We conducted this study in two medical facilities specialising in gynaecology and obstetrics in the city of Kisangani, in the Democratic Republic of the Congo. These were the Cliniques Universitaires de Kisangani (CUKIS) and the Clinique des Anges de Kisangani (CAKIS).

1) Study population: Our study population consisted of women aged 18 to 35 who consented to participate in this study.

2) Sample: Sampling in our study was voluntary non-probability, having recorded women aged 18 to 35 who had met the inclusion criteria. The sample size was calculated using Epi Info© 7.2.2.6 in its Stat Cal function. In order to compare the study factor (ovarian reserve) between female subjects with and without sickle cell disease, we allowed for a first order error of 5%, with an accuracy of 95%. The study sample consisted of 268 women. These women were categorised according to their haemoglobin status and according to the absence or presence of sickle cell crises, failing determination of the proportion of haemoglobin S in heterozygotes, into HbAA, asymptomatic HbAS (HbASa), symptomatic HbAS (HbASs) and HbSS, and then compared with each other.

In view of the common and comparable characteristics between the groups taken in pairs, the participants with HbAA were mixed with those with HbASa to form the control group. Those with HbSS were mixed with those with HbASs to form the patient group. Finally, the ovarian reserve was compared between two groups: a control group and a patient group.

For symptomatic homozygous and heterozygous sickle cell women:

For the comparison group:

All women in both groups were excluded from the study:

This was an analytical cross-sectional study conducted from 5 April to 31 October 2024.

Data collection was prospective. Data were collected using a previously prepared data collection form, then entered and encoded on the Excel spreadsheet program (Microsoft, CDC, 2010) before being analysed using R software version 4.4.2 (2024-10-31 ucrt). The sample was described in terms of frequencies and proportions. Proportions were compared using the chi-square test of independence, Fischer’s exact test at the P < 0.05 significance level. Quantitative variables were presented as means and standard deviations. Given the number of study arms, the means were compared by Student’s t test at the P < 0.05 significance level.

1) Dependent variable: Ovarian reserve AMH level and number antral follicles).

2) Independent variables:

Socio-demographic and clinical characteristics: Age, physical address, level of education, occupation, marital status, parity and gestational age, history of abortions, age at menarche and characteristics of the menstrual cycle (length, duration, quantity and frequency of menses), medical, surgical and alcoholic-smoking history, history of hormonal contraception, desire to conceive, use of ovulation inducers and pelvic X-ray, weight and height. In addition, we determined the frequency of vaso-occlusive crises and transfusions, haemoglobin levels, and treatment received or in progress in women with sickle cell disease.

3) Haemoglobin status

Reserve marker values [17] :

We compared markers of ovarian reserve between the different subgroups. To do this, serum levels antimüllerian hormone (AMH) and endovaginal ultrasound for counting antral follicles (CFA) were measured. The ultrasound was performed between the third and fifth day of the menstrual cycle. In addition to the above-mentioned ovarian reserve tests, the haemoglobin status of the participants was determined. In homozygous and heterozygous sickle cell patients, we also measured haemoglobin levels. AMH was measured by immunofluorescence using a Fluocare® quantitative immunofluorescence analyser, model MF-T 1000, SN: MF-T 1000-21120200. Haemoglobin status was determined using a hemotypeSC rapid test kit, designed to determine the presence of haemoglobin A, S and C in whole blood. Haemoglobin levels were determined using an automatic digital haemoglobimeter, BioAid HB (serial number 12000045526), after taking a drop of capillary blood from the pulp of the index finger. Endovaginal ultrasound was performed using an EDAN D3 ultrasound scanner, SN 317156-M 1661 1380005, 2016-06. Prior to folliculometry a pelvic ultrasound was performed to rule out ovarian pathology.

Before carrying out this study, we sought and obtained authorisation from the ethics committee at the University of Kisangani.

Each participant enrolled in the study received a clear and detailed explanation of the study. Those who agreed signed an informed consent form (in French or the local language) in which all the information relating to the study was recorded. All the information collected from the women was kept confidential. Security measures concerning confidentiality were guaranteed by anonymity during the collection, processing and analysis of the results, by limited access to the data, and by the impossibility of identifying the subjects when the results of the study were published.

Figure 1 shows the distribution of participants according to haemoglobin status and the presence of symptoms in heterozygotes (AS).

The following table shows that there was a significant difference between the two groups in terms history of abortion, with a P-value of 0.022 ( Table 1 ).

In the following table, there was a significant difference between the two groups in terms of age at menarche, family history of sickle cell anaemia, BMI and alcoholism ( Table 2 ).

