Introduction

Polycystic ovary syndrome (PCOS) is the most common endocrine disorder in reproductive age of women with a worldwide prevalence of 5–10 % [1, 2]. The characteristics of this syndrome are hirsutism, hyperandrogenism, menstrual disturbance, and infertility. Obesity, impaired glucose tolerance, insulin resistance, endothelial dysfunction, and increased inflammation are other well-known features [3–7]. Low-grade inflammation plays a key role in the development and progression of insulin resistance, type 2 diabetes mellitus (DM), and atherosclerosis [8–10]. Observation of coexistence of PCOS with insulin resistance and atherosclerosis may be the result of common pathway of inflammation in these clinical conditions [11, 12].

In previous studies, hsCRP was reported as a commonly used laboratory parameter in the detection of subclinical inflammation in patients with PCOS [13, 14]. Additionally, increased leucocyte counts were found to be an additional independent marker and prognostic factor in the development of inflammation and atherosclerosis [4, 7, 15–17]. In recent years, neutrophil to lymphocyte ratio (NLR) has gained a popularity in the detection of inflammation in different inflammatory diseases such as PCOS, DM, ulcerative colitis, and hypertension, and it has been shown that NLR is correlated with hsCRP levels [4, 18, 19]. As a marker, NLR is a cost-effective and convenient marker that could be an alternative to hsCRP for inflammation.

Similarly, increased levels of mean platelet volume (MPV) can also be used as a marker to detect inflammation [20–25]. It is known that larger platelets are metabolically more active and prone to inflammatory cytokine release. It was shown that MPV values are higher in patients with PCOS than controls [26–28].

In this study, we aimed to compare hsCRP, MPV, and NLR levels in lean and obese PCOS patients using two different control groups matched with age and body mass index (BMI).

Patients and methods

This prospective study was conducted in Konya Training and Research Hospital, in the clinics of Internal Medicine (Division of Endocrinology) and Obstetrics and Gynecology between June and December 2014. Informed consents were obtained from all participants. Study protocol was approved by the ethics committee of Necmettin Erbakan University, Meram Faculty of Medicine.

Obese PCOS patients with BMI ≥25 kg/m2 (n = 25) and lean PCOS patients with BMI <25 kg/m2 (n = 16) were enrolled into the study. Age- and BMI-matched 16 obese and 14 non-obese healthy subjects were included into the study as control group. All patients were aged between 18 and 40 years.

Diagnosis of PCOS was based on the Rotterdam criteria as having at least two of the following three criteria: first, oligomenorrhea (cycles lasting longer than 35 days) or amenorrhea (<2 menstrual cycles in the past 6 months); second, clinical or biochemical hyperandrogenism; third, polycystic appearance of ovary on ultrasonography (USG) and the exclusion of other causes of hyperandrogenism, such as Cushing’s syndrome, congenital adrenal hyperplasia, or virilization [29]. Subjects taking drugs in the last 6 months that may interfere or affect insulin resistance and inflammation, such as estrogens, oral contraceptives, corticosteroids, immunosuppressants, anti-hyperlipidemics, anti-hypertensive, anti-hyperglycemic drugs, insulin sensitizing drugs, and anti-inflammatory drugs in the last month or with any known active infection or inflammatory diseases such as Crohn, ulcerative colitis, rheumatoid arthritis, systemic lupus erythematosus or benign or malignant hematologic disorders, hypertension, or any surgical intervention within the past 6 months were excluded from the study.

Height (m) and weight (kg) were measured with underwear clothing. Waist circumference was measured as the minimum size between iliac crest and lateral costal margin. BMI was calculated as weight (kg) divided by height square (m2).

All blood samples were drawn after an overnight fasting between the third and fifth days of menstruation, separated by centrifugation, and stored in deep freeze at −70 °C until being analyzed.

Complete blood count was measured by Sysmex XE-2100 (Sysmex Corp. Kobe, Japan) with fluorescence flow cytometry or electrical impedance method. NLR was calculated as the ratio of the neutrophil count to lymphocyte count. hsCRP levels were measured by BN™ II (Siemens Healthcare Diagnostics, Siemens AG, Germany) with nephelometric method. Glucose levels [normal range (NR), 70–105 mg/dl] were measured by Abbott C-16000 Autoanalyzer (Abbott Laboratories, Abbott Park, IL) with glucose oxidized method. Insulin levels were measured by Advia Centaur XP (Siemens Healthcare Diagnostics, Siemens AG, Germany) with chemiluminescence method. Insulin resistance was calculated by homeostasis model assessment-insulin resistance (HOMA-IR) [fasting plasma glucose (mmol/l) × fasting serum insulin (µIU/ml)/22.5]. Intra- and inter-assay coefficients of variations for insulin were 4.6 and 5.9, respectively.

