Anti-Müllerian hormone versus antral follicle count as first-choice biomarkers in a low-resource setting: A cross-sectional study in Kumasi,Ghana

Abstract

Objective

To evaluate age-related variations in ovarian reserve among women in Kumasi, Ghana, and determine the preferred first-choice biomarker, either anti-Müllerian hormone or antral follicle count, in a low-resource setting.

Methods

A cross-sectional study was conducted at a private fertility hospital in Kumasi from July 2023 to June 2024. A total of 372 women undergoing fertility assessment were enrolled. Both anti-Müllerian hormone and antral follicle count were measured, stratified by age group, and used to generate age-related nomograms. Correlations between age and each biomarker were assessed.

Results

The median age was 23 years for oocyte donors and 33 years for self-cycle patients, with 80% classified as overweight/obese. Anti-Müllerian hormone and antral follicle count showed polynomial, nonlinear declines with age, peaking in the mid-20s, and approaching 0 by 45 years of age. Diminished ovarian reserve was observed at 38 years by antral follicle count and at 40 years by anti-Müllerian hormone. This disparity was also reflected in the poor agreement between antral follicle count and anti-Müllerian hormone in classifying participants as having normal ovarian reserve or diminished ovarian reserve (kappa = 0.167). Antral follicle count detected more diminished ovarian reserve cases (38%) than anti-Müllerian hormone (33%) and demonstrated a stronger negative correlation with age (r = −0.34, p = 0.00) than anti-Müllerian hormone (r = −0.30, p = 0.00). The biomarkers were positively correlated (r = 0.42, p = 0.00).

Conclusions

Ovarian reserve decline among Ghanaian women appears to be more gradual, leading to diminished ovarian reserve at least 3 years later than in reference models. Antral follicle count showed higher sensitivity and a stronger correlation with age, supporting its use as the preferred first-line biomarker for decision-making in low-resource settings.

Keywords

Age ovarian reserve anti-Müllerian hormone antral follicle count Ghana

Introduction

Ovarian reserve (OR) is defined as the number of oocytes remaining in a woman’s ovary that retain the potential to be fertilized and develop into an embryo¹ Although age is considered the primary determinant of OR, serum anti-Müllerian hormone (AMH) and transvaginal antral follicle count (AFC) are the most commonly used markers for its quantitative assessment. These measures are used to predict women’s response to controlled ovarian stimulation (COS) in assisted reproductive technology (ART).

OR testing, a proxy for ovarian response to COS or for detecting diminished ovarian reserve (DOR), is becoming increasingly relevant in ART for predicting treatment outcomes.² Predicting ovarian response to COS and ART treatment outcomes enables clinicians to choose individualized COS protocols. A woman’s age, AMH levels, and AFC are considered the most reliable predictors of ovarian response to COS,³ although some studies have reported that the ovarian response prediction index (ORPI), which combines age, AMH levels, and AFC, are superior to each biomarker alone.⁴

Conventionally, an AMH level of 1.2 ng/mL or an AFC of 5 is considered the normal lower threshold for OR.⁵ Based on these thresholds, AMH results are typically classified as indicating “low,” “normal,” and “high” OR or expected ovarian response.¹ However, these reference thresholds for normal AMH levels or AFC have not been universally accepted. Some studies have established age-specific reference values for AMH and AFC that may be more useful.^6,7

Conventional models of age-related variation in AMH show a polynomial pattern, with an accelerated decline beginning in the mid-30s and an increased prevalence of DOR from 35 years of age.⁸ Although some studies have reported a linear age-dependent pattern of change for AFC,⁷ others have described a biphasic decline.⁹ Notably, these findings suggest that AMH and AFC nomograms may vary across populations and should be determined for specific populations to facilitate more precise, patient-specific, individualized treatment plans.¹⁰ Mawusi et al.¹¹ determined age-related variation in AMH levels in Accra. However, no data exist on age-specific AFC in Ghana.

Many studies have suggested that AMH levels and AFC have comparable sensitivity and specificity for detecting OR status and similar predictive ability for ovarian response in in vitro fertilization (IVF)/intracytoplasmic sperm injection (ICSI). Therefore, these markers can be used interchangeably in clinical practice.^12–14 However, AFC is cheaper, easier to perform, and less invasive.¹⁵

The patterns of age-specific variation and reference values for AMH levels and AFC in the Ghanaian population need to be determined, as relatively few studies have investigated these markers in African ART settings.