The table below shows that AMH levels were significantly associated BMI (P-value = 0.005), the number of vaso-occlusive attacks (P-value = 0.041) and haemoglobin levels (P-value = 0.011) ( Table 3 ).

Table 3 shows that AMH levels were significantly associated with the number of haemolytic attacks (P-value = 0.008), treatment with folic acid (P-value = 0.042) and non-steroidal anti-inflammatory drugs (P-value = 0.009).

The number of antral follicles was significantly associated with BMI (P-value = 0.006), the number of vaso-occlusive attacks (P-value = 0.003) and Hb level (P-value = 0.006) ( Table 4 ).

The number of antral follicles was significantly associated with the number of haemolytic attacks (P-value = 0.011) and folic acid treatment (P-value = 0.041) ( Table 5 ).

There was a statistically significant difference between the control group and the patients with to AMH level (P-value = 0.001) and NFA (P-value = 0.001) ( Table 6 ).

Ovarian reserve in the patient group was significantly reduced compared with the control group ( Table 7 ).

The reduction in ovarian reserve was associated BMI (P-value = 0.008) and the number vaso-occlusive attacks (P-value = 0.013) ( Table 8 ).

Table 8 shows that the decrease in ovarian reserve was associated with the number of haemolytic attacks (P-value = 0.006) and the Hb level (P-value = 0.009).

In our study, most patients (HbSS and HbASs) had a low BMI (≤18.5 kg/m²). Our results corroborate those of M’Pemba-Loufoua, who states that patients with sickle cell disease tend to have a low BMI because of the chronicity of the disease, which leads to delayed growth in height and weight [18] .

Our study showed that there was a statistically significant difference between the control group and the sickle cell group in terms of history of abortion, mean age at menarche and BMI.

A history abortion was more common in the control group than in the patients. Although we did not determine the presence of a sexual partner or the frequency of sexual intercourse among the respondents, we think that the sickle-cell condition could confer a low sexual frequency on sickle-cell patients, which could explain the low rate of abortions among patients. In addition, given the average ages of the two groups, most of them being single and unemployed, we think that the majority of abortions could be induced, a risk that sickle cell sufferers would take less, compared with non-sickle cell sufferers.

The mean age of menarche in the sickle-cell group was 15.5 years, compared with 13 years in the control group, i.e. the sickle-cell group had an average delay of two and a half years. Our results were similar to those found by M’Pemba-Loufoua et al. in Brazzaville, who found that sickle cell patients had a delay in pubertal development of at least two years compared with non-sickle cell patients [18] , those of Platt et al. in the United States [19] , Oyedeji et al. in Nigeria [20] and Betoko et al. in Cameroon [21] who estimated a delay of 3 years.

The normal development of secondary sexual characteristics is the result of activation of the hypothalamo-pituitary-gonadal axis. However, this activation requires a sufficient supply of energy (leptin), determined by the abundance of adipose tissue and with an ideal weight (pondérostat) [22] . We think that the staturo-ponderal deficit in sickle-cell patients could therefore justify the delay in onset of menarche found in our study.

In our study, sickle cell patients had a lower average BMI than controls. The same trend was found by M’Pemba-Loufoua et al. in Brazzaville [18] . We believe that this could be due to chronic anaemia, resulting in increased diversion nutrients to active haematopoiesis, causing hypoproteinemia, which is detrimental to staturo-weight growth in sickle cell patients [23] .

We observed that AMH levels significantly associated with BMI (P-value = 0.005), the number of vaso-occlusive (P-value = 0.041) and haemolytic (P-value = 0.008) attacks, haemoglobin levels (P-value = 0.011) and treatment with folic acid (P-value = 0.042) and non-steroidal anti-inflammatory drugs (P-value = 0.009). In terms of BMI, 95% of patients with low AMH levels compared with 42.6% with normal levels had a BMI less than or equal to 18.5 kg/m². Patients with more than 10 vaso-occlusive attacks accounted for 75% of patients with low AMH, compared with 33.7% with normal AMH; the same trend was observed for haemolytic attacks. 70% of patients with low AMH compared with 42.6% with normal AMH had Hb level of less than 7 g%. 72.2% of patients with normal AMH were on folic acid compared with 45% with low levels. For patients with NSAIDs, 80% had a low AMH level, compared with 40% with a normal level.

A normal BMI is an expression of good nutritional status, suggesting an energy reserve of lipids available for metabolism. At ovarian level, lipids drive follicular growth by making energy substrates available after β-oxidation, for meiosis and cytoplasmic maturation of follicular cells. This could justify the relationship between BMI and AMH levels in our study [24] .