Statistical analysis

The student’s t test for independent samples was used to compare normally distributed measurements. The Kolmogorov–Smirnov test was used to test normal distribution of variables. The Pearson and Spearman’s correlation analyses were performed to determine the correlation of parametric and non-parametric data, where appropriate. The Chi-square test was used to compare categorical data. Mean ± standard deviation and frequencies (%) were used to summarize parametric and categorical data, respectively. The statistical analyses of data were carried out using SPSS 21 statistical software for Windows version (IBM Corp., Armonk, NY). Statistical significance level was set to 0.05.

Results

The anthropometric, hormonal, and metabolic characteristics of patients are summarized in Table 1.

Table 1 Some anthropometric, metabolic, and hormonal data of the study population
Full size table

Before subgroup analyses of lean and obese PCOS patients, our participants were compared in two groups as all PCOS patients (n = 41) and all controls (n = 30). Compared with controls, PCOS group had significantly higher waist circumference (WC), HOMA-IR, insulin, glucose, MPV, neutrophil count, neutrophil to total leucocyte ratio, NLR, and basophil count. hsCRP levels were similar in both groups (Table 1).

In PCOS group (n = 41), correlation analyses revealed that hsCRP was positively correlated with BMI, WC, total leucocyte count, neutrophil count, and NLR, while MPV was negatively correlated with BMI, total leucocyte count, platelet count, and neutrophil count. NLR was positively correlated with HOMA-IR, hsCRP, BMI, WC, insulin, total leucocyte count, and neutrophil count while negatively correlated with lymphocyte count (Table 2).

Table 2 Correlation analyses in PCOS group and controls
Full size table

Subgroup analyses

Correlation analysis of subgroups is shown in Table 3. In the comparison of obese PCOS patients (n = 25) and controls (n = 16), obese PCOS group had higher insulin and HOMA-IR levels. The counts of total leucocyte, its subgroups, MPV, and hsCRP were not significantly different among groups. In the comparison of lean PCOS patients (n = 16) and controls (n = 14), lean PCOS group had higher WC, NLR, MPV, and basophil count. hsCRP levels and counts of total leucocyte and its subgroups except for basophil were similar.

Table 3 Some anthropometric, metabolic, and hormonal data of the obese and lean PCOS groups
Full size table

In the correlation analyses of obese PCOS group, hsCRP was positively correlated with BMI (Table 4). In lean PCOS group, hsCRP was positively correlated with BMI, WC, insulin, and HOMA-IR levels, as well as positive correlation of NLR with BMI, insulin, and HOMA-IR. In obese controls, hsCRP was positively correlated with glucose. In lean PCOS group, hsCRP was also positively correlated with BMI. In lean controls, hsCRP was negatively correlated with HOMA-IR and insulin levels.

Table 4 Correlation analyses in subgroup of patients
Full size table

According to the presence of obesity, our cases were divided into two main groups as the lean and the obese (BMI ≥25 kg/m2), each including both patients with PCOS and healthy controls. Obese group had higher age, BMI, WC, glucose, insulin, HOMA-IR, hsCRP, leucocyte, neutrophil and platelets levels, and lower lymphocyte, compared to lean group. MPV and NLR levels were found to be similar (Table 5).

Table 5 Some anthropometric, metabolic, and hormonal data of study population [cases were divided into two main groups as the lean (BMI < 25 kg/m2) and the obese (BMI ≥ 25 kg/m2)]
Full size table

Discussion

In this study, we found higher neutropil and basophil counts, NLR, MPV levels in patients with PCOS, compared to controls. In addition, in the subgroup analysis, we found that obese PCOS patients had higher insulin and HOMA-IR levels while lean PCOS group had higher WC, NLR, MPV, and basophil counts, compared to age- and BMI-matched controls.

In our study, similar hsCRP levels were found in both groups. Increased inflammatory markers such as hsCRP have been reported in patients with PCOS [14, 30]. The underlying mechanism of increased hsCRP is yet to be elucidated, and whether it is related to PCOS itself or the accompanying obesity still remains uncertain. According to the presence of obesity, we reclassified our participants into two as the lean and the obese (BMI ≥25 kg/m2), and obese group showed higher hsCRP levels than lean group. It is well known that obesity is an important factor leading to an increase of hsCRP [31]. In a study with small sampling size by Kahal et al., as consistent with our findings, it was reported that hsCRP levels in obese patients with PCOS were similar to those of controls [13]. We consider PCOS as an inflammatory disease, and obesity a contributing factor to the elevation of hsCRP; therefore, we anticipated to observe higher level of hsCRP in PCOS patients. However, because of small sample size of groups in our study, we failed to demonstrate the increase of hsCRP in our patients with PCOS. Despite this limitation in our study, we were able to demonstrate increased levels of MPV and NLR. We consider that the condition reflects higher sensitivity of these markers in PCOS patients, and that further studies are required to support these hypotheses.

