Impact of polycystic ovary syndrome on quality of life of women in correlation to age, basal metabolic index, education and marriage

Polycystic ovary syndrome (PCOS) is the major endocrine related disorder in young age women. Physical appearance, menstrual irregularity as well as infertility are considered as a sole cause of mental distress affecting health-related quality of life (HRQOL). This prospective case-control study was conducted among 100 PCOS and 200 healthy control cases attending tertiary care set up of AIIMS, Patna during year 2017 and 2018. Pre-validated questionnaires like Short Form Health survey-36 were used for evaluating impact of PCOS in women. Multivariate analysis was applied for statistical analysis. In PCOS cases, socioeconomic status was comparable in comparison to healthy control. But, PCOS cases showed significantly decreased HRQOL. The higher age of menarche, irregular/delayed menstrual history, absence of child, were significantly altered in PCOS cases than control. Number of child, frequency of pregnancy, and miscarriage were also observed higher in PCOS cases. Furthermore, in various category of age, BMI, educational status and marital status, significant differences were observed in the different domain of SF-36 between PCOS and healthy control. Altogether, increased BMI, menstrual irregularities, educational status and marital status play a major role in altering HRQOL in PCOS cases and psychological care must be given during patient care.


Introduction
Polycystic ovary syndrome (PCOS) is a major endocrine disorder in young age women affecting their health-related quality of life (HRQOL) and their mental well-being as well [1,2]. Moreover, this develop into lifelong health condition that continues far beyond the young ages and affects around 5 million young age population in the United States of America [3,4]. In India, PCOS has been reported to vary between racial counterparts with an estimated prevalence of 9.13% in adolescents [5]. The major changes in physical appearance, obesity, along

Participants
Patients diagnosed with PCOS, based on criteria derived from the 2003 ESHRE/ASRM (Rotterdam criteria) were arbitrarily enrolled in the study. PCOS is diagnosed as the presence of at least two of three of the following: 1) Oligo/anovulation, 2) hyperandrogenism, 3) Polycystic ovaries [18]. A healthy control (HC) was selected from participants of the same population and having regular menses and had no clinical features of hyperandrogenism as well as infertility.

Data sources
Data was collected after describing both written and verbal information about the study. After explaining, the informed consent form was signed by each participant and then they were requested to complete the questionnaires. Face-to-face interviews were conducted by investigators to the subjects meeting the inclusion criteria and consented for the participation into the study in three parts: Part A: Semi-structured, pre-validated questionnaires were used for collecting information on the socio-demographic, economic and reproductive history. Part B: Pre-validated SF-36 questionnaire is a standard diagnostic tool for evaluating various aspects of the HRQOL over the previous 4 weeks [19]. Its validity, sensitivity, reliability, internal consistency and stability, as well as test-retest reliability have frequently been confirmed in various studies [20][21][22]. SF-36 contains 8 domains: general health, physical functioning, and role limitations due to physical health, role limitation due to the emotional problem, body pain, social functioning, energy/fatigue and emotional well-being. The scores for each domain range from 0-100, where higher scores indicate better condition.

Study size
The sample size was estimated post assuming α-error of 0.05, power of 80%, percentage of controls having a poor quality of life to be 20% based on previous studies and odds of poor QOL among cases to be twice than among controls. Hence, a total of 100 PCOS cases and 200 healthy control cases were enrolled in the study.

Inclusion criteria
We included all diagnosed case of PCOS only, female from menarche to menopausal age between the age of 10-49 years, and those given informed consent.

Exclusion criteria
Patients having cognitive or developmental disabilities/another major illness that substantially influenced the HRQOL of women, confirmed malignancy and deformities, as well as breastfeeding women were excluded from the study.

Statistical analysis
Data were analyzed by using statistical software-Stata Version 14.0 (Stata Corp, Texas, USA). After checking for the normality condition for continuous variables, the appropriate statistical test was applied. Confounders like excessive body weight were taken into consideration. Quantitative data expressed as mean±SD, minimum and maximum followed normal and skewed distribution respectively. Analysis of covariance model (ANCOVA) was used to address potential confounders. Categorical variables expressed as frequency and percentage. Pearson Chi-Square test and Fisher exact test were used to checking the association between qualitative variables and categorical variables. Logistic regression analysis was used to estimate odds (95% CI) and models were robust for PCOS and other variables. Multivariable linear regression analysis was performed to observe the association between the variables. Independent t-test and One Way ANOVA used to compare normally distributed continuous variable between two and three categories respectively. Rank sum/Kruskal Wallis test used for comparing skewed continuous variables among categories and to look association between demographic categories. For all statistical tests, P-value < 0.05 is considered as statistical significant.

