Prolactin levels among women attending infertility clinics

Study area

This study was conducted at the gynecological clinics of the University of Uyo Teaching Hospital (UUTH) and St. Luke’s Hospital, both located in Uyo metropolis, the capital city of Akwa Ibom State, situated in Nigeria’s South-South geopolitical zone.

The UUTH is the primary tertiary healthcare center, while St. Luke’s Hospital, Anua, is the only secondary healthcare facility. They both provide comprehensive medical services to all patients, including specialized care in obstetrics and gynecology.

Study design and study population

This research adopted a comparative cross-sectional study design to evaluate serum prolactin levels among women attending infertility clinics over a 4-month period, from December 01, 2020 to March 31, 2021. The findings were compared with those of fertile women seeking family planning services within Uyo metropolis.

Determination of sample size

The sample size was determined using the statistical formula for comparing two proportions:

$n=\frac{{\left(Z_{\alpha 2}+Z\beta \right)}^{2}P{\left(1-P\right)}^2}{{\left(P_1-P_2\right)}^2}$ [10]

In a similar study conducted in Dhaka, Bangladesh,^[11] a 14% prevalence of hyperprolactinemia among women experiencing infertility was reported, while only 4% of fertile controls exhibited this condition. These prevalence rates were instrumental in determining the sample size. The initial sample size determination indicated that 128 participants were needed for each group, resulting in a total of 256 participants. However, to account for potential dropouts or non-responses, an adjustment was necessary. A 10% non-response rate was anticipated, and this was factored into the sample size calculation. The adjustment was made using the formula:

Adjusted Sample Size = (Initial Sample Size × 100)/(100 - Expected Non-Response Rate).

Where the expected non-response rate, denoted as ×, was 10%.

Therefore, this study needed 142 cases and 142 controls, resulting in a total of 284.

Participant recruitment and sampling

This study employed a consecutive recruitment strategy, targeting women attending infertility clinics at the UUTH and St. Luke’s Hospital, Anua. Participants were enrolled upon meeting the predefined inclusion criteria, continuing until the target sample size for each facility was achieved. For each infertile case recruited, a control participant, a fertile woman of comparable age and not using hormonal contraception, was subsequently recruited from the family planning clinics of the respective hospitals.

The study’s objectives, procedures, and potential benefits were comprehensively explained to all prospective participants. Emphasis was placed on the voluntary nature of participation, the confidentiality of collected data, and the right to withdraw at any point without repercussions.

Reviewing departmental records at UUTH from January to June 2019 revealed 137 infertility clinic visits, averaging 23 visits per month. St. Luke’s Hospital Anua reported 75 visits during the same period, with a monthly average of 13. Combined, the expected monthly attendance for both sites was 36 (23 + 13).

Over a 4-month period, an average of about 92 (23 × 4) women attended the infertility clinics of UUTH, while 52 (13 × 4) attended those of St. Luke’s Hospital Anua, totaling 144 across both locations.

Participant allocation to each center was proportional to their respective attendance rates. The allocation factor was determined by dividing the minimum sample size by the total attendance: 142/144 = 0.986.

Consequently, UUTH recruited 92 × 0.986 = 91 participants, and St. Luke’s Hospital Anua recruited 52 × 0.986 = 51 participants, summing to the required 142 infertile participants.

Control participants were recruited from the family planning clinics of both hospitals. The average monthly attendance at these clinics from January to June 2019 was 40 at UUTH and 25 at St. Luke’s Hospital Anua. During the study period, 91 and 51 fertile women, respectively, matching the age and sociodemographic profiles of the study group, were consecutively recruited from these clinics.

Data collection instrument

Data were collected using a semi-structured, self-administered questionnaire. The questionnaire was pretested at UUTH’s gynecological clinic on clients who were not part of the study population, with their records marked to prevent double recruitment. The questionnaire comprised three sections: Section A for anthropometric and sociodemographic data; Section B for clinical, obstetric, and gynecological information; and Section C for prolactin assay results.

Trained research assistants, consisting of resident doctors, assisted in data collection at both study sites.

Inclusion criteria for the study group

Consenting women unable to conceive after 1 year of regular, unprotected intercourse.

Exclusion criteria for the study group

Women who were not infertile based on the definition, those who declined consent, women with known psychiatric disorders, and those using dopaminergic agonists or medications affecting prolactin levels.

