Introduction

Menopause, the permanent cessation of menses, retrospectively refers to the passage of 12 months from the last menstrual period and occurs with a mean age of 52 years [1]. This physiologic process reflects the discharge of ovarian follicles, which causes a lack of estrogen and can be associated with signs and symptoms such as hot flashes (HFs), vaginal dryness, urogenital symptoms, sleep disorders, and depression. HFs are the most critical symptom and the main reason to refer menopausal women for care [1,2]. The prevalence of this annoying symptom in late menopausal transition and early postmenopause reaches 60–80% [3]. HFs are a sudden sensation of heat in the face and chest lasting 2–4 minutes and often accompany perspiration, irritability, and palpitations [2]. The Study of Women’s health Across the Nation (SWAN 2012) showed that day and night HFs were strongly associated with decreased quality of life [4].

Menopause, through declining endogenous estrogen and causing HFs, may potentially contribute to dyslipidemia, according to some experts. This approach can be reciprocal [5]. The results of the SWAN study showed an inauspicious relationship between HFs and lipid profiles (LPs) [6]. Some other studies on menopausal women, however, did not reveal such a relationship [7]. Changes in LPs directly or indirectly (by affecting oxidative stress [OS]) can affect the inflammatory process. This event may be in terms of the increased abdominal fat accumulation in postmenopausal women and therefore increased abdominal adipose tissue. This process can increase some systemic inflammatory markers (SIMs) by reducing estrogen and, consequently, causing HFs [5].

HFs, dyslipidemia/OS, some SIMs, and their interaction can play a part in the pathogenesis of endothelial dysfunction, hypertension, diabetes, and cardiovascular disorders (CVD) [5]. These ailments enhance the cost of treatment and health care [8]. Moreover, there is an increasing tendency to investigate the association of increased SIMs and pro-inflammatory cytokines (hormone-like glycoproteins produced by the immune system cells [9]) with HFs. Several studies in this field reported contradictory results [10–13]. Tumor necrosis factor-alpha (TNF-α), Interleukin-6 (IL-6), and Interleukin-1 (IL-1) are key pro-inflammatory cytokines [14]. Their association with menopause and its outcomes has drawn more consideration from researchers. Conversely, some of these cytokines, such as interleukin-17 (IL-17), have been studied less in this field. IL-17, produced by T helper 17 (Th17) lymphocytes, can produce IL-6 by affecting endothelial cells, which may successively result in atherosclerosis, thrombosis, and CVD [15]. Menopause appears to activate IL-17 receptors on the surface of epithelial cells by diminishing estrogen [16]. Some investigations revealed an association between elevated IL-17 and decreased estrogen in menopausal women [16,17]. Others, however, did not notice such a relationship [14]. Besides, as per searching in databases, just a single study (Huang 2017) assessed the association of IL-17 with HFs and observed no significant relationship between them [10].

Researchers also consider C-reactive protein (CRP) as an indicator of systemic inflammation. Increased CRP, particularly in women, is one of the non-lipid CVD markers [7,18]. Assessing the relationship of this marker with menopause and its outcomes revealed conflicting results. Experts agreed to do more research in this field [7,13,19]. Neutrophil-to-Lymphocyte Ratio (NLR) and Lymphocyte-to-Monocyte Ratio (LMR) are also SIMs measures that are becoming increasingly popular with researchers due to their easier accessibility and lower cost. Some studies observed differences in the composition of blood leukocytes between women before and after menopause. In these studies, women over 50 (due to lower estrogen levels) showed lower neutrophil counts and higher lymphocyte counts (and lower NLR) than younger women [20]. Besides, increasing IL-17 appears to elevate neutrophil production [21–23]. Thus, there might be a controversy among the recent findings. Moreover, it makes sense to assess the relationship of NLR with HFs in menopausal women and observe changes in neutrophil count and NLR among them. Considering the alterations in hematopoiesis with decreasing estrogen levels in menopause [20], some studies also observed an increase in LMR and lymphocyte count in perimenopausal and postmenopausal women [24].

