The first important observation resulting from our study is that significant differences occur in the studied population of women in terms of the severity of insulin resistance evaluated using the HOMA-IR index at the time of GDM diagnosis. According to the International Diabetes Federation criteria, the HOMA-IR cut-off point to differentiate between low and high insulin resistance is 2.38, and HOMA-IR values range from 0.34 to 1.29 in the lowest quartile, whereas in the 4th quartile these values range from 2.89 to 20.39. Such differences in the HOMA-IR indicate a significant pathophysiological variation in the severity of insulin resistance at the diagnosis of GDM in the studied population.
Consistent with our results, several previous studies performed on smaller populations have demonstrated that HOMA-IR index assessed at diagnosis of GDM ranged from 1.6 to 25 [13–17].
Interestingly 1/4 of the women in our study (the first quartile), exhibited a low degree of insulin resistance, while chronic insulin resistance is present in most of the GDM women [18–20], and it progresses in the course of pregnancy.
It is also worth of noticing that the highest rate of insulin resistance (fourth quartile) occurred in 1/4 of our female patients, while most women with this type of diabetes exhibit a mild degree of insulin resistance [18, 19].
The results of our study clearly indicate a pathophysiological and clinical variation of GDM in respect to HOMA-IR values. The value of HOMA-IR index higher than >1.29 is associated with a higher value of the pre-pregnancy body mass index, fasting glucose level, insulin level, and severity of GDM. Values of HOMA-IR in the range of 1.92-2.89 are specifically associated with reduced insulin secretion, while values of >2.89 are additionally associated with a higher HbA1c level and an unchanged beta cell function, suggesting insufficient compensation for metabolic demands. It should be emphasised that there were no statistically significant differences between the HOMA-IR quartiles in terms of age, parity and time of diagnosis of GDM.
Our results are consistent with the general opinion that chronic insulin resistance lies at the core of the development of the most common form of beta cell dysfunction present in GDM . The function of the beta cells decreases around a surprisingly low value of insulin resistance – as low as >1.29. The obtained results stress the need for implementing a treatment aimed at improving tissue susceptibility to insulin in GDM women with a HOMA-IR of >1.29. It is especially important to introduce healthy lifestyle changes, as well as a proper diet based on individualized caloric needs and weight gain . It should be pointed out, that limiting excessive growth of fatty tissue may also decrease the adverse effect of adipokines on the insulin secretion  and reduce the intensity of insulin resistance of skeletal muscles .
Since metformin reduces insulin resistance, metformin treatment seems to be pathophysiologically sound in most of the GDM women, with the exception of subjects with a suspected of developing type 1 diabetes or with type 1 diabetes exhibiting itself in the course of pregnancy. Women with HOMA-IR values in the range of 1.92-2.89 exhibit a decreased beta cell function, suggesting insulin substitution, rather than the introduction of oral hypoglycemic drugs.
It is possible, based on existing evidence that starting insulin substitution in the latter subpopulation of women may delay the development of subsequent disturbances in carbohydrate metabolism [24–26] and thus may partly reduce future cardiovascular risk related to diabetes.
Women with HOMA-IR values of >2.89 are likely to benefit from metformin treatment combined with insulin therapy.
Another interesting finding in this study is the association between the degree of insulin resistance, beta cell function and the severity of GDM. Severity of GDM evaluated by starting insulin therapy and daily insulin dose, exhibits the strongest correlation with HOMA-IR index values. An increase in the HOMA index value of >2.89 is accompanied both by an increase in insulin secretion and by an increase in insulin requirement, which suggests inefficient secretion in response to insulin resistance.
Our findings are in discordance with several up to day published data performed in the small populations of Caucasian GDM women [13, 16, 17] as well as in Japanese women . In these studies minimal increase in insulin secretion assessed by HOMA B in the range of 80–130 to 170–180 [13, 16] or by OGTT- derived indices of beta cell function  fail to compensate for increased insulin resistance assessed by HOMA-IR index between 1–1.8 to 2,3 contributed to the development of gestational diabetes. In contrast, we show that basal insulin secretion decrease in the range of HOMA-IR between 1 to 2,39. On the other hand, these results my support our current suggestion that there is a major need for a treatment aimed at improving tissue insulin sensitivity at low insulin resistance index-HOMA-IR values. In another observations, beta-cell activity assessed by median of HOMA-B was equal 154 at the diagnosis of GDM  and 200  alongside with the HOMA-IR- 2.1 and 2.4 respectively.
Interestingly, most of these estimates of beta cell activity are lower than those reported in our current study. We suggest that the relatively early week of gestation at the time of diagnosis of GDM as well as a younger age of studied subjects in our data compared with most of other studies [7, 13, 16, 17, 28] are partly responsible for these differences. These discrepancies may be due to the small size of populations studied as well as may reflect the differences in diagnostic procedures, severity of glucose intolerance, or anthropometric and ethnical  characteristics. In the another study of large size population of GDM women basal insulin secretion assessed by mean of HOMA-B in whole population was equal 200 + −300 while the insulin sensitivity index derived from-OGTT was reduced by 42% versus than in women with normal glucose regulation . These results are in accordance with basal insulin secretory capacity values derived from our current study and suggest pathophysiologic heterogeneity of the studied population of GDM women with respect to this pathophysiological parameter. It is suggested that results from HOMA-B a measure of basal insulin secretory capacity should be interpreted alongside with insulin sensitivity measure by HOMA-IR .
In general, our observations seem to be in line with data which investigated insulin resistance during pregnancy and its association with insulin therapy and beta cell dysfunction during pregnancy [30–32].
It has been demonstrated in several of the presented studies that insulin use during pregnancy is both a parameter of GDM severity and a strong predictor for the development of type 2 diabetes [31–33]. Our present results suggest that the degree of insulin resistance, as assessed by the HOMA-IR index at the diagnosis of GDM, could be a potential predictor of GDM severity and of the future development of type 2 diabetes, because this parameter is associated with both beta cell dysfunction and insulin therapy.
The importance of insulin resistance for beta cell function in post-GDM has been partly illustrated by the results of the Troglitazone In Prevention Of Diabetes and Pioglitazone in Prevention of Diabetes studies, in which the protection of beta cell function was closely related to the degree of reduction in endogenous insulin requirements .
We acknowledge, that estimation of insulin sensitivity using HOMA-index is less precise than measurement with an insulin clamp, but we consider that HOMA method is more feasible in a large size cohort. Further, we have established, that HOMA would be an accurate measure of total insulin sensitivity thoroughout pregnancy in women with gestational diabetes . Additionally, results of some studies rise the pathophysiologic significance of rather chronic hepatic insulin resistance  than the whole-body insulin resistance in declining of beta cell function in GDM women during pregnancy as well as postpartum .
A limitation of our study is the use the fasting state HOMA of beta cell function in population of GDM women. There are no evidences that insulin secretion as assessed by HOMA-B index a surrogate of basal insulin secretion is an adequate or is not adequate measure of beta cell function in GDM women. Taken into consideration the results of last data which suggest the pathophysiologic importance of hepatic insulin resistance in GDM women, the evaluation of basal insulin secretion may be a valuable metabolic parameter in these group of women [36–38].
Certain limitation is the lack of information on the dietary habits and the lifestyle prior to the study. Last data revealed, that dietary factors, both before and during pregnancy as well as physical activity are associated with GDM risk .