Our study showed that patients who had more vaso-occlusive and haemolytic attacks had low AMH levels. Our results corroborate the argument put forward by Chase et al. in a clinical case report, proposing that frequent episodes intravascular sickle cell disease, vascular occlusion and infarction, as well as tissue hypoxia associated with chronic anaemia, could explain the ovarian dysgenesis, which could justify the low AMH levels in our series [25] .

Haemosiderosis induced by multiple haemolytic crises can affect oocyte quality through the accumulation of reactive oxygen species (free radicals) generated by free iron, creating increased oxidative stress. As, the window of the iron concentration range in biological systems is narrow for optimal beneficial effects with minimised adverse consequences [26] . There is a delicate balance between reactive oxygen species and antioxidant enzymes in ovarian tissues. Antioxidant enzymes neutralise the production of reactive oxygen species and protect the follicles against ageing. The granulosa and luteal cells respond to hydrogen peroxide with a loss of gonadotropin action and inhibition of progesterone secretion [27] . We believe that this oxidative stress, associated with a chronic inflammatory state in sickle cell disease, could be the reason for impaired ovarian reserve in sickle cell disease. Chronic anaemia in sickle cell disease leads to hypoxaemia, resulting in ovarian hypoxia, which could justify premature ovarian dysgenesis and failure manifested by a decrease in AMH levels in sickle cell disease patients [25] .

Our study showed that most patients with low AMH levels were taking NSAIDs. Although the effects of NSAIDs on ovulation inhibition are well described in the literature [28] [29] , we have not been able to access any literature determining the effects of NSAIDs on ovarian reserve. Nevertheless, as NSAIDs inhibit the production of prostaglandins, we believe that inhibition of the latter would deprive the ovary of the beneficial effect of prostaglandins on the ovarian vascular system, namely vasodilatation and inhibition of platelet aggregation [30] . This will result in a reduction in ovarian blood flow, which may justify ovarian dysgenesis.

In our study, we found that most women with normal AMH levels were taking folic acid. In fact, a Polish cohort study had revealed that women who had received folic acid supplementation prior to infertility treatment by medically assisted reproduction (MAP) had better quality follicles and a higher degree of mature follicles than women who had not folic acid [31] . We therefore believe that this improvement in follicle quality could be the reason for normal AMH levels in some women with sickle cell disease.

We observed that the number of antral follicles was significantly associated with BMI (P-value = 0.006), the number of vaso-occlusive attacks (P-value = 0.003), the Hb level (P-value = 0.006), the number of haemolytic attacks (P-value = 0.011) and folic acid treatment (P-value = 0.041). The number of antral follicles (ANF) tended to improve as BMI and haemoglobin improved, with a reduction in vaso-occlusive and haemolytic attacks. We believe that this would be justified in the same way as the AMH level.

In our series, there was a statistically significant difference between the control group and the patients with regard to AMH level (P-value < 0.001) and NFA (P-value < 0.001). There were 5.1% in the control group with low AMH compared with 22.5% in the patients. 2.8% in the control group had a low NFA compared with 53.9% among patients.

Taking AMH and NFA together, thus defining ovarian reserve in our series, decreased ovarian reserve (DOR) was found in 1.7% of the control group, compared with 20.2% of patients. This difference was statistically significant (P-value < 0.001). In patients, ovarian reserve was associated with MC (P-value = 0.008), the number of vaso-occlusive crises (P-value = 0.013), the number of haemolytic crises (P-value = 0.006) and the Hb level (P-value = 0.009). 94.4% of patients with reduced ovarian reserve versus 62% of patients without reduced ovarian reserve had a BMI ≤ 18.5 kg/m² and the trend was the opposite with a BMI greater than 18.5 kg/m². 11.1% of patients with reduced ovarian reserve had fewer than 5 vaso-occlusive attacks, compared with 38% without reduced ovarian reserve, and with more than 10 vaso-occlusive attacks, 77.8% of patients with reduced ovarian reserve had more than 10, compared with 36.6% of patients without reduced ovarian reserve.

The same trend was observed with haemolytic crises. In terms of Hb levels, the majority of patients with a reduced ovarian reserve (72.2%) had Hb below 7, compared with 38% without a reduced ovarian reserve.

These results corroborate those found in the literature, indicating that there were more sickle cell women with low ovarian reserve compared with non-sickle cell women of the same age [1] [8] [11] [13] . This can be explained by the pathophysiology of sickle cell disease and its management as described above (vaso-occlusive and haemolytic crises, haemosiderosis, chronic anaemia and the use of non-steroidal anti-inflammatory drugs).