Does body mass index compromise assisted reproductive technique outcomes? A cross-sectional study

Abstract Background Overweight and obese people face several health problems. Female obesity has been shown to reduce fertility in the general population. Assisted reproductive technology outcomes in obese cases are widely studied, but the results are inconclusive. Objective This study aimed to compare live birth rate (LBR) among women with 4 different types of body mass index (BMI). Materials and Methods In this cross-sectional study, data of 1611 women, who were candidates for fresh and frozen embryo transfer cycles, was extracted from 2051 medical files at the Reproductive Sciences Institute, Yazd, Iran from May 2019-May 2021. The participants were divided into 4 groups (underweight, normal, overweight, and obese) according to their BMI, and LBR was considered to be the main outcome. Results Of 1611 women, 39 were underweight, 585 were normal, 676 were overweight, and 311 were obese. Underweight women had the lowest LBR (12.8%), but there was no statistically significant difference (p = 0.55). In addition, LBR was compared in the 4 BMI groups according to age, type of transfer cycle (fresh or freeze), and cause of infertility, and there was comparable LBR in the 4 BMI groups. However, metaphase 2 oocyte rate, doses of gonadotropin usage in the cycles, and estradiol level had statistically significant differences (p < 0.001). Conclusion According to our study, obesity does not affect LBR in the IVF cycle, regardless of fresh or frozen embryo transfer cycles, different age groups, and causes of infertility.


Introduction
Lifestyle changes in modern societies have led to elevated body mass index (BMI) and obesity.
In some countries, nearly 50% of the population is either overweight or obese (1).According to the World Health Organization (WHO), there are about 2 billion overweight and 650 million obese adults across the world.The issue is also common among people of reproductive age (2,3).A systematic review showed that the prevalence of obesity and overweight in adults and children in Iran was about 12.8-76.4and 2.4-35.4%,respectively (4).Also, a study in 2020 reported that around 59% of adults in Iran were overweight or obese (5).
Overweight and obese people face several health problems that are extensively documented.
The negative effects of obesity on general health that are associated with the reproductive system are also remarkable.Research indicates that obesity in women can lead to reduced fertility and increase the risk of abortion in general population (6).However, similar conclusive results regarding the population requiring assisted reproductive technology (ART) have largely remained elusive.Some studies have linked obesity to diminished LBR, implantation rate, fertilization rate, and increased abortion rate due to decreased endometrial receptivity (7)(8)(9)(10).Others have claimed that the oocytes of overweight and obese women are smaller than those of normal BMI women and have a faster growth rate after fertilization (11).Oocytes of overweight and obese cases have lower quantity and quality compared to those with healthy weight (12,13).On the other hand, studies have found BMI to have no impact on the outcome of ART, and only obese women may need further doses of gonadotropins (14), and recommended reducing the BMI before ART is not necessary (15).

Given the inconclusiveness of the results
reported in the literature, as well as the fact that all research in this field of data has been carried out in European and Northern American populations, this study aims to investigate the potential impacts of BMI on ART outcomes in both fresh and frozen cycles in Iranian population.

Materials and Methods
In this cross-sectional study, data of 1611 women IVF outcomes were compared between groups.
The primary outcome was the live birth rate, and secondary outcomes were the implantation rate, abortion rate, and chemical and clinical pregnancy rate.

Results
From a total of 1611 women who underwent ART treatment, 39 were underweight, 585 were normal, 676 were overweight, and 311 were declared obese.The mean age was statistically different between the 4 groups; however, this difference was not clinically significant.

Duration of infertility and serum levels of AMH
were increased by an increment of BMI.However, this increase was not statistically significant.
The total gonadotropin dose (p < 0.05) was found to be significantly different across the BMI categories, obese women needed further gonadotropin doses.The percentage of good quality embryos (grade A, B) and endometrial thickness (ET) did not differ between the groups.
The percent of M2 oocytes was markedly lower in underweight women, and the difference between the 4 groups was statistically significant (p < 0.001).Also, the cause of infertility (p < 0.001), estradiol level (p < 0.001), and the number of transferred embryos (p < 0.001) had statistically significant differences between groups; the malefactor was the most common cause of infertility in underweight women.However, polycystic ovarian syndrome (PCOS) was the most common cause of infertility in other groups.
Estradiol level decreased by an increment of BMI (Table I).

ART outcomes such as chemical and clinical
pregnancy rate, live birth rate, and implantation rate did not differ significantly between the 4 groups.
However, these outcomes had the lowest range in underweight women (Table II).
Subgroup analysis was done in 2 age groups (< 35 yr, ≥ 35 yr) according to BMI (16).BMI had no significant effect on ART outcomes in each age group (Table III).Also, ART outcomes were assessed for different causes of infertility according to BMI, and the results were compared (Table IV).
ART outcomes in fresh and frozen embryo transfer cycles did not differ in the 4 BMI groups separately (Table V).No significant difference was observed between LBR in fresh and frozen cycles.In normal BMI, LBR in the frozen embryo transfer group was https://doi.org/10.18502/ijrm.v21i12.15040International Journal of Reproductive BioMedicine Bahrami et al.
higher than the fresh cycle (24.6%vs. 18.3%), and the difference was near to statistically significant, p = 0.07 (Table VI).
We considered that the baseline characteristics were statistically significant between the groups.Therefore, we considered confounding factors as categorical variables and investigated primary and secondary outcomes in subgroups.No difference in the subgroups of age was observed.We adjusted the confounding effect of age by logistic regression analysis and did not find a significant effect on outcomes.

