{"paper_id":"957ce7fb-9ffa-4e33-b907-cd43e710d803","body_text":"Intrauterine insemination (IUI) is used to transport spermdirectly\ninto the uterus. It is a simple, non-invasive, andcost-effective\ntechnique used for assisted reproduction. Themost common indication\nfor IUI is cervical infertility, and\nit is also used in male subfertility, anovulation, endometriosis\ncases in which at least one tube is healthy, as well as unexplained\ninfertility ( 1 ,  2 ). Although there may be a trendtowards higher\npregnancy rates when the number of IUIsper cycle is increased,\na recent meta-analysis has shownthat increased IUI numbers do not\nincrease pregnancy ( 3 ).\nPrevious investigations reported that IUI had a success rateof 10-20% \nfor clinical pregnancies ( 4 ).\nRecently, the effect of vitamin D (VD) has been investigated on not only the musculoskeletal system, but also inthe reproductive and other systems ( 5 ). The biologic actionsof VD are mediated through the vitamin D receptor (VDR).\nVDR was found to be in the ovary (particularly the granulosacells), \nuterus, placenta, and testis, suggesting VD may have a\nsignificant role in human reproduction ( 6 ). Two studies supporting\nthis data indicated that VD deficiency is responsible\nfor reduced fertility and reproductive capacity in female rats\n( 7 ,  8 ). Research conducted on human subjects also supports\nthis role as in experimental animal studies ( 9 ).\nCalcitriol (1, 25 dihydroxyvitamin D 3 ) which is the active\nform of VD stimulates CYP19 expression (CYP19\nencodes the aromatase enzyme) that results in increased\nestrogen production, when it was bound to VDR ( 10 ).\nFurthermore, it has been reported that decidua secretes \ncalcitriol during blastocyst implantation, and calcitriol \nhas been reported to regulate the immune response in the \nmaternal-fetal interface during pregnancy ( 11 ).\nThere are several studies which presented controversial\nresults on the differences in 25-hydroxyvitamin D 3 [25 (OH) D 3 ]\nlevels of the patients undergoing different \ninfertility treatment modalities ( 12 ,  13 ). The aim of our \nstudy was to investigate the 25 (OH) D 3  levels in patients \nwho underwent ovulation induction with IUI and then to \ndetermine the relationship between 25 (OH) D 3  levels and \nthe occurrence of pregnancy.\n\nThis case-control study was conducted between March \n2014 and June 2014 in the infertility outpatient clinics of \nZekai Tahir Burak Women’s Health Education and Research \nHospital. This is a government supported tertiary \nlevel maternity hospital located in the capital city of Turkey. \nThe institutional review board approved the study and \ninformed consent was obtained from each patient (approval \nnumber: 23.09.2013/9). All of the study protocols were carried \nout in accordance with the Helsinki Decleration ( 14 ).\nWe defined the infertile patients as those reproductive \nage couples who were unable to become pregnant in the \nabsence of contraception. For the women below 35 years \nof age, infertility was diagnosed as a minimum of 1 year \nof trying to become pregnant, whilst for the women above \n35 years of age, the diagnosis was limited to 6 months \nof unprotected sexual intercourse. After we obtained detailed \ninformation about age, duration of infertility, infertility \ntype, previous history of surgery, and any systemic \ndisturbances (such as diabetes mellitus, hypertension, and \nthyroidal disease), a complete physical and gynaecological \nexamination was performed on all of the women. We \nconfirmed tubal patency in the women using hysterosalpingography \n(HSG) and if there was bilateral tubal occlusion \ndetected with HSG, we applied laparoscopy and \nhysteroscopy to define any pathology such as pelvic adhesions \nor endometriosis. When we suspected an intracavitary \nlesion in the uterus after HSG, or transvaginal \nultrasound, we performed hysteroscopy.\nWe included women with mild male factor infertility, \nunexplained infertility, and polycystic ovary syndrome \n(PCOS). We excluded patients who had advanced age \n(above 40 years of age), any systemic or endocrine diseases, \nstage 3-4 endometriosis, or intracavitary lesions in \nuterus (such as endometrial polyp, submucous myoma, \nand uterine septum), smokers and women who used of \nany kinds of drugs or substances likely to affect levels \nof VD. We also excluded patients whose partner had a \nmotile sperm count lower than 5 million/mL. The fertile \ngroup consisted of patients who applied to the family \nplanning unit of our hospital for contraceptions. These patients \nhad given birth in the previous 12 months, has not \nbreastfed their neonate, and had no history of infertility.