IGF-1 Levels are Inversely Associated With Metabolic Syndrome in Obstructive Sleep Apnea

Abstract

Study Objectives:

This study examined insulin-like growth factor-1 (IGF-1) production and its association with the metabolic syndrome (MS) in men with obstructive sleep apnea (OSA).

Methods:

In total, 47 overweight and obese men who had been referred for suspected OSA underwent polysomnography and were classified based on the apnea-hypopnea index (AHI) into three groups: no OSA, < 5 events/h (n = 11); mild OSA, ≥ 5 to < 15 events/h (n = 8); and moderate-severe OSA, ≥ 15 events/h (n = 28). The assessment of the somatotropic axis function included IGF-1 measurement. MS was diagnosed according to the National Cholesterol Education Program guidelines.

Results:

IGF-1 level in the moderate-severe OSA group was lower than in the no-OSA group (156.8 ± 54.3 μg/L versus 225.5 ± 80.5 μg/L; p = 0.013). IGF-1 level was negatively correlated with body mass index, waist circumference (WC), AHI, and sleep duration with oxygen (O₂) saturation < 90% and positively correlated with the average and minimum O₂ saturation (p = 0.027). In a multivariable linear regression, considering WC and minimum O₂ saturation as independent variables, only the minimum O₂ saturation was a predictor of low IGF-1 levels. The proportions of patients with MS were different between the three groups (18.2% in no OSA; 25% in mild OSA, and 57.1% in moderate-severe OSA; p = 0.047). Furthermore, in the lowest tertile of IGF-1 value, 66.7% of patients were affected by MS (p = 0.049). Hemoglobin (Hb)A1c correlated negatively with the minimum O₂ saturation and IGF-1 levels. However, in multivariable linear regression only IGF-1 levels were a predictor of HbA1c levels.

Conclusion:

The occurrence of OSA is associated with a reduction in IGF-1 levels. IGF-1 alterations in OSA also seem to be associated with a higher prevalence of MS.

Citation:

Izumi S, Ribeiro-Filho FF, Carneiro G, Togeiro SM, Tufik S, Zanella MT. IGF-1 levels are inversely associated with metabolic syndrome in obstructive sleep apnea. J Clin Sleep Med 2016;12(4):487–493.

INTRODUCTION

Obstructive sleep apnea (OSA) has been associated with reduced secretion of growth hormone (GH) and insulin-like growth factor-1 (IGF-1).¹ However, the GH/IGF-1 axis is disrupted in OSA by mechanisms that are not yet fully understood.^2,3

Obesity has been associated with increasing OSA severity, and weight gain represents the most important risk factor for OSA.^4,5 It is also recognized that obesity results in reduced GH secretion and subnormal IGF-I levels.^6,7 The exact mechanisms responsible for the association between obesity and dysregulation of GH/IGF-1 levels have yet to be clarified.^8,9 Neuroendocrine changes in growth hormone-releasing hormone (GHRH), somatostatin and ghrelin pathways have been associated with low plasma GH levels in obesity. Hyperinsulinemia, probably by stimulating greater hypothalamic somatostatin release, can also contribute to reduced GH secretion and may affect IGF-1 levels, which are associated with insulin resistance.¹⁰ Factors derived from excess ectopic fat, including proinflammatory mediators and free fatty acids, can affect IGF-1 secretion and bioactivity independently or in concert.¹¹ Finally, some authors have suggested that OSA may mediate the negative association between obesity and the GH/IGF-1 axis.¹²

BRIEF SUMMARY

Current Knowledge/Study Rationale: OSA has been associated with reduced secretion of growth hormone (GH) and IGF-1. Moreover, a growing body of evidence suggests that OSA may contribute to the development of metabolic syndrome. This study examined IGF-1 production and its association with the MS in men with OSA.

Study Impact: The association between OSA and MS has been demonstrated in several studies, and therefore it is important to understand the influence of the GH/IGF-1 system on metabolic changes among patients with OSA. This study aims to understand the independent association of reduced IGF-1 production with increased incidence of MS in patients with OSA.