Exclusion criteria for the control group

Women who did not consent to participate in the study, those using hormonal contraception, and lactating mothers.

Study protocol and sample collection

The participants’ weight and height were measured, and their body mass index (BMI) was calculated. Venous blood samples (5 mL) were collected from consenting women on the 2^nd or 3^rd day of their menstrual cycle following an 8–12 h fast, at least 30 min after their arrival to minimize stress-induced prolactin fluctuations. For women with amenorrhea or irregular cycles, random samples were collected after fasting, after excluding pregnancy. Samples were transported to UUTH’s Chemical Laboratory within 1 h, labeled, and centrifuged. Serum was separated, labeled, and stored at −10°C–−20°C until analysis.

Prolactin immunoassay

Serum prolactin levels were measured using a commercial enzyme-linked immunosorbent assay kit (Diagnostic Automation Inc.). A designated Chemical Pathologist at UUTH performed the analysis. The upper limit of normal prolactin was set at 19.5 ng/mL. Women with levels exceeding 100 ng/mL were advised to undergo imaging to rule out prolactin-secreting pituitary tumors.

Data analysis

Data were analyzed using STATA version 15. Categorical variables were summarized as frequencies and percentages and quantitative variables as means and standard deviations. The Wilcoxon rank-sum test compared median prolactin levels between groups. The Chi-square test assessed associations between selected factors and prolactin levels, with a significance level of P < 0.05.

Ethical considerations

Ethical approval was obtained from UUTH’s Research Ethical Committee. Participation was voluntary, and a written informed consent was obtained from all participants. Confidentiality was maintained using initials and serial numbers for identification.

Sociodemographic and clinical characteristics

Table 1 presents the sociodemographic and clinical characteristics of the participants. Significant differences emerged between the two groups. The control group (fertile women) exhibited a statistically significant higher mean age compared to the infertile group (P = 0.027). The average BMI was significantly higher in the infertile group (25.95 kg/m²) than in the control group (23.74 kg/m²), with a P < 0.001.

Marital status was significantly associated with the study group (P < 0.001). The median duration of relationships was also significantly longer in the control group (P = 0.002). Educational level and ethnic group were significantly associated with the study groups (P = 0.009 and P = 0.001, respectively).

Furthermore, parity, number of previous pregnancies, number of previous abortions, menstrual pattern, and the presence of galactorrhea were all significantly associated with the study group (P < 0.001 for each).

A notable disparity in median prolactin levels was observed between the study groups. Specifically, the infertile group exhibited a significantly elevated median prolactin level compared to the fertile control group, with values of 14.6 ng/mL and 6.4 ng/mL, respectively. This difference was statistically significant, as indicated by a P > 0.001. These data are detailed in Table 2, which presents the median prolactin levels for both groups, encompassing a total of 280 participants.

Within the subset of 44 infertile participants diagnosed with hyperprolactinemia, a distribution of severity was observed: 33 women (75.0%) had mild hyperprolactinemia, 9 (20.5%) presented with moderate hyperprolactinemia, and 2 (4.5%) had severe hyperprolactinemia. Conversely, among the seven fertile controls with hyperprolactinemia, six individuals (85.7%) had mild elevations, and one individual (14.3%) had moderate elevation. Notably, no fertile control participants exhibited severe hyperprolactinemia [Table 3].

Within the group of infertile participants, 18 individuals (13%) were classified as having primary infertility, while the majority, 122 (87.1%), had secondary infertility.

Table 4 shows the associations between various sociodemographic and clinical characteristics and the prolactin levels of the participants. Statistically significant associations were observed between infertility status and parity, menstrual pattern, presence of galactorrhea, and prolactin levels (P < 0.001, P < 0.001, P = 0.01, and P < 0.001, respectively).

Conversely, no significant associations were found between BMI, age, marital status, and prolactin levels (P = 0.91, P = 0.10, and P = 0.18, respectively). Similarly, ethnic group, number of previous pregnancies, and number of previous abortions did not demonstrate significant associations with prolactin levels (P = 0.33, P = 0.60, and P = 0.05, respectively).

Predictors of elevated prolactin levels: Binary logistic regression analysis

Table 5 shows the results of a binary logistic regression analysis, examining the predictors of elevated prolactin levels among the study participants.