Result
The outcome of socioeconomic status (SES) of a woman with PCOS and HC cases are mentioned in Table 1. The women with PCOS and HC were comparable in respect of marital status and family type. Statistically significant differences were observed between PCOS and HC in terms of age (P<0.020), BMI (P<0.001), educational status (P<0.001), marital status (P>0.05) and work category (P<0.001). Total 97% of PCOS case was below the age of 30 years in comparison to 78% of control. Among all PCOS cases, 60% was student and almost 54% received higher education. Among the HC group, 39% was student and only 15% received higher education (P<0.001). A higher percentage of PCOS cases (16%) belong to greater BMI (>30) in comparison to HC (2%). As shown in Table 2, among the PCOS group, a significant percentage of women (33%) has menarche at age greater than 14 years and there was no any HC cases lies in this category (P<0.001). In respect of menstrual history, PCOS cases have a higher percentage of irregular  Table 3, Overall differences of mean in PCOS and HC case comparable in respect of BMI (P<0.125).
However, the highest range of BMI was more in PCOS cases in comparison to HC. Whereas, a statistically, significant differences in mean were observed in respect of age (P<0.009), age at marriage (P<0.001), age of menarche (P<0.001), number of children (P<0.001) and number of pregnancy (P<0.006) between PCOS and HC cases. In case of age, women with PCOS at age �19 showed significantly higher score for of general health (P<0.001), role limitation due to physical health (P<0.001), role limitation due to the emotional problem (P<0.022), pain (P<0.025) and social function (P<0.010) in comparison to age >30. However, comparable differences were observed in physical function (P<0.116), energy or fatigue (P<0.087) and emotional well-being (P<0.108). In HC cases, women of age �19 showed a statistically higher score in general health (P<0.001), physical health (P<0.001), role limitation due to the emotional problem (P<0.005) and energy/fatigue (P<0.001). Comparable differences were observed for role limitation due to physical health (P<0.818), pain (P<0.424), social functioning (P<0.110) and emotional well-being (P<0.147; Table 4). As per BMI is concerned, PCOS women scored high and statistically significant differences were observed in case of BMI those who have value <18 in comparison to >30. In addition, significant differences were observed for general health score (P<0.001), physical health (P<0.001), energy and emotion (P<0.001). Whereas, comparable differences observed in role limitation due to physical health (P<0.085), role limitation due to the emotional problem (P<0.565), pain (P<0.189), social function (P<0.549) and emotional well-being (P<0.127). In HC case, there no significant difference was observed in all the eight domains of SF-36 ( Table 5). As per the level of education is concerned, HRQOL score was higher in all eight domain of SF-36 in well-educated women in comparison to illiterate or women having education of primary level, but this difference was observed to be statistically non-significant and comparable. In case of HC women, HRQOL score in graduation level was higher and significant differences were observed in relation to the level of education for SF-36 domains like general health (P<0.001), physical health (P<0.039) and energy/fatigue (P<0.003). However, we observed comparable differences among all other domains of SF-36 (Table 6).
We observed significant differences between married and unmarried PCOS cases in terms of general health (P<0.001), physical functioning (P<0.027), role limitation due to physical health (P<0.006), role limitations due to emotional problems (P<0.002), pain (P<0.001), social functioning (P<0.001), energy/fatigue (P<0.003) and emotional well-being (P<0.001). Whereas, in HC cases, no differences were observed between married and unmarried cases in regarding SF-36 domain score for role limitation due to physical health (P<0.538), role limitation due to the emotional problem (P<0.105), Pain (P<0.044), social functioning (P<0.225), emotional well-being (P<0.857). However, significant differences were observed for general health (P<0.001), physical health (P<0.002) and energy/fatigue (P<0.001) among married and unmarried HC cases ( Table 7). The Fig 1 and Table 8 exhibits the regression analysis data plot and we was observed strong association between infertility and menstrual irregularities (P<0.049) as well as emotional well being (P<0.001) of PCOS patients. We also observed infertility (P<0.001) and hirsutism (P<0.05) as a major predictor affecting in all domain scores (Table 9).

