Patients with sleep apnea commonly develop the metabolic syndrome, the hallmark of which is insulin resistance. It appears to be one of the major risk factors for atherosclerotic and cardiovascular diseases . The mechanisms by which obesity and intermittent hypoxia predispose to insulin resistance in that patient group are of interest.
Adipose tissue dysfunction that is characterized by the imbalanced secretion of different adipokines has been largely disputed in clinical studies. Adiponectin and leptin are adipose-specific hormones that improve glucose tolerance. Both have been thoroughly investigated in different studies. Several other adipose-derived hormones oppose insulin action, and they circulate at increased levels in obesity, thus also possibly contributing to insulin resistance. Prominent among them are tumor necrosis factor-α and interleukin-6, two cytokines that are also produced at high levels in macrophages and were first identified as inflammatory mediators [27, 28]. Results are rather controversial to shed light and promote the implementation of a certain biomarker neither for screening of impaired glucose tolerance in OSA, nor as a factor for stratification of patients at high risk of developing metabolic syndrome.
Resistin is a 12-kDa polypeptide that was initially linked to insulin resistance in animal models . Early reports suggested that resistin is associated with obesity and insulin resistance in rodents. Deletion of the resistin gene reduces the impact of obesity on glucose homeostasis . Conversely, acute administration of resistin impairs glucose tolerance and insulin action . Mice with chronic hyperresistinemia exhibit modest fasting hyperglycemia and glucose intolerance, associated with increased hepatic glucose production in the setting of hyperinsulinemia. These results indicate that chronic hyperresistinemia leads to impairment of glucose homeostasis .
A number of studies have examined plasma resistin levels or adipose resistin expression, and have found variable associations with insulin resistance [11, 33–35]. A recent large study involving the Framingham offspring cohort found a significant relationship between insulin resistance and resistin. This relationship however was considerably weaker than the relationship with adiponectin, and was lost after adjustment for BMI . Resistin decreases after TZD treatment of humans, although resistin was also decreased by metformin treatment . Therefore, resistin is clearly an important adipokine that likely plays a role in the development of insulin resistance.
An investigation of resistin in OSA will therefore provide a better understanding of the complex interrelation between metabolic disorders, inflammation and cardiovascular involvement.
According to the results of our study there is not a significant difference between the plasma levels of resistin in patients with impaired and normal glucose tolerance. The two groups were comparable regarding the age, BMI, waist circumference and the severity of OSA. The duration of sleep, the average desaturation index and the time spent under SpO2 <90% also did not differ much between the groups.
The analysis of resistin plasma levels on the other hand showed correlation neither to the anthropometric (age, BMI, waist circumference, smoking status), nor to the sleep study or glucometabolic characteristics. The other biomarkers, indicative of insulin resistance and oxidative stress also turned to play no role on resistin plasma levels. Assuming this data it seems that resistin alone is not a trigger of insulin resistance in obese OSA patients. It more likely plays a secondary role in the complex adipokine signaling, accompanying adipose tissue dysfunction.
Our results support those of Yamamoto et al. , who find no association between resistin plasma levels and insulin resistance, presented by the should be HOMA-I in OSA patients. Their study however should be consciously reconsidered as it is performed in Asian population, with significantly lower BMI (BMI-28). Wysozka et al.,  investigated different adipokines in Caucasian obese (BMI 30–39,9) and overweight (BMI – 25–29,9) patients with and without OSA. They conclude that in both groups – obese and overweight - OSA per se causes a decrease in resistin plasma levels.
Having this in mind and considering that each of the three groups we studied consisted of extremely obese patients with an average BMI >40 (only 10% have BMI 30–35) we can speculate that the secretion of resistin could be blunted. If true this makes it rather difficult to detect the slight difference in resistin plasma levels between OSA patients with IGT and NGM if any exists.
Wysoczka et al. , describe that in both obese groups – with and without OSA resistin correlated to an increased fasting glucose. Similar is the data, presented by Rangwala et al. . They detect that under chronic hyperresistinaemia normal weight and on normal diet mice have high fasting glucose. This they prove is due to increased hepatic gluconeogenesis. Moreover impaired fasting glucose persists in the model despite the adaptive hormonal rearrangements that take place. In our study however we can not find such a relation in none of the studied groups.
We found no association between high levels of insulin and HOMA-I and resistin neither in patients with normal, nor in those with IGT. This is in controversy to what is reported in obese patients (BMI-33) without diabetes . Again it is very probable to assume that extreme OSA per se blunts resistin secretion.
In human adipose tissue, resistin seems to be produced mainly by infiltrating macrophages .
Recent studies have shown the causative association between resistin and systemic inflammation , especially in the vascular endothelium .It is notable that plasma resistin concentrations increase with increasing inflammatory mediator levels, predicting the severity of atherosclerosis . These observations detected in the general population are also confirmed in OSA patients by Yamamoto and Lee. The data presented by Yamamato et al.,  shows that plasma resistin levels increase with the severity of OSA. This correlates best with AHI and is associated to increased inflammation, reflected by the higher concentrations of IL-6. Similar are the findings of Lee et al. . They also demonstrate that plasma resistin levels increase with the severity of OSA and that this trend correlates best to AHI. We should take in mind that both studies are performed in Asian population which deters the application of their findings in Caucasians. In pediatric OSA however where much of the confounding factors are abolished the increase in plasma resistin levels with the severity of AHI is very persuasive .
In our study we find a tendency for an increase of resistin with the severity of OSA that is not statistically significant. The plasma levels of the marker in the group with moderate apnea, compared to those with severe one are respectively – 3,92vs4,96ng/ml, not reaching statistical significance. A reason for this can be that our patients were extremely obese – the average BMI >40. In the study of Yamamoto et al. , and Lee et al. , the average BMI is 28. Even though the distinction in the levels of resistin are best remarked when comparison is made between patients with severe OSA and the control group (p<0,01). Plasma resistin levels between moderate and severe OSA even in slight degrees of obesity – (BMI – 28) do not reach statistically important difference. The influence of the intermittent hypoxia, accompanying OSA, can therefore be obscured in extremely obese patients.
The limitations of our study are that: First patients were of similar age and BMI, but extremely obese, which complicates the applicability of data to the general OSA population; Second patients were predominantly men as a gender dimorphism is present in plasma resistin [9, 43] the results can not be translated to women with OSA.
Last but not least we performed a cross-sectional study, encompassing a relatively small number of subjects with various impairments of glucose metabolism, which does not allow the establishment of any causality, but only the propositions of certain hypothesis about the role of resistin in the metabolic derangements in OSA.