Metabolic syndrome (MS), which is a commonly used term for the clustering of obesity, insulin resistance, hypertension, and dyslipidemia, affects millions of people worldwide and is associated with an increased risk of type 2 diabetes. A growing body of evidence suggests that OSA may contribute to the development of MS and diabetes. Despite substantial evidence from studies to suggest an independent link between OSA and metabolic abnormalities, the issue still remains controversial.^13–17

One hypothesis to explain the metabolic changes found in patients with OSA is somatotropic axis dysfunction. Growth hormone deficiency (GHD) may predispose adults to the development of type 2 diabetes mellitus because it is characterized by a tendency toward obesity with visceral adiposity.¹⁸ The structural homology of IGF-1 with insulin also appears to play a key role in glucose homeostasis in these patients.¹⁹ Reduction in IGF-1 level leads to reduced peripheral glucose uptake and increased hepatic glucose production, thus inducing insulin resistance.^20–22

Some investigators have evaluated the relationship between OSA and MS; however, the mechanisms have yet to be clarified.^13–17 Moreover, studies suggest that OSA may contribute to reduced secretion of GH and IGF-1,^12,23,24 which could be the mechanism involved in the association between OSA and MS. The aim of the current study was to examine IGF-1 production and its association with the MS in men with OSA.

METHODS

Population

Between January 2013 and November 2014, 47 men age 18– 60 y with a body mass index (BMI) of 25 to 45 kg/m² were referred for polysomnography (PSG) for suspected OSA and selected for the study. Patients were classified according to their apnea-hypopnea index (AHI) into three groups: No OSA (AHI < 5 events/h), (n = 11); mild OSA (AHI between 5–14.9 events/h), (n = 8); or moderate to severe OSA (AHI ≥ 15 events/h), (n = 28).

Exclusion criteria included a history of smoking, sleep apnea treatment, cardiovascular and pulmonary diseases, malignant tumors, untreated thyroid disease, severe depression, diabetes mellitus, metformin therapy, chronic kidney disease, liver failure, and changes in antihypertensive class drugs in the past 3 mo. Because there are differences in OSA severity between men and women, only men were considered for evaluation.²⁵

This study was approved by the Ethics Committee of the Universidade Federal de São Paulo, and written informed consent was obtained from all of the participants.

Clinical, Anthropometric, and Biochemical Parameters

A questionnaire was used to document patient personal and medical history as well as drug therapy. A physical examination was performed, and anthropometric measurements including weight (in kilograms), height (in meters) and waist circumference (in centimeters) were recorded. Waist circumference was measured at the midpoint between the last rib and the iliac crest at the end of a normal expiration.²⁶ BMI was calculated by dividing weight by height squared.

Waking blood pressure was measured between 08:00 and 11:00 in the supine position after a 5-min rest period and was recorded as the mean of three measurements taken at 1-min intervals.

Blood specimens were obtained after a 12-h overnight fast from all of the subjects for the measurement of plasma glucose, total cholesterol, triglycerides, high-density lipoprotein (HDL), very-low-density lipoprotein (VLDL), and liver function tests. A standard 75-g oral glucose tolerance test (OGTT) was performed by measuring fasting and 2-h plasma glucose levels. IGF-1 levels were measured at the first visit. Single blood samples were drawn between 08:00 and 09:00 following the sleep period. Serum samples were stored at –70°C until assessment.

Polysomnographic Parameters

PSG was performed using a digital system (EMBLA S7000, Embla Systems, Inc, Broomfield, CO, USA). The following physiological variables were monitored simultaneously and continuously: four channels for the electroencephalogram (EEG); two channels for the electrooculogram; four channels for the surface electromyogram (submental region, anterior tibialis muscle, masseter region, and seventh intercostal space); one channel for an electrocardiogram; airflow detection via two channels through a thermocoupler (one channel); nasal pressure (one channel); respiratory effort of the thorax (one channel) and of the abdomen (one channel) using inductance plethysmography; snoring (one channel) and body position (one channel); oxy-hemoglobin (Hb) saturation (SpO₂); and pulse rate. All of the sleep studies were scored in a blinded fashion. Trained technicians recorded the polysomnographic data according to standardized criteria for sleep investigation.²⁷ EEG arousals were scored according to the criteria established by The American Academy of Sleep Medicine Manual for Scoring Sleep and Associated Events.²⁸ Analysis of SpO₂ was performed to obtain a profile of average and minimum oxygen saturation as well as the percentage of total sleep time, during which SpO₂ was < 90%.