In literature, there are several studies investigating MPV levels in patients with PCOS. Gursoy et al. [26] and Kebapcilar et al. [28] reported that patients with PCOS had higher MPV levels, compared to controls. In another study performed by Kabil Kucur et al. [27], MPV levels were found higher in patients with PCOS, and after the treatment with ethinyl estradiol/cyproterone acetate or metformin, a significant decrease was observed in MPV levels. It could also be speculated that anti-inflammatory effects of these treatment modalities might contribute to the protective effect on cardiovascular diseases in patients with PCOS. Additionally, platelets with higher volumes are known to be associated with increased agility in thrombotic processes. In our study, as well as lean subgroup, PCOS patients had higher MPV levels than controls. However, we found no similar results in obese subgroup, compared to obese controls. Silfeler Beng et al. [32] and Dogan et al. [33] reported similar MPV levels in patients with lean PCOS and BMI- matched controls, and our findings are different from theirs. They concluded that as to obesity, their results are more important contributors to MPV levels than the existence of PCOS per se. Interestingly, as opposed to the findings in previous studies by Coban et al. [34], we observed a significant negative correlation between BMI and MPV levels (r = −0.336, p = 0.03), which could not be reproduced in subgroup analysis. When we divided our cases into two main groups as the lean and the obese (BMI ≥25 kg/m2), MPV levels were found to be similar. Although some other contributing factors related to this finding might be present, the further discussion and comparison with the previous studies would be impossible due to the paucity of data on the subject. Therefore, we consider that this condition should be further discussed, and more comprehensive studies to enlighten the correlation between MPV, obesity, and PCOS are needed.

Increased leucocyte count is an independent risk and prognostic factor in the development of inflammation and atherosclerosis [17], and there are several studies reporting increased leucocyte counts in patients with PCOS [4, 7, 15, 16, 35]. In our study, leucocyte counts were similar in patients with PCOS and controls; still, differential count of leucocytes revealed that patients with PCOS had higher neutrophil count and ratio, basophil count and lower lymphocyte count, compared to controls. Basophil count was higher only in lean PCOS patients than controls, and no significant difference regarding to other leucocytes was present. To our knowledge, no study investigating and accounting for the increase of basophil is present, and our study is the first to indicate this increase. However, we failed to elucidate such an increase, and so further studies should be performed. In a study similarly designed to ours by Keskin Kurt et al., it was reported that higher hsCRP, leucocyte count, and NLR were present in PCOS patients [4]. Additionally, they also reported a moderate, but significant correlation among NLR and HOMA-IR, BMI, and hsCRP, reflecting the higher statistical power of their study. Considering the common inflammatory process, leucocyte count and NLR were expected to be increased along with CRP. As opposed to the findings in the study by Keskin Kurt et al., we found higher NLR with similar hsCRP levels in PCOS group compared to matched controls. However, we failed to demonstrate the same findings as those in obese PCOS patients, probably due to small sample sizes of subgroups. Nonetheless, we consider that even in case of similar hsCRP levels, NLR might be a more sensitive parameter to demonstrate ongoing subclinical inflammation in PCOS patients. In a recent study published by Papalou et al., it has been reported that higher leucocyte counts were observed in patients with PCOS than controls, and while androgenism did not seem to affect leucocyte counts, obesity and insulin resistance were major contributing factors in the development of leukocytosis [35]. In our study, although a significant association was present between leucocyte and BMI in PCOS and control groups, the association was observed between HOMA-IR and leucocyte only in PCOS group (Table 2). In an analysis investigating the association between androgens and leucocyte counts, no correlation was found except for between NLR and free testosteron (p = 0.451, r = 0.021) (unexisting data).

On the other hand, the analysis of NLR and BMI, as expected, showed a significant correlation (r = 0.351, p = 0.024, Table 2). However, this finding was not persistent in obese PCOS group, as opposed to the one found in lean PCOS patients (Table 4). We consider that this condition was a reflection of decreased power of the study in subgroup analyses. Similarly, positive correlations were observed among total leucocytes, neutrophil, and BMI. Additionally, when we divided our cases into two main groups as the lean and the obese (BMI ≥25 kg/m2), NLR levels were found to be similar.

In conclusion, we found that the patients with PCOS had higher NLR and MPV levels even when similar hsCRP levels were present. This result may be attributed to the fact that NLR and MPV levels are good indicators in the detection of inflammation as independent of hsCRP levels in patients with PCOS. On the other hand, no similar results were shown when patients were reclassified as the lean and the obese, and this may be due to small sample size in our study. New studies with larger sample size are needed.