This study aimed to ascertain the pattern of variation in OR with age and the correlation between AMH levels and AFC in a low-resource setting (Kumasi). Establishing the correlation between AMH levels and AFC in this study will provide preliminary data for context-specific biomarker prioritization in a low-resource decision-making setting. The evidence will serve as a reference for clinicians in determining tailored COS protocols for Ghanaian women.

Material and methods

Study setting and design

This study was conducted at Hallmark Medicals (HM), a private hospital at Ahensan Estate in the Asokwa Municipal of Ashanti Region, Ghana. The study site was selected based on convenience. HM is a specialist general hospital with ART as its flagship service. It offers medical, surgical, child health, obstetrics, and gynecology services. The hospital provides almost all primary ART services, including IVF or ICSI and embryo transfer (ET) (IVF–ET or ICSI–ET), egg and embryo cryopreservation, sperm freezing, gamete donor services, and surrogacy.

The hospital’s ART clientele includes local patients, cross-border Ghanaians, and international patients. The hospital runs an ART clinic twice weekly, with an average of 3–4 new infertility cases per clinic day. A total of 12–15 patients are registered at the clinic on each clinic day.

This cross-sectional study was conducted among normally menstruating women aged 18–42 years who presented to HM, Kumasi, between 1 July 2023 and 21 June 2024, either to achieve pregnancy or to voluntarily and anonymously donate oocytes for the first-time for ART treatment of women with infertility. Written informed consent was obtained from all eligible women.

Women >42 years of age were excluded from blood sampling for AMH measurement or AFC nomogram development because, conventionally, most women in this category are excluded from self-cycle IVF (IVF with autologous oocytes) programs because of expected poor OR.¹⁶ Women using hormonal contraception; those diagnosed with amenorrhea/anovulation; and those who had previously undergone COS for IVF/ICSI, commissioned surrogacy, or had a history of premature ovarian insufficiency, chemotherapy, or radiotherapy were also excluded.

Sample size and sampling technique

Sample size

The prevalence of infertility among women in Ghana was estimated at 15%.¹⁷ Using Cochran’s formula for calculating sample size for a simple proportion, with 5% desired precision, a 95% confidence interval (CI), and a 10% non-response rate, the minimum required sample size was estimated to be 216. However, 372 women with infertility or those attending the clinic to donate their oocytes for the treatment of women with infertility consented to participate in the study.

Sampling

The study participants were conveniently selected. The research team reviewed HM’s hospital records daily during the study period to identify eligible participants.

Ethical considerations

The study was conducted in accordance with the principles of Declaration of Helsinki and regulations for data collection. Ethical approval (CHRPE/AP/468/23), dated 27 June 2023, was obtained from the Committee on Human Research and Publication Ethics (CHRPE) of the Kwame Nkrumah University of Science and Technology (KNUST), Kumasi. A patient information leaflet was used to explain the study’s objectives, risks, and benefits to each potential participant in a language they understood. Participants were assured of anonymity and confidentiality and were informed of their right to withdraw from the study. Additionally, they were informed about available alternatives for obtaining quality care and appropriate channels for complaints and redress. Only participants who voluntarily provided written informed consent, either by signature or thumbprint, were enrolled. A written informed consent obtained from at least the female partner of a couple was considered sufficient (Clinical trial number: not applicable).

Data collection

Demographics

The demographic characteristics of all eligible participants were collected on the first day of presentation. These included age, religion, employment status, marital status, weight, and height. Each participant’s weight and height were measured using a calibrated scale and stadiometer. Body mass index (BMI), calculated as weight (in kilograms) divided by the square of height (in meters), was determined.

Assessment of OR

OR was assessed using transvaginal AFC and serum AMH measurements.

AFC

Follicles measuring 2–10 mm in the ovary were considered antral follicles. AFC was determined at HM using a 7.5-MHz transvaginal ultrasound probe attached to a Mindray ultrasound machine (Shenzhen, Guangdong, China) following standard procedures to count and sum the antral follicles in each ovary. The Mindray ultrasound machine used was a battery-powered portable DP series model with real-time imaging, digital storage and reporting, and a user-friendly interface, making it highly suitable for a resource-limited setting. The transvaginal ultrasound scan for AFC assessment was performed prior to pituitary downregulation and COS, independent of the day of the menstrual cycle. All AFC assessments were performed by a single reproductive endocrinology and infertility specialist.