Apneas were considered when there was a decrease in the amplitude of airflow ≥ 90% from baseline lasting ≥ 10 sec. An obstructive apnea was classified by continued/increased inspiratory effort during the event and a central apnea was characterized by absent inspiratory effort during the event. Only patients with obstructive apnea were selected. Hypopneas were recorded when there was a 30% reduction in the amplitude of airflow lasting ≥ 10 sec followed by a ≥ 3% decrease in arterial oxygen saturation or an EEG arousal.²⁸ The number of apnea and hypopnea episodes per hour of sleep (AHI) was calculated.

Laboratory Analyses

Plasma glucose, total cholesterol, triglyceride, HDL-cholesterol, aspartate aminotransferase (AST), and alanine amino-transferase (ALT) levels were measured using an ADVIA 2400 Chemistry System (Siemens, Tarrytown, NY, USA). Glucose was analyzed by hexokinase. The detection limit was 0 mg/dL, and the average within-assay coefficient of variation was 0.7%. Total cholesterol, HDL-cholesterol, and triglycerides were analyzed by enzymatic colorimetric methods. The detection limits were: cholesterol 10 mg/dL, HDL-cholesterol 20 mg/dL, triglycerides 0 mg/dL, and the average within-assay coefficients of variation were: cholesterol 0.6%, HDL-cholesterol 0.4%, triglycerides 0.6%. Low-density lipoprotein (LDL) levels were calculated using the Friedewald formula.²⁹ AST and ALT were analyzed according to the International Federation of Clinical Chemistry (IFCC) method (detection limits: AST 0 UI/L, ALT 0 UI/L, and average within-assay coefficient of variation: AST 2.3%, ALT 2%). HbA1c levels were measured by high performance liquid chromatography (HPLC) (2.2 Tosoh Plus A1C, Tosoh Corporation, Tokyo, Japan).

Serum IGF-1 levels were measured using a chemilumines-cent enzyme immunoassay system (DiaSorin, Saluggia, Italy) and a commercially available kit (LIAISON IGF-I, Saluggia, Italy). Assay sensitivity was 3 μg/L. The interassay and intra-assay coefficients of variation were 5.6–9.6% and 3.0–5.1%, respectively. Samples were assayed in a single large batch, and quality assessment samples were well within the manufacturers' defined ranges. All of the biochemical assays were performed in the Sleep Institute Laboratory.

Metabolic Syndrome

MS was diagnosed according to the National Cholesterol Education Program (NCEP) guidelines.³⁰ Patients had MS if they had three or more of the following risk factors: waist circumference > 102 cm, triglycerides ≥ 150 mg/dL, HDL cholesterol < 40 mg/dL, blood pressure ≥ 130/85 mmHg or current use of antihypertensive drugs, and fasting plasma glucose level ≥ 100 mg/dL.

Statistical Analysis

The data are expressed as the mean (standard deviation) or the median (interquartile range). Clinical, anthropometrical, biochemical, and polysomnographic characteristics were compared between no OSA, mild OSA, and moderate-severe OSA groups using analysis of variance or the Kruskal-Wallis test. To assess the differences between categorical variables, we used chi-square statistics. Correlations between variables of interest were determined using the linear correlation test or the Spearman test. Multivariate linear regression analysis was used to identify associations between polysomnographic, anthropometric, and metabolic parameters with IGF-1 and HbA1c levels. IGF-1 levels were divided into tertiles, and the prevalence of MS in the lowest tertile (tertile 1) was compared with the prevalence in the highest tertile (tertile 3). Continuous variables were compared between tertile 1 and tertile 3 using unpaired Student t-test or the Mann-Whitney U test. A value of p < 0.05 was considered to be statistically significant. Data analysis was performed using Statistical Package for Social Sciences for Windows version 20.0 (SPSS Inc, Chicago, IL).

RESULTS

The general, laboratory, and polysomnographic characteristics of all subjects are summarized in Table 1. There were no significant differences in age, BMI, waist circumference (WC), and blood pressure between the three groups. However, the moderate-severe OSA group had marginally lower HDL-cholesterol levels than the mild OSA group. As expected, significant differences in the following polysomnographic parameters were observed: AHI, arousal index per hour of sleep, duration of stages 1 and 3 sleep, average and minimum oxygen saturation, and sleep period with SpO₂ < 90% (Table 2).

Table 1.

Clinical, anthropometrical, and biochemical characteristics of all subjects.