Regarding the interaction of SIMs, LPs, HFs, and the pivotal impact of SIMs and LPs on endothelial dysfunction and its outcomes, HFs might also act as a potential risk factor for these disorders. This study aimed to determine the association between some SIMs, LPs, and HFs in healthy menopausal women. We hypothesized that there was some discrepancy between SIMs and LPs levels in the two groups of menopausal women without and with HFs.

Materials and methods

Participants and study design

In this cross-sectional study, we compared the serum levels of IL-17, high-sensitivity CRP (hs-CRP), NLR, LMR, and LPs in menopausal women without and with HFs. The Ethics Committee of Guilan University of Medical Sciences, Rasht, Iran, approved the review (Ethics Code: IR.GUMS.REC.1399.636). Using the results asserted by Huang et al. (2017), we determined the sample size. We used the mean (SD) of IL-17, declared in women with different HFs severity, and considering the expected clinical difference (1 unit). Therefore, 40 subjects in the group without HFs and 120 subjects in the group with HFs would be necessary, concerning a confidence of 95%, a power of 80%, and an alpha value of 0.05 [10].

Using consecutive sampling, we assigned eligible women from healthy menopausal women referred to one governmental teaching hospital, Guilan/Rasht, Iran, from April to September 2021 for health management or as a companion to women admitted to the obstetrics and gynecology wards. Inclusion criteria included age 40–60 years; having natural menopause; lapse of at least 12 months since the last menstruation, according to the Stages of Reproductive Aging Workshop + 10 (STRAW+10) criteria [25]; no history of any disease, including diabetes, hypertension, cardiovascular, respiratory, infectious, liver, kidney, and thyroid diseases; not taking any medication, including anti-hypertensive and lipid-lowering agents, immunosuppressive drugs, anti-inflammatory drugs, anti-oxidant supplements, neuropsychiatric drugs, and hormone replacement therapy; not consuming tobacco products and alcohol; BMI of 18.5–30 kg/m2. Exclusion criteria included unwillingness to cooperate during the study for any reason.

Based on Fig 1, the first author (ND) individually interviewed a total of 1987 women in a consecutive manner, according to the inclusion criteria. As the sampling proceeded, she excluded 1816 women from the study because they did not meet the inclusion criteria. Finally, she collected 171 eligible women, out of which 11 were unwilling to continue the cooperation. Therefore, a total of 160 women were enrolled in the study. According to their response to HFs status, the participating women were divided into two groups without HFs (N = 40) and with HFs (N = 120). The women with HFs, who had experienced HFs during the three months before enrollment, were given a card and asked to record the number of HFs they experienced over a two-week period as part of a pilot study. According to the modified Kupperman index, the severity of HFs in each study sample was determined by calculating mean HFs (severity 1: <3 HFs/day, severity 2: 3–9 HFs/day, severity 3: ≥10 HFs/day) [26].

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Recruitment of the study samples to determine the association of SIMs and LPs with HFs.

https://doi.org/10.1371/journal.pone.0291804.g001

We fully informed all recruited women about the study goals and the voluntary nature of their participation. Then, all of them signed a written informed consent form. Through face-to-face interviews, the demographic and reproductive information of the participating women was recorded in the related form, the face validity of which was approved by ten faculty members specialized in midwifery and reproductive health. Then, their height (cm) and weight (kg) were measured with minimum clothing and without shoes, and the body mass index (BMI) (kg/m2) was obtained by dividing weight by height squared. Moreover, their waist circumference (WC) and hip circumference (HC) were measured with a minimum of clothing using a tape measure in centimeters. WC was obtained by measuring the rim of the abdomen at the midpoint of the border between the last rib and the iliac crest without applying pressure to the soft tissues. Also, HC was obtained by measuring the largest circumference at the level of the buttocks. WC/HC was calculated by dividing WC by HC.