AMH measurement

Venous blood samples for AMH level determination were obtained on-site at HM, independent of the day of the menstrual cycle and as part of the initial investigations. Blood samples were collected using standard procedures and in accordance with standard precautions for venipuncture. Each sample bottle was labeled with the participant’s unique identification number. The collected samples were immediately processed and analyzed using a Finecare FIA Meter Plus device (Guangzhou Wondfo Biotech, China) in conjunction with the Finecare™ AMH rapid quantitative test fluorescence immunoassay, following the manufacturer’s sample preparation and analysis instructions. Quality control materials simulating the characteristics of patient samples were included every 24 h to monitor the performance of immunological assays, as samples were processed in a random “access” format rather than a “batch” format. The frequency of quality control was in accordance with good laboratory practice and laboratory accreditation requirements.

Data management and analysis

Each day, all data collection checklists were examined for completeness and consistency. All participant information was deidentified, coded, and edited. The edited data were entered into Excel and transferred to STATA version 17 and R statistical software for analysis.

Continuous demographic variables, such as age and BMI, were summarized using medians and interquartile ranges (IQRs) or means and SDs, whereas categorical variables, such as employment status and religion, were summarized using frequencies and percentages. The Kruskal–Wallis or Wilcoxon rank-sum tests were used to compare medians of continuous variables, with the latter used for non-normally distributed continuous variables between oocyte recipients and self-cycle groups. The chi-squared test or Fisher’s exact test was used to compare categorical outcomes and associations between groups, with Fisher’s exact test used when expected cell counts were small.

AFC and AMH nomograms were generated using a natural cubic spline with the model formula ns (age, 3) and the model:

A M H_{i} = β_{0} + β_{1} B_{1} (a g e_{i}) + β_{2} B_{2} (a g e_{i}) + β_{3} B_{3} (a g e_{i}) + ϵ_{i}

where,

β_{0}

is an intercept;

β_{1}, β_{2}, {and β}_{3}

are spline coefficients;

B_{1}, B_{2}, {and B}_{3}

are basis functions generated by the natural cubic spline; and

ϵ_{i} \sim N (0, σ^{2})

The CI was represented by the colored border along the spline in each nomogram.

Agreement between AFC and AMH for predicting normal ovarian reserve (NOR) versus DOR was assessed using a confusion matrix and Cohen’s kappa coefficient. All statistical analyses were performed using R software (version 4.2.1). Appropriate corrections for multiple comparisons were applied to control for Type I errors. All tests were conducted at a 5% significance level with 95% CIs. Kappa values were interpreted as follows: <0.20, poor; 0.21–0.39, fair; 0.40–0.59, moderate; 0.60–0.79, good; and >0.80, excellent.¹⁸ Missing values were excluded from the analyses. The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.¹⁹

Results

Socio-demographic characteristics

A total of 372 women attended the fertility clinic during the study period. Of these, 73 dropped out of treatment, whereas 98 voluntarily presented to anonymously donate their oocytes for the ART treatment of women with infertility. Table 1 presents the socio-demographic characteristics of the participants, including age, BMI, employment status, religion, and marital status.

Table 1.

Socio-demographic characteristics of women seeking infertility treatment in Kumasi.

Characteristic	Overall, N = 372^a	Treatment dropouts, N = 73^a	Oocyte donors, N = 98^a	Oocyte recipients, N = 142^a	Self-cycles, N = 59^a	p value^b
Age	34 (26, 39)	34 (30, 37)	23 (21,25)	40 (38, 44.8)	33 (29.5, 35.5)
Employment status
Government employed	115 (32%)	32 (44%)	4 (4.5%)	51 (37%)	28 (47%)
Private employed	47 (13%)	10 (14%)	7 (7.9%)	23 (17%)	7 (12%)
Self-employed	134 (38%)	22 (31%)	36 (40%)	55 (40%)	21 (36%)
Unemployed	61 (17%)	8 (11%)	42 (47%)	8 (5.8%)	3 (5.1%)
Religion
Christian	80 (89%)	0 (NA%)	80 (90%)	0 (NA%)	0 (0%)
Muslim	10 (11%)	0 (NA%)	9 (10%)	0 (NA%)	1 (100%)
Marital status
Cohabiting	6 (1.6%)	3 (4.2%)	0 (0%)	2 (1.4%)	1 (1.7%)
Married	261 (70%)	68 (94%)	0 (0%)	136 (96%)	57 (97%)
Single	104 (28%)	1 (1.4%)	98 (100%)	4 (2.8%)	1 (1.7%)
BMI	28.52 (25.39, 31.79)	29.30 (25.39, 32.87)	27.18 (24.25, 30.09)	28.87 (25.66, 31.89)	29.30 (26.79, 33.20)	0.002
BMI category
Underweight	3 (1.0%)	0 (0%)	3 (3.1%)	0 (0%)	0 (0%)
Normal	81 (22%)	14 (19%)	29 (30%)	27 (19%)	11 (19%)
Overweight	139 (37%)	24 (33%)	40 (41%)	54 (38%)	21 (36%)
Obese	149 (40%)	35 (48%)	26 (27%)	61 (43%)	27 (46%)

