., 1996).It has become clear that during the transport of hormones intocells (e.g. steroid and thyroid hormones) the unbound hormonefraction is not the only component of the hormone in thecirculation which enters the cells (Griffin, 1996). Thismight also be true for IGF-I. Dissociation of IGFBP-boundIGF-I might also occur within the capillary bed, suggesting



Table 2 Pearson’s correlation coefficientsbetween (total and free) IGF-I, IGFBP-1,IGFBP-3, insulin and glucose levels for thewhole study population

Total IGF-I Free IGF-I IGFBP-1 IGFBP-3 Insulin

Total IGF-IFree IGF-I 0.35†IGFBP-1 ¹0.30† ¹0.17*IGFBP-3 0.30† 0.19§ ¹0.17*Insulin ¹0.03 0.05 ¹0.19§ 0.07Glucose 0.11 0.09 ¹0.12 0.03 0.46†

* P < 0.05, §P < 0.01, †P < 0.001.

Table 3 Pearson’s (age-adjusted) correlationcoefficients between (total and free) IGF-I,IGFBP-1, IGFBP-3, sex steroids and SHBG inmen (A) and women (B)

Total IGF-I Free IGF-I IGFBP-1 IGFBP-3

MenOestradiol 0.15 0.10 ¹0.08 ¹0.13Free oestradiol index 0.24* 0.13 ¹0.15 ¹0.07Testosterone ¹0.03 0.08 ¹0.08 ¹0.14Free androgen index 0.20 0.12 ¹0.35† 0.28*SHBG ¹0.20* ¹0.05 0.38† ¹0.24*DHEAS 0.13 0.01 ¹0.05 ¹0.03Androstenedione 0.08 0.01 0.07 0.01Oestrone 0.12 ¹0.009 ¹0.03 ¹0.07

WomenOestradiol ¹0.003 0.08 0.01 0.07Free oestradiol index 0.26§ 0.16 ¹0.15 0.38†Testosterone 0.07 ¹0.009 ¹0.05 ¹0.07Free androgen index 0.29§ 0.12 ¹0.21* 0.31§SHBG ¹0.30§ ¹0.18 0.31§ ¹0.43†DHEAS 0.06 0.24* ¹0.22* ¹0.002Androstenedione ¹0.002 0.07 ¹0.002 ¹0.02Oestrone ¹0.01 0.15 ¹0.12 ¹0.03

* P < 0.05, §P < 0.01, †P < 0.001.

that the active fraction is larger than the unbound fractionmeasuredin vitro. Although not yet proven, measurementof the easily dissociable free IGF-I fraction probably reflectsthe ‘true’ bioavailable IGF-I (the amount of IGF-I availablefor entry into most tissues) better than the currently used assays.In a previous study three fractions of IGF-I have been foundto be present in the circulation: the total amount, theultrafiltrated free fraction (measured by an ultrafiltratedcentrifuge method) and the easily dissociable free fraction(measured by the same free IGF-I assay as used in ourstudy) (Juulet al., 1996). These authors concluded that itremains to be proven to what extent the two latter fractionsreflect bioavailable IGF-I.

It should be realized that cross-sectional studies do notreflect the rate of change of free IGF-I in an individual. Asour study was cross-sectional, the higher free IGF-I levels insubjects above 70 years might be the consequence of a selectivesurvival in the cohort: subjects with high free IGF-I levels maylive longer. Alternatively, subjects with lower free IGF-I levelsmight have been excluded from the study because their physicalcondition (illness, frailty, or other causes) prevented a visit tothe research centre.

Apart from these limitations in the study itself, anexplanation for our observation of higher free IGF-I levels insubjects above 70 years might be that free IGF-I levels indeedincrease with age. An age-related increase has also beendescribed for IGF-II levels (Frystyket al., 1994). It has becomeclear that IGF-I levels are not only GH-dependent and that thereare a number of other factors which influence IGF-I production(Blum, 1996). The increase in free IGF-1 levels might be due toother age-related phenomena, such as a decreased metabolismand clearance of IGF-1, increased IGFBP-3 protease activity,increased abdominal adiposity, hyperinsulinaemia and hyper-glycaemia, which might modify IGF-I levels independently,despite the decrease in GH, total IGF-1 and IGFBP-3 levelsduring ageing (Corpaset al., 1992, 1993; Copelandet al.,1990). Alternatively, an increase in free IGF-I levels may alsorepresent an age-related IGF-I receptor or post-receptor defect(O’Connoret al., 1996).