Results

According to the inclusion criteria, a total of 160 menopausal women were recruited in this study. Concerning their HFs status, they were divided into one of two groups of menopausal women without HFs (N = 40) or with HFs (N = 120). Menopausal women with HFs comprised three groups of severity values of 1 (N = 42), 2 (N = 43), and 3 (N = 35) based on their HFs severity. The mean (SD) of the age of the eligible menopausal women was 54.64 (4.26) years, and the mean (SD) of the menopausal age was 48.66 (3.73) years. Table 1 presents the demographic and reproductive characteristics of the participants. Given the baseline characteristics, the two groups of participants without and with HFs were not significantly different. Based on Table 2, the mean age of the participants in severity values of 1, 2, and 3 were 55.85 (4.22), 54.39 (4.56), and 52.31 (3.70) years, respectively. Also, the mean elapsed time since the last menstruation and the mean elapsed time since the onset of HFs decreased from severity 1 to 3. The HFs severity-based groups were not significantly different, considering the baseline characteristics.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Comparison of baseline characteristics in menopausal women without and with HFs.

https://doi.org/10.1371/journal.pone.0291804.t001

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Comparison of baseline characteristics in menopausal women in HFs severity-based groups.

https://doi.org/10.1371/journal.pone.0291804.t002

Association of SIMs with HFs and their severity

According to the Mann-Whitney U test, the two groups of menopausal women without and with HFs were not significantly different in terms of IL-17, hs-CRP, NLR, and LMR levels (Table 3).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Comparison of SIMs, and LPs in menopausal women without and with HFs.

https://doi.org/10.1371/journal.pone.0291804.t003

HFs severity-based groups were not significantly different in terms of the SIMs, apart from NLR, as indicated by Table 4. The median (IQR) of NLR significantly expanded in HFs severity values of 1, 2, and 3 (1.25, 1.57, 1.71 respectively; P = 0.013) [S1 Fig]. Given the pairwise comparison, the two groups with severity values of 1 and 2 were significantly different in terms of the median (IQR) of NLR (P = 0.011), according to Table 5. After Bonferroni adjustments for multiple comparisons, this difference was no more significant (P = 0.068). Moreover, the group with a severity value of 3 showed a higher median (IQR) of NLR compared with the group with a severity value of 1 (P = 0.003), which was still higher using the Bonferroni test (P = 0.016). After controlling age, marital duration, menopausal age, and elapsed time since the last menstruation, there was a significant association between NLR and HFs. In other words, NLR had an increasing effect on HFs severity after controlling the confounding upshots of demographic and reproductive variables (P = 0.005, B = 0.779, Odds Ratio = 2.180, 95%CI: 1.270–3.744). Furthermore, we observed the increasing effect of White Blood Cells (WBC) and WC/HC on HFs (Table 6).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Comparison of SIMs, and LPs in menopausal women in HFs severity-based groups.

https://doi.org/10.1371/journal.pone.0291804.t004

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. The pairwise comparison of NLR, TC and TG in HFs severity-based groups.

https://doi.org/10.1371/journal.pone.0291804.t005

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Results of the logistic regression (backward stepwise) and the ordinal regression comparing SIMs and LPs between two groups without HFs and with HFs, and in HFs severity-based groups.

https://doi.org/10.1371/journal.pone.0291804.t006

SIMs correlation with LPs in the group of menopausal women with HFs

In the group of menopausal women with HFs, lymphocyte count positively correlated with TG (P = 0.007, r = 0.243), hs-CRP negatively correlated with TG (P = 0.039, r = -0.189), hs-CRP also negatively correlated with TC/HDL-C (P = 0.003, r = -0.268), using Spearman correlation analysis. Other SIMs and LPs were not significantly correlated (Table 7).