\nAfter initial clinical assessment, infertile patients were \nevaluated for clomiphene citrate (CC) or gonadotropins \n(Gn) and IUI use. Those patients who had used CC with \nIUI treatments for three times or were above 35 years of \nage were directed into the Gn with IUI regimen (n=63), \nwhilst the other infertile patients were directed into the \nCC and IUI regimen (n=41). When 18-20 mm (dominant \nfollicles) were found through ultrasound, 2 human chorionic \ngonadotropin (hCG, Pregnyl, MSD, Netherlands) \nampoules containing 5,000 units each, were injected intramuscularly, \nand IUI applied 36 hours after the injection. \nWhen there were 3 or more dominant follicles, or \nendometrial thickness was less than 6 mm, hCG was not \nadministered. Then 2 weeks later, a blood sample was \nobtained from patients for ß-hCG measurement. Clinical \npregnancy was diagnosed 5 weeks after IUI, when the evidence \nof fetal heart activity or presence of the gestational \nsac in the uterine cavity was detected.\nThe concentration of serum 25 (OH) D 3  was used to determine \nthe status of VD in the body for this study since it has \nbeen proven to be the best biomarker for VD insufficiency. It \nalso reflects VD levels from both dietary intake and in-skin \nsynthesis ( 6 ). The two groups were matched in term of veiling \nhabits, daily exposure to sunlight, and dietary intake of \nVD-rich foods which was determined by a dietician.\nThe serum levels of 25 (OH) D 3  levels and baseline \nhormones including estradiol, follicle stimulating hormone \n(FSH), luteinizing hormone, prolactin, and thyroid \nstimulating hormone were measured on the third day of \nthe menstrual cycle when ovulation induction was started. \nWe performed the recruitment of study volunteers in a single \nseason, because the blood levels of VD have seasonal \nvariabilities ( 15 ,  16 ). In addition, patients living in the same \ngeographical region were selected for the study ( 17 ).\nAfter overnight fasting, venous blood samples were obtained \nearly in the morning and transferred to the laboratory \nin a non-transparant box to avoid exposure to light, and \nthen serum was separated by centrifugation at 5,000 rpm \n(2,236 g) for 10 minutes. The serum 25 (OH) D 3  levels \nwere measured using an enzyme linked immunosorbent \nassay kit (Immunodiagnostic AG, Leverkusen, Germany), \nand presented in ng/mL. The intra-assay and inter-assay \ncoefficients of variation for serum 25 (OH) D 3 , were 8.9 \nand 10.6% respectively. Serum 25 (OH) D 3  concentrations \n<20 ng/mL was considered as VD deficiency. Types \nof VD deficiency were also classified as mild (10-20 ng/\nmL), moderate (5-10 ng/mL), and severe (<5 ng/mL). Serum \n25 (OH) D 3  concentrations between 20 and 30 ng/\nmL was accepted as VD insufficiency whereas a threshold \nvalue of =30 ng/mL was considered sufficient serum VD \nlevels. Basal hormone levels were measured using an Immulite \n2000 analyzer (EURO/DPC Ltd., Gwynedd, UK). \nBody mass index (BMI) was defined as the weight in kilograms \ndivided by the square of the height in meters.\nWe examined the women who had a positive result for \nß-hCG using transvaginal ultrasound at at least weeks 6-7 \nof gestation to detect fetal cardiac activity. The difference \nbetween the two subgroups (pregnant and non-pregnant) \nof infertile patients in terms of 25 (OH) D 3  levels was the \nprimary outcome measured of this study. The secondary \noutcome was the comparison of serum 25 (OH) D 3  levels \nbetween infertile and fertile groups.\nData were recorded and analysed using the Statistical \nPackage for the Social Sciences program for Windows \nversion 17.0 (SPSS Inc, Chicago, IL, USA). The normal \ndistribution of the variables was assessed using the Shapiro-\nWilk’s test. Continuous variables were presented as \nthe mean with standard deviation (SD) or median (range), \nand categorical variables were presented as the number \n(percentage) of subjects. Continuous variables were \ncompared using independent samples t-test if they were \nnormally distributed or with the Mann-Whitney U test if \nthey were non-normally distributed. Categorical variables \nwere analyzed using the Chi-square (χ 2 ) test or Fisher’s \nexact test. Correlations were calculated using Spearman's \ncorrelation analysis. In all analyses, two-tailed P<0.05 \nwere considered as statistically significant. Post-hoc power \nanalysis demonstrated that we achieved a power of 0.95 \nwith a 5% level of significance and a 0.5 effect size by \nusing a two sample comparison ( 18 ). Power analysis was \ncarried out on G-power software (G-power v3.1.9.2, Universitat \nKiel, Kiel, Germany).