Table 2.

Polysomnographic characteristics of all subjects.

IGF-1 levels in the moderate-severe OSA group (156.8 ± 54.3 μg/L) were significantly lower than in the no-OSA group (225.5 ± 80.5 μg/L; p = 0.013). IGF-1 values were significantly negatively correlated with BMI (r = −0.340; p = 0.019), WC (r = −0.291; p = 0.047), AHI (r = −0.312; p = 0.033) and with the sleep period with SpO₂ < 90% (r = −0.308; p = 0.035) whereas a positive and significant correlation was observed with average (r = 0.329; p = 0.024) and minimum oxygen saturation (r = 0.323; p = 0.027) (Figure 1A). There was no significant correlation between IGF-1 levels and stage 3 sleep duration (p = 0.254).

Figure 1. Insulin-like growth factor levels.

Correlation between insulin-like growth factor-1 levels and (A) minimum oxygen saturation, (B) hemoglobin (Hb)A1c levels.

The association of central obesity and OSA with IGF-1 levels was assessed. In a subsequent multivariable linear regression (Table 3) in which WC and minimum oxygen saturation were the independent variables, and IGF-1 level was the dependent variable, only the minimum oxygen saturation was associated with lower IGF-1 levels.

Table 3.

Linear regression analysis of insulin-like growth factor-1 level predictors.

HbA1c levels in the moderate-severe OSA group (5.9 ± 0.4%) were significantly higher than in the no OSA group (5.5 ± 0.3%; p = 0.033) (Table 1). Moreover, HbA1c levels negatively correlated with the minimum oxygen saturation (r = −0.306; p = 0.025) and IGF-1 levels (r = −0.408, p = 0.004) (Figure 1B). However, in a subsequent multivariable linear regression in which IGF-1, WC, and minimum oxygen saturation were the independent variables, and HbA1c was the dependent variable; only IGF-1 levels was associated with HbA1c levels (Table 4).

Table 4.

Linear regression analysis of hemoglobin A1c predictors.

The proportions of patients with MS were different between the three patients groups (18.2% in no OSA; 25% in mild OSA and 57.1% in moderate-severe OSA; p = 0.047). Furthermore, in the whole group, there was an association between IGF-1 levels and MS. In the lowest tertile of IGF-1 values (IGF-1 ≤ 148.0 μg/L), 66.7% of patients were affected by MS compared with 31.2% in the subgroup of IGF-1 levels ≥ 203.0 μg/L (p = 0.049) (Table 5).

Table 5.

Clinical, anthropometrical, and biochemical characteristics of all subjects according to insulin-like growth factor-1 tertile.

The general, laboratory, and polysomnographic characteristics of the subgroups with the lowest tertile and the highest tertile of IGF-1 values are summarized in Tables 5 and 6. The subgroup with the lowest tertile of IGF-1 values had higher BMI, WC, HbA1c, AHI, and sleep period with SpO₂ < 90%, whereas HDL-cholesterol and average and minimum oxygen saturation were lower than in the subgroup with the highest IGF-1 levels.

Table 6.

Polysomnographic characteristics of all subjects according to insulin-like growth factor-1 tertile.

DISCUSSION

This study demonstrated that hypoxemia induced by OSA is associated with reduced IGF-1 levels and reduced IGF-1 is in turn associated with higher HbA1c values regardless of the severity of obesity.

GH secretion is strongly related to sleep and is mainly released during slow wave sleep.³¹ Thus, individuals with sleep apnea, who demonstrate polysomnographic reduction of slow wave sleep, may also display alterations in GH secretion. However, studies on the somatotropic axis in patients with OSA has revealed mixed results, possibly reflecting different methodologies used or patients with different degrees of OSA.^12,23,24

Our study demonstrated that somatotropic dysfunction in the moderate-severe OSA group was characterized by low IGF-1 levels compared to the no- OSA group. Reduced plasma IGF-1 levels in OSA has also been demonstrated in previous studies.^24,32 Intermittent hypoxemia seen in OSA has been associated with decreased IGF-1 levels.³³ Previous animal studies already demonstrated that chronic hypoxemia may decrease IGF-1 levels.² Consistent with this finding, in experiments with rats, hyperoxia increased IGF-1 and GH receptor expression.³⁴ Additionally, Ursavas et al.¹² demonstrated that the low- circulating IGF-1 levels in patients with OSA were correlated with several PSG parameters, such as AHI, arousal index, average desaturation, and oxygen desaturation index. However, Grunstein et al.³³ highlighted the role of hypoxemia, which was a more important contributor than sleep fragmentation in the pathogenesis of this hormonal change. Our findings are in line with the prior studies that have suggested an important role for hypoxemia for reduction in IGF-1 level in patients with OSA.