Download:

• PNG
  larger image
• TIFF
  original image

Table 7. Results of Spearman correlation between SIMs and LPs in menopausal women with HFs.

https://doi.org/10.1371/journal.pone.0291804.t007

Prediction levels of TG and NLR for HFs incidence

According to the ROC curve, TG and NLR predictive levels to determine HFs incidence were statistically significant (P<0.001, AUC = 0.694, 95%CI:0.600–0.788; P = 0.002, AUC = 0.676, 95%CI:0.578–0.773, respectively) in menopausal women with HFs (Table 8). Therefore, by analyzing the sensitivity and specificity of different points of TG and NLR, we noticed the best cut-off points for TG and NLR as per the following: 143 mg/dl, with 61.5% sensitivity and 60% specificity for TG, and 1.36, with 65.4% sensitivity and 60% specificity for NLR (Fig 2).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. ROC curve of TG (mg/dl) and NLR predictive levels to determine the incidence of HFs.

https://doi.org/10.1371/journal.pone.0291804.g002

Download:

• PNG
  larger image
• TIFF
  original image

Table 8. ROC curve of TG and NLR predictive levels to determine incidence, severity 1, severity 2, and severity 3 of HFs.

https://doi.org/10.1371/journal.pone.0291804.t008

Regarding the predictive levels of TG and NLR, we noted no significant difference in determining the prognosis of the HFs severity value of 1 (Fig 3). However, these predictive levels to assess the prediction of the HFs severity value of 2 were statistically significant (P<0.001, AUC = 0.681, 95%CI:0.598–0.763; P = 0.003, AUC = 0.634, 95%CI:0.548–0.720, respectively) [Table 8]. Similarly, by examining the sensitivity and specificity of different points of TG and NLR in the HFs severity value of 2, we determined the best cut-off points for TG and NLR in this group as below mentioned: 148 mg/dl with 60.3% sensitivity and 60% specificity for TG and 1.41 with 61.5% sensitivity and 56% specificity for NLR (Fig 4).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. ROC curve of TG (mg/dl) and NLR predictive levels to determine the prognosis of HFs severity 1.

https://doi.org/10.1371/journal.pone.0291804.g003

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. ROC curve of TG (mg/dl) and NLR predictive levels to determine the prognosis of HFs severity 2.

https://doi.org/10.1371/journal.pone.0291804.g004

Likewise, the predictive levels of TG and NLR were significant at the assessment of the HFs severity value of 3 (P <0.001, AUC = 0.733, 95%CI:0.629–0.836; P = 0.037, AUC = 0.616, 95%CI:0.509–0.723, respectively) [Table 8]. Therefore, by examining the sensitivity and specificity of different points of TG and NLR at the HFs severity value of 3, we noticed that the best cut-off points for TG and NLR were as follows: 188.5 mg/dl with 71.4% sensitivity and 72% specificity for TG, and 1.45 with 62.9% sensitivity and 57% specificity for NLR (Fig 5).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. ROC curve of TG (mg/dl) and NLR predictive levels to determine the prognosis of HFs severity 3.

https://doi.org/10.1371/journal.pone.0291804.g005

Discussion

The present study aimed to determine the association of some SIMs, LPs, and HFs in healthy menopausal women. Thus, we assessed the main obtained results. Menopause may activate IL-17 receptors on the surface of epithelial cells by lowering estrogen levels [16]. IL-17 can induce IL-6 and Matrix Metallo Proteinase (MMP) release by affecting endothelial cells, which results in reperfusion damage, thrombosis, and atherosclerosis [15]. Molnar and Waliullah et al. observed diminished estradiol and raised IL-17 levels in menopausal women [16,17]. Moreover, women with HFs may have lower levels of estradiol than women without HFs, according to Gast et al. [27]. Accordingly, we expected menopausal women with HFs to have lower estradiol and higher IL-17 levels than menopausal women without HFs. We actually detected no significant association between IL-17 and HFs. However, compared to menopausal women without HFs, those with HFs tended to have lesser estradiol and increased IL-17 levels. In this study, we planned to gauge the serum IL-17 levels of menopausal women, while a simple expansion in circulating IL-17 may not be the reaction to estradiol depletion. Likewise, some studies noticed a rise in this cytokine in the synovial fluid of patients with rheumatoid arthritis [28]. Given the design of the study, we did not perform invasive procedures. Evaluating the local levels of IL-17 may lead to a better understanding of this issue. As per searching databases, only one study (Huang 2017) assessed the association of circulating IL-17 with HFs and uncovered no significant association [10]. Thus, it makes sense to do more research in this field.