Discussion
PCOS has no any constant treatment due to its multifaceted features. However, lifestyle modification, hormonal contraceptives and some other drugs like inositol, clomiphene, eflornithine, finasteride, flutamide, letrozole, metformin, spironolactone has been reported to ameliorate the PCOS symptoms [ PCOS is an endocrine disorder and its long term complications affect various aspects of HRQOL in women [23,24]. Despite the various evidence about compromised HRQOL in women with PCOS, we further explored the other determinants that may help the clinician in care of the patient well-being [17,25,26]. Overall, we demonstrated the drastically compromised HRQOL in young women suffering from PCOS. As earlier reported, the woman with PCOS belongs to a lesser age group in relation to HC indicating a higher prevalence of PCOS cases in young age woman especially in adolescents [27]. As per SDS is concerned, the major difference between PCOS and control cases was observed in case of age, BMI and level of education [28]. Thus, all this indicated that PCOS affects HRQOL more in the young woman and the SDS definitely affect the prevalence of PCOS. The age of menarche in the majority of women was > 18 years as earlier reported [29]. This was in contrast to other reports [30]. Consistently, increased age of menarche was observed in PCOS cases indicating the impact of first menstruation in young women life and in the development of PCOS and another reproductive as well as metabolic disorder [31,32].
As previously reported, we observed a direct correlation between the PCOS and irregular or delayed menses and having no children that can be taken as a symptom of PCOS diagnosis [33]. The higher number of women having lesser than two children and lesser number of times get pregnant further supported this compromised HRQOL [30]. This compromised HRQOL in PCOS case was further supported by our study.
The overall decreased mean of BMI and age at menarche indicated as PCOS symptoms. Concurrently, mean age, age at marriage, number of children and frequency of pregnancy was less in PCOS cases than control. Consistent to the previous study, these indices corroborate above findings and strongly indicated the deterioration of HRQOL in PCOS women [28]. https://doi.org/10.1371/journal.pone.0247486.g001 Table 8. Multiple linear regression analysis of possible related factors in HRQOL. The independent variable is MI. The independent variable is EWB.

Model Summary and Parameter Estimates (B)
EWB: Emotional well being; P value less than 0.05 is considered as significant differences between the group. MI: Menstrual irregularities; P value less than 0.05 is considered as significant differences between the group. https://doi.org/10.1371/journal.pone.0247486.t008 Altogether, higher fertility disorder in PCOS cases was observed that directly affects their HRQOL due to physical, social as well as emotional issues. Furthermore, in PCOS cases, physical, social and emotional well-being more affected as evidenced from all eight domains of SF-36 indicating strongly compromised HRQOL than HC cases [34]. In particular, we also compare the mean score in relation to age, BMI, educational level and marital status. Consistent with the previous report, increasing the age had a more negative impact on different domains of SF-36 in PCOS cases than HC cases. Comparable scores in PCOS women with increasing age for physical health, energy and emotional wellbeing may be due to improved regular menses with age concurrent to improved PCOS features and loss of societal fear. Whereas, in HC cases, changes in normal life trend in the prospect of HRQOL was seen with increasing age indicating normal HRQOL [35].
In a similar fashion, with the increase in the BMI, physical activity was not affected in PCOS cases as observed from different domains of SF-36 but the emotional problem was more affected in PCOS cases in comparison to HC. This may be a major reason for compromised HRQOL in PCOS women [36,37]. In HC cases, none of the scores of SF-36 domains was different between BMI groups. Consistent to previous studies, with increasing the level of education, all the domains of SF-36 in PCOS cases have improved HRQOL. Similar to other HRQOL studies in different diseases, where well-qualified patients have better HRQOL than illiterate cases [38]. We also observed that well-qualified group probably have higher number of PCOS cases that directly support that improved SDS is a major contributing factor in developing PCOS. The differences in scores of all the domains of SF-36 were observed in married PCOS cases in contrast to unmarried. Thus, consistent to the previous report, the HRQOL of the unmarried cases was better in comparison to married women [39]. Whereas, in HC cases, all physical, social, as well as emotional wellness, were similar in both married and unmarried women. This may be due to their social independency and quality of education in young women with PCOS. As reported earlier, infertility and hirsutism emerges as the major problem affecting the overall HRQOL and a strong association has been observed between infertility and emotional well-being [40].

Conclusion
Our data compares the relations between PCOS and HC cases of overall HRQOL. We explored the strong association between PCOS and SES, and suggest that with increasing age and BMI PCOS patients had lower scores on SF-36; opposite association was with education level. However, Infertility emerges as the major predictor affecting overall HRQOL in PCOS cases. The present study does have its limitations of not measuring biochemical assessment and ultrasonography indices.