Discussion
A total of 1611 women were grouped based on their BMI, and 42% of the majority of the subjects were in the overweight group.The main objective of this study was to compare LBR among women with 4 different types of BMI.The outcomes were compared among the underweight group which had the lowest LBR.However, due to the small number of cases in this group (only 2.5% of the entire study population), a significant p-value was not observed.In the literature, we found increased BMI to impact IVF.In a retrospective cohort study in 2020 in China on 14,782 cycles, the correlation of BMI and cumulative live birth rate (CLBR) was found to be an Specifically, CLBR was highest in underweight, plateaued in normal and overweight, and decreased in obese women.
However, the BMI group's definition was not according to the WHO (17).In another retrospective cohort study on 2,39,127 fresh embryo transfers, which was published in 2016, LBR decreased progressively with increasing BMI (8).These results were confirmed by other studies (7,18).
Contrary to these studies, in 2020, in a retrospective cohort of 1415 cases of blast transfer, the LBR in obese people who had frozen blast embryo transfer was examined and was found to be the same as normal people.Therefore, it was concluded that obesity has no harmful effect on endometrial receptivity (3) In this study, doses of gonadotropin were significantly higher in obese cases, which is similar to the result of a previous study (20).Although the requirement for higher doses of gonadotropin in obese and overweight cases is logical, another study found no correlation between BMI and doses of gonadotropin (21).
Regarding the impact of BMI on M2 oocytes, some have concluded that high BMI reduces M2 oocyte count (22), while others have reported contradictory results, indicating no impact from higher BMI (23).Our results suggest a significant reduction in M2 oocytes in the underweight group.
Nevertheless, in agreement with other reports in the literature, this reduction in M2 oocytes did not affect good-quality embryo numbers (6,24). International Therefore, the reduced M2 oocyte count is not clinically important.Yet, there are studies such as that believe in a reduction of good-quality embryos with increased BMI (25).
In In all the previous studies, most analyses focused on fresh cycles, with a few studies on freeze cycles.However, in the present study, we analyzed both fresh and freeze cycles.
Pregnancy results among the 4 BMI groups were compared, and in comparing fresh and freeze cycles regarding the LBRs, it was found that in the normal BMI group, the freeze cycle was more effective with close to a significant p-value of 0.07.
However, for the obese group, no meaningful difference was found between the fresh and freeze cycles.
It seems that the main discrepancy in our result with studies is the genetic diversity of our population and perhaps the larger sample size of those works (8,17).Thus, we recommend a large multicentric study about this context in our country.The main limitation of our study is its retrospective nature, leading to some missing data points in the participants' files.As such, a number of the patient files were removed from the study due to missing required information.However, the analysis of multiple aspects of ART cycles, only a few of which had been considered in earlier studies, may cover the main shortcomings of the work.

Conclusion
According to our findings and some similar studies, BMI may not be a prognostic factor for ART outcomes.Factors such as age, antral follicular count, and AMH are the best prognostic factors.It is recommended to conduct a multicenter prospective study with a larger sample size and the impact of some factors like lifestyle, previous ART cycles, or comorbidity.

( 18 -
42 yr), who were candidates for fresh and frozen embryo transfer cycles, was extracted from a total of 2051 medical records at the Reproductive Sciences Institute, Yazd, Iran from May 2019-May 2021.Women with uncontrolled underlying diseases, uterine anomaly, severe male factors like testicular samples or ejaculates with below 1 × 10 6 spermatozoa/ml, uterine surrogacy, and ovum donation were excluded.All relevant information, such as anti-Mullerian hormone (AMH) levels, BMI, cause and duration of infertility were all obtained from the participants' files.All women were categorized into 4 BMI groups according to the WHO classification: group 1 (underweight, BMI < 18.5 kg/m 2 ), group 2 (normal, BMI = 18.5-24.9kg/m 2 ), group 3 (overweight, BMI = 25-29.9kg/m 2 ), and group 4 (obese, BMI ≥ 30 kg/m 2 ).The protocol for ovarian stimulation was agonist or antagonist protocol, and the dosage of gonadotropin was adjusted according to the women's age, antral follicular count, and AMH levels.In frozen embryo transfer cycles, endometrial preparation was done by estradiol valerate starting on the 2 nd day of the cycle, when endometrial thickness reached ≥ 7 mm, progesterone supplementation was administered.One or 2 International Journal of Reproductive BioMedicine BMI and pregnancy outcomes embryos on day 3 were transferred on the 4 th day of progesterone administration.
Statistical analysis was done by Statistical Package for the Social Sciences, version 18.0, SPSS Inc., Chicago, Illinois, USA (SPSS).Data processing was done by Chi-square and Fisher's exact tests in categorical variables.The distribution of continuous variables was checked by the Kolmogorov-Smirnov test, and the Kruskal-Wallis test was used for comparison in continuous variables.Subgroup analysis was performed by Chi-square or Fisher's exact test to assess LBR in different age groups and different embryo transfer cycles according to BMI.The significance level was considered as < 0.05.

Table II .
Comparison of pregnancy outcomes according to BMI Data are presented as n (%), the Chi-square test.BMI: Body mass index

Table III .
Pregnancy outcomes in each age group according to BMI Data are presented as n (%), the Chi-square test.BMI: Body mass indexTable IV.Pregnancy outcomes of each infertility cause according to BMIGroupsGroup 1 (n = 39) Group 2 (n = 585) Group 3 (n = 676) Group 4 (n = 311)

Table VI .
Pregnancy outcome (live birth rate) in BMI groups according to fresh and frozen cycle