Concerning the univariate analysis, the SWAN study ascertained a statistically significant association between hs-CRP and HFs, which was no more significant after controlling covariates. This point was probably a matter of shared risk factors, particularly overweight/obesity, the researchers stated [13]. Menopause can increase fat and accumulate abdominal adipose tissue in women directly or by declining estradiol, which may result in SIMs increase, including CRP, and thus, HFs occurrence [9]. Moreover, Tuomikoski et al. observed no association between hs-CRP changes and HFs [7]. Furthermore, CRP levels were not associated with total menopause rating scale (MRS) symptom score, Kaya et al. realized [29]. Since estradiol can be involved in CRP production in the liver [30], we determined estradiol as a likely confounding variable. Therefore, we compared estradiol levels in menopausal women without and with HFs to ensure their homogeneity. We observed no significant association between hs-CRP alterations and the HFs incidence and severity. These findings were in line with the SWAN study. Also, given the prevalence of overweight women in Guilan province, Iran, we measured BMI and WC/HC in both menopausal women without and with HFs to confirm their homogeneity in terms of these variables. In total, CRP may not have a simple increasing impact on HFs and further studies are needed in this field.

Chen et al. realized a drop in NLR after a decrease in neutrophil count and an increase in lymphocyte count in menopausal women. They ascribed this alteration to menopause-induced estradiol deficiency [20]. As stated before, some studies noted a connection between declined estradiol and extended IL-17 levels in menopausal women [16,17]. Tahmasebinia et al. also reported raised IL-17 levels associated with increased neutrophil counts [23]. Thus, as IL-17 enhanced, we expected neutrophil count and NLR to increase and estradiol to decrease. We found no significant association between NLR and HFs. However, we observed a tendency to expand IL-17 and increase the neutrophil count and NLR in menopausal women with HFs compared to those without HFs. Even though not significant, this finding was consistent with the pathophysiology of inflammation and dyslipidemia association with HFs [9]. Conversely, it contradicts Chen’s data (dropped counts of neutrophils and NLR following decreased estradiol levels) [20]. We also observed that NLR significantly increased as the HFs severity extended. Adjustment for the covariates still retained this association. Furthermore, Lee et al. noted increased LMR in menopause and considered it would be associated with decreased estradiol levels [24]. We observed women with HFs tended to expand LMR levels compared to those without HFs. However not statistically significant, this finding agreed with Lee’s report and the pathophysiology of inflammation and dyslipidemia association with HFs [9,20]. We did not observe a significant association between the HFs severity and LMR.

The present study explored significant associations between some components of LPs and HFs. We found higher TG levels in menopausal women with HFs than those without HFs after the multivariate regression. As we observed, the menopausal women with HFs showed more elevated levels of TC, TG, and LDL-C than those without HFs. This increase, although not significant, was consistent with the pathophysiology of the relationship of dyslipidemia with HFs [9]. TC, LDL-C, and TC/HDL-C levels increased, and HDL-C levels diminished in menopausal women, as indicated by Taleb-Belkadi et al. These alterations were ascribed to estradiol deficiency in menopause [19]. Furthermore, we observed that HDL-C tended to increase and TC/HDL-C tended to decrease in the menopausal women with HFs, which was not consistent with the pathophysiology of the association of dyslipidemia with HFs [9]. Also, this finding was in contrast to those of the Taleb-Belkadi et al. study that reported TC/HDL-C levels increased, and HDL-C levels diminished in menopausal women [19]. Moreover, the levels of these TC and TG in menopausal women with HFs increased as HFs severity extended. This finding could be according to the relationship between dyslipidemia and HFs [9]. After controlling the demographic and reproductive variables, we realized higher TG and HDL-C levels were associated with the HFs severity. The cumulative effect of HDL-C on the HFs severity is not justifiable. Therefore, further research is required to confirm such an association in menopausal women. Besides, Kaya et al. revealed that elevated TG levels were associated with higher total MRS symptom scores. However, they did not observe any significant difference in symptom severity-based groups regarding HDL-C levels [29]. Similarly, increased levels of TG were associated with higher total MRS scores in menopausal women, Cengiz et al. noticed [31].