\n\nOne hundred and four infertile and one hundred and \nthree fertile women were included into this cross-sectional, \ncase-control study. Examination of the infertile and fertile \npatients showed that there was no statistically significant \ndifference between the groups regarding their mean age \nand BMI. Obstetric history characteristics were statistically \nsignificantly different between the two groups (P<0.001 for \nall). The mean FSH levels were higher in the infertile patients \nthan in the fertile patients (7.4 ± 2.1 mU/mL vs 6.2 ± \n1.6 mU/mL, P=0.001), but it was within the normal range \nin either group. Mean prolactin levels of the fertile group \n(14.4 ± 5.4 ng/mL) were higher than the infertile group’s \n(12.2 ± 4.6 ng/mL). This difference was statistically meaningful \n(P=0.002), but those values were in the normal range \nas with mean FSH levels. There were no statistically significant \ndifferences in 25 (OH) D 3  levels between the 2 groups \n[7.3 (3-25.5) ng/mL vs. 6.8 (3.4-37.1) ng/mL, P=0.512], as \nseen in Table 1. No significant correlation between serum \n25 (OH) D 3  and FSH levels was observed either in the entire \nstudy population (Spearman’s r=0.051, P=0.466).\nDescriptive characteristics and serum 25 (OH) D 3  levels of infertile and fertile patients\nBMI; Body mass index, FSH; Follicle-stimulating hormone, LH; Luteinizing hormone, TSH; Thyroid stimulating hormone, *; Median (minimum-maximum), **; Mean (SD),  a ; Student t Test, and  b ; Mann Whitney U test. P<0.05 is considered as statistically significant.\nThe severity of VD deficiency in the infertile and fertile \ngroups showed that most of the participants were deficient \nfor VD (96.2 vs. 97.1%) and only 1 (1%) participant in \nthe fertile group had 25 (OH) D 3  levels =30 ng/mL. 19 \n(18.3%) patients were in the severe deficiency group, 56 \n(53.8%) cases were in the moderate deficiency group, and \nlastly 25 (24%) women were in the mild deficiency group \namong the infertile group. The number of fertile patients \nin the same groups was 19 (18.4%), 58 (56.3%), and 23 \n(22.3%), respectively (P=0.776).\nAfter IUI treatment, the numbers of clinical pregnancies \nand live births among 104 infertile patients were 14 \n(13.3%) and 10 (9.61%), respectively. When infertile patients \nwere divided into two subgroups (pregnant and non-\npregnant), there was no statistically significant difference \nbetween these subgroups regarding age, BMI, obstetrical \nhistory, baseline hormone levels, or ovulation induction \nagent used. Similarly, no significant difference was observed \nbetween the pregnant and non-pregnant subgroups \nof infertile patients in terms of serum 25 (OH) D 3  levels \n(P=0.267). Ten (71.4%) patients out of the 14 clinical \npregnancies had moderately deficient VD levels. The only \nsignificant parameter that may predict pregnancy was the \nage of the patients, namely the pregnant group was statistically \nsignificantly younger than the non-pregnant group \n( Table 2 ).\nIndividual characteristics, ovulation induction type and vitamin D levels in pregnant and non-pregnant patients after IUI\nIUI; Intrauterine insemination, BMI; Body mass index, FSH; Follicle-stimulating hormone, LH; Luteinizing hormone, TSH; Thyroid stimulating hormone, CC; Ovulation induction with clomiphene citrate, Gn; Ovulation induction with gonadotropin, *; Median (minimum-maximum), **; Mean (SD), ***; n (%),  a ; Student’s t test,  b ; Mann-Whitney U test, and  c ; Fisher’s exact test. P<0.05 is considered as statistically significant.\n\nOur study showed that infertile and fertile patients had \nsimilar serum VD levels and that there was no statistically \nsignificant difference in serum VD measurements \nbetween the pregnant and non-pregnant groups after IUI.