Although several studies have reported a significant association between obesity and the reduced GH and IGF-1 levels, the exact mechanisms remain unclear.^8–11 Importantly, many prior studies exploring the relationship between obesity and somatotropic dysfunction did not account for the presence and severity of OSA. A few investigators have indeed shown an important association between OSA severity and IGF-1 levels independent of obesity.^12,24 In fact, in our study, although low serum IGF-1 level was also associated with various measures of adiposity (i.e., BMI and WC), the association seemed to be more dependent on OSA-induced hypoxemia.

As in our study, the association between OSA and metabolic syndrome has been demonstrated in several studies and seems to be responsible for many of the features of this syndrome. Men with OSA have a higher prevalence of both insulin resistance and MS than men with a similar BMI but without OSA.^13–17

Somatotropic dysfunction may partially explain the higher prevalence of MS in these patients with OSA. Verhelst et al.³⁵ demonstrated that MS is highly prevalent in hypopituitary patients with adult-onset GH deficiency. Consistently, MS in patients with GH deficiency is partially reversed by GH replacement therapy.³⁶ GH-deficient patients have increased central adiposity, reduced lean body mass, and impaired insulin sensitivity. In our study, 66.7% of the individuals in the lowest tertile of IGF-1 levels (< 148.0 μg/L) were also affected by MS compared with 31.2% of those in the highest tertile of IGF-1 levels.

Studies suggest that IGF-1 decreases insulin secretion and increases glucose disposal, and thus plays an important role in glucose metabolism.³⁷ Insulin and IGF-1 have significant homology and interact with differing affinity on the same receptors. Thus, insulin resistance, which is a component of MS, may partially result from low IGF-1 levels. This explains why the higher HbA1c values in the moderate-severe OSA group correlated with IGF-1 levels in our study. Some investigators also reported that low circulating IGF-1 levels are independently associated with hyperglycemia in adults.^38–40 Reduced levels of GH and/or IGF-1 in patients with OSA may be an important mechanism responsible for the higher prevalence and incidence of type 2 diabetes seen in patients with OSA.

Our study has some limitations including a small sample size and the fact that the study results cannot be generalized to women. The small sample size limits our ability to adjust for some important confounders in our statistical models. Last, the cross-sectional nature of the study does not address the direction of causality. Indeed, only rigorously designed intervention studies will provide causal evidence and insights into mechanisms by which OSA may lead to somatotropic dysfunction.

Despite the aforementioned limitations, our results indicate that the of OSA severity, particularly hypoxemia, is associated with reduced in IGF-1 levels. Furthermore, alterations in IGF-1 levels in OSA seem to be associated with higher prevalence of MS. Strengths of our study include measurement of fasting and postprandial glucose levels as well as objective assessment of sleep and OSA by using in-laboratory PSG. Undoubtedly, further studies are needed to clarify the mechanisms that are involved in the GH/IGF-1 axis alteration in patients with OSA as well as to evaluate the effect of positive airway pressure on neuroendocrine dysregulation.

DISCLOSURE STATEMENT

This was not an industry supported study. The authors have indicated no financial conflicts of interest.

ABBREVIATIONS

AHI

apnea-hypopnea index

ALT

alanine aminotransferase

AST

aspartate aminotransferase

BMI

body mass index

EEG

electroencephalogram

GH

growth hormone

GHD

growth hormone deficiency

GHRH

growth hormone-releasing hormone

Hb

hemoglobin

HDL

high-density lipoprotein

HPLC

high performance liquid chromatography

IFCC

international federation of clinical chemistry

IGF-1

insulin-like growth factor-1

LDL

low-density lipoprotein

MS

metabolic syndrome

NCEP

national cholesterol education program

OGTT

oral glucose tolerance test

OSA

obstructive sleep apnea

O₂

oxygen

PSG

polysomnography

SpO₂

oxy-hemoglobin saturation

VLDL

very-low-density lipoprotein

WC

waist circumference