According to previous findings, the decline in serum GHconcentration with age in men and women correlates withchanges in gonadal steroid levels (Hoet al., 1987; Hartman,1996; Pfeilshifteret al., 1996). Indeed, we confirmed a closerelationship between total IGF-I and free oestrogen levels inboth elderly men and women. Total oestradiol levels in ourstudy were slightly higher in elderly men than in women, whichis in accordance with the study reported by Pfeilshifteret al.(1996). Mean total IGF-I levels were comparable in both sexesin both studies. In neither study was a correlation observedbetween total IGF-I and total oestradiol levels.

The absence of a relationship between free IGF-I and free

androgen and oestrogen levels suggests that during ageing thereis no direct relationship between IGF-1 bioactivity in peripheralblood and free sex steroid levels. Such a conclusion might haveimportant consequences with regard to therapeutic inter-ventions in the elderly, which aim to increase serum IGF-Ilevels with combinations of IGF-I modifying drugs (e.g. GH,growth hormone releasing hormone (GHRH), growth hormonebinding protein (GHBPs)) and sex steroids (e.g. androgensand/or oestrogens).

In our study serum IGFBP-1 levels were positively relatedto SHBG in both men and women. The direction of theassociations between IGFBP-1 and SHBG were the mirrorimage of the observed associations between IGFBP-1 and thefree androgen and oestradiol indices. These results suggest acommon regulatory mechanism for the hepatic production ofIGFBP-1 and SHBG. Previous studies have also reported apositive correlation between SHBG and IGFBP-1 levels(Holly et al., 1989; Weaveret al., 1990; Holly et al., 1991).In cultures of human hepatoma cells both IGF-I and insulininhibit the production of SHBG and IGFBP-1 (Leeet al.,1993; Craveet al., 1995). In agreement with this, both totaland free IGF-I levels were negatively related to SHBGlevels.

It has been previously found that DHEA administrationincreases IGF-I concentrations in middle-aged and elderlyindividuals (Moraleset al., 1994). The relationships betweenfree IGF-I and DHEAS and between IGFBP-1 and DHEAS inwomen, as observed in our study, support the view that DHEASinfluences free IGF-I levels in women via an unknownmechanism. DHEAS might thus exert a direct effect on thehepatic production of IGF-1 in women. Free IGF-I may alsoinfluence DHEA production in the adrenals by its previouslydemonstrated positive effects on the number of corticotrophin(ACTH) receptors (Penhoatet al., 1994). The absence of arelationship between serum IGF-I and DHEAS in men might bethe consequence of a considerable contribution of testicularDHEA to serum DHEAS levels (Weustenet al., 1987), whichmay have masked a relationship between free IGF-I and adrenalDHEAS.

In conclusion, free IGF-I levels did not decrease with age.On the contrary, levels were higher in individuals over 70years compared with those aged 55–70 years. There was norelationship between free IGF-I and free androgen or oestrogenindices. We hypothesize that high free IGF-I levels in olderpeople may be the consequence of selective survival in thecohort: subjects with high free IGF-I levels may live longer.The absence of a relationship between free IGF-I levels and freeandrogen and oestrogen indices suggests that there is no directinteraction between the biological activity of circulating IGF-Ilevels and sex hormone production in a healthy ageingpopulation.



Acknowledgement

This study was supported by a grant from the NetherlandsDiabetes Fund.

References

Blum, W. (1996) Insulin-like growth factors and IGF-binding proteins:their use for diagnosis of growth hormone deficiency. InGrowthhormone in adults(eds A. Juul & J.O.L. Jorgensen), pp. 48–74.Cambridge University Press, Cambridge.

Bang, P., Brismar, K. & Rosenfeld, R.G. (1994) Increased proteolysisof insulin-like growth factor binding protein-3 (IGFBP-3) innoninsulin-dependent diabetes mellitus serum, with elevation of a29 kilodalton (kDa) glycosylated IGFBP-3 fragment contained in theapproximately 130–150 kDa ternary complex.Journal of ClinicalEndocrinology and Metabolism, 78, 1119–1127.

Bereket, A., Wilson, T.A., Blethen, S.L., Fan, J., Frost, R.A., Gelato,M.C. & Lang, C.H. (1996) Effect of short-term fasting on free/dissociable insulin like growth factor I concentrations in normalhuman serum.Journal of Clinical Endocrinology and Metabolism,81, 4379–4384.