We controlled the impact of physical activities and nutrition on participants. Some studies showed that more strenuous walking (such as Nordic walking) and walking duration might affect LPs [32]. We did not control the severity of walking in participants, which may be effective on LPs changes in menopausal women with HFs. As estradiol deficiency seems to be associated with dyslipidemia [9], we measured this sex hormone as a confounding variable. We did not observe any significant difference between the two groups of menopausal women without and with HFs in this regard.

The present study estimated the predictive levels of TG and NLR in the incidence and prognosis of various HFs severity levels, which was the first of its type to our knowledge. Relatedly, we noticed that with each 100 mg/dl increase in TG, the incidence of HFs can expand by nearly 1.5 times which can be practicable in the clinical approach.

In this study, hs-CRP negatively correlated with TG and TC/HDL-C in menopausal women with HFs, as we realized. Those correlations appear to be infirm. Moreover, they were not justified, given the pathophysiology of the association of inflammation and dyslipidemia with HFs [9]. Even though hs-CRP is an inflammatory marker, it sounds to require further study on menopause and its outcomes.

This study comprised some strengths: First, this study simultaneously assessed the relationship between SIMs, LPs, and the presence and severity of HFs. This is crucial because flushing is likely related to vascular dysfunction [9]. Therefore, investigating the alterations of SIMs and LPs in HFs severity-based groups also seems reasonable. Second, the participants with HFs experienced flushing for at least three months before enrollment and reported relatively reliable HFs frequency. Third, the participants only declared their daily HFs number, and we determined the HFs severity accordingly. Thus, we could prevent the participants from an unconscious bias concerning HFs severity report. Fourth, regarding low prices and ease of access, studies on NLR and LMR as hematologic inflammatory markers appear cost-effective.

This study included some limitations we need to address: First, structurally (cross-sectional study), it could not determine the causal relationship of SIMs and LPs with HFs. Second, we used the minimum sample size necessary to assess this relationship. As we noticed, the alterations of some variables, however not significant, were in line with the pathophysiology of the association between some SIMs, LPs, and HFs. Therefore, using a larger sample size may result in significant differences between groups of menopausal women. Third, in this study, we did not use a standard tool to evaluate food intake and physical activities in menopausal women. Instead, we asked them about their physical activities (type, frequency and duration of physical activity if done) and nutrition (frequency of having oily and fast foods if consumed). Fourth, participants recorded the number of their daily HFs subjectively. Accordingly, they may not have registered their accurate number of HFs.

Conclusion

Concisely, we found that TG was related to HFs in menopausal women. Indeed, menopausal women with HFs had higher levels of TG than those without HFs. We also noticed a rise in NLR, HDL-C, and TG levels as HFs severity enhanced. We furthermore noted negative correlations between hs-CRP and TG; and hs-CRP and TC/HDL-C in menopausal women with HFs. We also observed that TG and NLR could predict the incidence of HFs and determine the prognosis of the HFs severity values of 2 and 3. Hence, our study revealed that additive changes in some SIMs and LPs could be related to HFs and their severity. This event may suggest HFs as links between alterations in SIMs/LPs and their outcomes. Considering the interaction of these variables, surveying and focusing on SIMs’ and LPs’ levels in perimenopausal women may predict and decline the incidence and severity of HFs. Regarding the probable association of HFs with endothelial dysfunction [9], adjusting SIMs’ and LPs’ levels in perimenopausal women appears to prevent HFs and their potential outcomes through menopause. Moreover, researchers better perform longitudinal studies to elucidate the causal relationship among inflammation, dyslipidemia, and HFs.

Supporting information

Acknowledgments

We appreciate the Vice Chancellor of Research and Technology of Guilan University of Medical Sciences (GUMS), Rasht, Iran, for approval of this research project. We also value Hamzeh Yousefi Amin, the research laboratory supervisor, and his associates for their assistance and performance of the lab tests. Moreover, we would like to express our special thanks to all women who participated in this study.