\nVD has an essential role in both male and female reproductive \nsystem ( 19 ). It was found that its deficiency is \nhighly prevalent among women undergoing ovarian stimulation \n( 9 ). Considering the previous data, we designed \nsuch a study assuming that VD could be lower in infertile \npatients, but we found no relationship between them. This \nresult may be due to the fact that VD deficiency is very \ncommon in our study population, because 200 of the 207 \npatients also including women with no fertility problem \nhad VD deficiency at the initial examination. This was a \nsurprise and suggests that fertile patients who had a delivery \nin the preceding 12 months may have exhausted their \nVD stores during the most recent pregnancy and they had \nnot been able to replace it yet. VD deficiency is one of the \ngeneral public health matters in our country, with similar \ninferences having been suggested in other studies from \nour country ( 20 ,  21 ).\nOvulation induction with IUI is the most utilized method \nof infertility treatment in our unit. The success of IUI \ntreatment is multifactorial, and pregnancy rates per cycle \nhave been estimated as 10.2% in a IUI cycle with controlled \novarian stimulation ( 22 ).\nA study by Ott et al. ( 23 ) demonstrated that 25 (OH) \nD 3  levels may predict ovarian response to ovarian stimulation. \nThis suggestion is consistent with another study \nshowing that VDR exists in human ovaries and is important \nfor sex steroid synthesis ( 10 ). However, when we \ncompared pregnant patients with non pregnant patients \nin terms of serum VD levels, there was no statistically \nsignificant difference between them. This may be associated \nto the differences of individual VDR receptivity and \nVDR polymorphism ( 24 ). Although VDR polymorphism \nhas been reported as not being related to infertility in an \nendometriosis study, there is a need for further research \nto clarify this particular issue ( 25 ). Another noteworthy \nresult of our study is that patients who became pregnant \nafter IUI treatment were younger than those who did not. \nIt has already been shown that age is one of the most important \nfactors in infertility management ( 26 ).\nIn two rat studies, VD deficiency was shown to significantly \nincrease infertility, decrease probability of viable \nbirths and healthy full-grown individuals ( 8 ,  27 ). \nAlthough the exact mechanism remains to be elucidated, \ncompromised ovarian folliculogenesis and infertility were \nfound in two studies conducted on VD deficient mice ( 28 , \n 29 ). A recent human study supporting these inferences \nshowed that there is a negative correlation between serum \nlevels of 25 (OH) D 3  and FSH ( 30 ). However, we found \nno correlation between them.\nVD has been found to be related with the activation of \nkey enzymes in steroidogenesis such as 3-beta-hydroxysteroid \ndehydrogenase, and it has been shown to induce \nthe production of progesterone that consequently leads to \nuterine quiescence ( 5 ). Thus, VD may play a protective \nrole for ongoing pregnancies through this mechanism.\nA recent randomized controlled trial by Asadi et al. ( 31 ) \nshowed that endometrial thickness was enhanced by the \nadministration of VD in patients undergoing Gn and IUI \ntreatment. Another study found that higher serum and \nfollicular fluid 25 (OH) D 3  levels were associated with \nhigher pregnancy rates in women undergoing IVF ( 32 ). \nSimilarly, a Greek study group found that follicular fluid \nVD levels significantly correlated with the quality of \nembryos ( 18 ). However, the same authors suggested that \nexcess serum and follicular fluid vitamin levels may have \na detrimental effect on IVF outcomes. These findings led \nus to think that an optimal level of VD is necessary for \novulation, fertilization, and implantation.\nThe strength of this study is that our data is single-centered \nand reliable. The data were obtained prospectively \nfrom the patients living in the same geographical region \nduring the same season. There is a limitation to our study; \nVD deficiency was wide-spread in our study population, \nconsistent with the results of previous studies on this issue \n( 19 ,  20 ). The similarity of the groups in terms of high \nprevalence of VD deficiency may have caused 25 (OH) \nD 3  levels to not be distinguishable between the groups.\n\nNo significant difference was observed between pregnant \nand nonpregnant women who underwent ovulation \ninduction with IUI treatment with regard to serum 25 \n(OH) D 3  levels. No association was found between infertility \nand serum 25 (OH) D 3  levels either. Further research \nwhich compares women who have deficient and sufficient \nserum VD levels is warranted.","source_license":"CC-BY-4.0","license_restricted":false}