Bolton, N.J., Tapanainen, J., Koivisto, M. & Vihko, R. (1989)Circulating sex hormone-binding globulin and testosterone innewborns and infants.Clinical Endocrinology, 31, 201–207.

Cohen, P., Ocrant, I., Fielder, P.J., Neely, E.K., Gargosky, S.E., Deal,C.I., Ceda, G.P., Youngman, O., Pham, H., Lamson, G., Giudice,L.C. & Rosenfeld, R.G. (1992) Insulin-like growth factors (IGFs):implications for aging.Psychoneuroendocrinology, 17, 335–342.

Copeland, K.C., Colletti, R.B., Devlin, J.T. & McAuliffe, T.I. (1990)The relationship between insulin-like growth factor-1, adiposity andaging.Metabolism, Clinical and Experimental, 39, 584–587.

Corpas, E., Harman, S.M. & Blackman, M.R. (1992) Serum IGF-1binding protein-3 is related to IGF-1, but not to spontaneous GHrelease, in healthy old men.Hormone and Metabolic Research, 24,543–545.

Corpas, E., Blackman, M.R., Roberson, R., Scholfield, D. & Harman,S.M. (1993) Oral arginine/lysine does not increase growth hormoneand insulin-like growth factor-1 secretion in old men.Journal ofGerontology, 48, M128–M133.

Cox, C., Caulier, C., Havelange, G., Meunier, J.C. & Renzi, A. (1992)Two-sites immunoradiometric assay using monoclonal antibodies forthe determination of serum human sex hormone binding globulin.Journal of Immunoassay, 13, 355–373.

Crave, J.C., Lejeune, H., Brebant, C., Baret, C. & Pugeat, M. (1995)Differential effects of insulin and insulin-like growth factor I on theproduction of plasma steroid binding globulins by human hepato-blastoma-derived (Hep G2) cells.Journal of Clinical Endocrinologyand Metabolism, 80, 1283–1289.

Daughaday, W.H. & Rotwein, P. (1989) Insulin-like growth factors Iand II. Peptide, messenger ribonucleic acid and gene structures,serum and tissue concentrations.Endocrine Reviews, 10, 68–91.

Frystyk, J., Skjærbæk, C., Dinesen, B. & Orskov, H. (1994) Freeinsulin-like growth factors (IGF-I and IGF-II) in human serum.FEBSLetters, 384,185–191.

Griffin, J.E. (1996) Assessment of endocrine function. InTextbook ofEndocrine Physiology(eds J.E. Griffin & S.R. Ojeda), pp. 86–100.Oxford University Press, New York.

Hall, K., Brismar, K., Hilding, A. & Lindgren, B. (1993) Insulin-like

growth factor-binding protein-1, a marker of insulin production.Clinical Pediatrics Endocrinology, 2 (Suppl. 2), 51–55.

Hartman, M.L. (1996) Physiological regulators of growth hormonesecretion. InGrowth Hormone in Adults(eds A. Juul & J.O.L.Jorgensen), pp. 5–35. Cambridge University Press, Cambridge.

Ho, K.Y., Evans, W.S., Blizzard, R.M., Veldhuis, J.D., Merriam, G.R.,Samojlik, E., Furlanetto, R., Rogol, A.D., Kaiser, D.L. & Thorner,M.O. (1987) Effect of sex and age on the 24-hour profile of growthhormone secretion in man: importance of endogenous oestradiolconcentrations.Journal of Clinical Endocrinology and Metabolism,64, 51–58.

Ho, K.K.Y. & Hoffman, D.M. (1993) Aging and growth hormone.Hormone Research, 40, 80–86.

Hofman, A., Grobbee, D.E., de Jong, P.T.V.M. & van den Ouweland,F.A. (1991) Determinants of disease and disability in the elderly: theRotterdam elderly study.European Journal of Epidemiology, 7,403–422.

Holly, J.M.P., Smith, C.P., Dunger, D.B., Howell, R.J.S., Chard, T.,Perry, L.A., Savage, M.O., Cianfarani, S., Rees, L.H. & Wass, J.A.H.(1989) Relationship between the fall in sex hormone binding globulinand insulin-like growth factor binding protein-1. A synchronizedapproach to pubertal development?Clinical Endocrinology, 31,277–284.

Holly, J.M.P., Cotterill, A.M., Jemmott, R.C., Shears, D., Al-Othman, S.,Chard, T. & Wass, J.A.H. (1991) Inter-relations between growthhormone, insulin, insulin-like growth factor-I (IGF-I), IGF-bindingprotein-1 (IGFBP-1) and sex hormone-binding globulin in acrome-galy. Clinical Endocrinology, 34, 275–280.

Juul, A., Flyvbjerg, A., Frystyk, J., Muller, N., Skakkebæk, N.E. (1996)Serum concentrations of free and total insulin-like growth factor-1,IGF binding proteins -1 and -3 and IGFBP-3 protease activity inboys with normal or precocious puberty.Clinical Endocrinology,44, 515–523.

Khaw, K.-T., Chir, M.B.B., Tazuke, S. & Barett-Connor, E. (1988)Cigarette smoking and levels of adrenal-androgens in post-menopausal women.New England Journal of Medicine, 318,1705–1709.

Lee, P.D.K., Suwanichkul, A., Depaolis, L.A., Snuggs, M.B., Morris,S.L. & Powell, D.R. (1993) Insulin-like growth factor (IGF)suppression of IGFBP-1 production: evidence for mediation by theIGF-1 receptor.Regulatory Peptides, 48, 199–206.

Lee, P.D.K., Powell, D., Baker, B., Liu, F., Mathew, G., Levitsky, I.,Gutierrez, O.D. & Hintz, R.L. (1994) Characterization of a direct,non-extraction immunoradiometric assay for free IGF-I.Presented atthe 76th annual meeting of the Endocrine Society, Anaheim.

Morales, A.J., Nolan, J.J., Nelson, J.C. & Yen, S.S.C. (1994) Effects ofreplacement dose of dehydroepiandrosterone in men and women ofadvancing age.Journal of Clinical Endocrinology and Metabolism,78, 1360–1367.

O’Connor, K., Stevens, T.E. & Blackman, M.R. (1996) Growthhormone and aging. InGrowth Hormone in Adults(eds A. Juul& J.O.L. Jorgensen), pp. 323–366. Cambridge University Press,Cambridge.

Orskov, H., Weeke, J., Frystyk, J., Kaal, A., Nielsen, S., Skjærbæk, C.,Christiansen, J.S., Moller, N. & Jorgensen, J.O.L. (1996) Growthhormone determination in serum from patients with growthdisorders. InGrowth Hormone in Adults(eds A. Juul & J.O.L.Jorgensen), pp. 109–121. Cambridge University Press, Cambridge.

Penhoat, A., Rainey, W.E., Viard, I. & Saez, J.M. (1994) Regulation ofadrenal cell-differentiated functions by growth factors.HormoneResearch, 42, 39–43.



Pfeilschifter, J., Scheidt-Nave, C., Leidig-Bruckner, G., Woitge, H.W.,Blum, W.F., Wuster, C., Haack, D. & Ziegler, R. (1996) Relationshipbetween circulating insulin-like growth factor components and sexhormones in a population-based sample of 50- to 80-year-old menand women.Journal of Clinical Endocrinology and Metabolism, 81,2534–2540.

Rudman, D., Feller, A.G., Nagraj, S.H., Gergans, G.A., Lalitha, P.Y.,Goldberg, A.F., Schlenker, R.A., Cohn, L., Rudman, I.W. & Mattson,D.E. (1990) Effects of human growth hormone in men over 60 yearsold. New England Journal of Medicine, 323,1–6.

Verjans, H.L., Cooke, B.A., de Jong, F.H., de Jong, C.M.M. & van

der Molen, H.J. (1973) Evaluation of a radioimmunoassay for testo-sterone estimation.Journal of Steroid Biochemistry, 4, 665–676.

Weaver, J.U., Holly, J.M.P., Kopelman, P.G., Noonan, K., Giadom,C.G., White, N., Virdee, S. & Wass, J.H. (1990) Decreased sexhormone binding globulin (SHBG) and insulin-like growth factorbinding protein (IGFBP-1) in extreme obesity.Clinical Endocrinol-ogy, 33, 415–422.

Weusten, J.J.A.M., Smals, A.G.H., Hofman, J.A., Kloppenburg, P.W.C.& Benraad, T.J. (1987) Early time sequence in pregnenolonemetabolism to testosterone in homogenates of human and rat testis.Endocrinology, 120,1909–1913.



Documents

Serum free IGF-I, total IGF-I, IGFBP-1 and IGFBP-3 levels in an elderly population: relation to age and sex steroid levels