^aMedian (IQR); n (%).

^bKruskal–Wallis rank-sum test.

BMI: body mass index; NA: not applicable; IQR: interquartile range.

Overweight and obesity emerged as significant characteristics, with only 81 (22%) of the participants having a normal BMI. Almost 4 out of 5 participants were either overweight (37%) or obese (40%). Participants in the self-cycle group had a significantly higher median BMI of 29.30 (26.79, 33.20) kg/m² than the oocyte recipients (28.87 (25.66, 31.89) kg/m²; p = 0.002). Additionally, they were more likely to be obese (46%) compared with the oocyte recipients (43%). However, oocyte donors were more likely to have a normal BMI, with a median BMI of 27.1 kg/m².

Nomogram for AMH and AFCs with age

The nomograms of serum AMH and transvaginal AFC demonstrated initial increases in OR with age from the early 20s, peaking at 25 years of age. This increasing trend was followed by a gradual, geometric decline toward 0 by 45 years of age. However, transvaginal AFC identified DOR, defined as a transvaginal AFC of less than 5 or a serum AMH level of less than 1.2 ng/mL, 2 years earlier, at age 38 years, compared with serum AMH at age 40 years (Figures 1 and 2).

Figure 1.

Variation in serum anti-Müllerian hormone (AMH) levels with age. The lower limit of normal AMH level was considered to be 1.2 ng/mL.

Figure 2.

Variation in antral follicle count (AFC) with age. The lower limit of normal follicle count was considered to be 5.

ORPI

Figure 3 shows a spline plot of the ORPI over age. It followed a trajectory similar to that observed for AMH and AFC.

Figure 3.

Variation in the ovarian response prediction index (ORPI) with age.

Correlation between AMH and total AFC among women seeking infertility treatment in Kumasi

Both serum AMH levels (r = −0.30, p < 0.001) and transvaginal AFC (r = −0.34, p < 0.001) demonstrated a weak negative nonlinear correlation with age. However, AFC had a slightly stronger correlation with age than AMH (Figures 4 and 5). The age distribution in the cohort appeared approximately normal. AFC and AMH showed a moderate positive linear correlation (r = 0.42, p < 0.001) (Figure 6).

Figure 4.

Correlation between anti-Müllerian hormone (AMH) and age (correlation coefficient r = −0.30, p < 0.001).

Figure 5.

Correlation between antral follicle count (AFC) and age (correlation coefficient r = −0.34, p < 0.001).

Figure 6.

Correlation between anti-Müllerian hormone (AMH) and antral follicle count (AFC) (correlation coefficient r = 0.42, p < 0.001).

Cohen’s kappa coefficient confirmed the disparity (poor agreement) between the two biomarkers in classifying participants as having NOR or DOR (Table 2), with a kappa value of 0.167. However, in 68% of cases, transvaginal AFC correctly identified women without DOR. The novelty of this study lies in the superior performance of transvaginal AFC as an ovarian reserve biomarker compared with serum AMH, as summarized in Table 3.

Table 2.

Agreement and diagnostic performance of AFC using AMH as the reference standard.

AFC vs. AMH	AMH: DOR	AMH: NOR	Total
AFC: DOR	38	50	88
AFC: NOR	39	106	145
Total	77	156	233
Performance metric	Estimate	95% CI/p value
Accuracy	0.618	0.552, 0.681
Sensitivity	0.494	—
Specificity	0.680	—
PPV	0.432	—
NPV	0.731	—
Balanced accuracy	0.587	—
Kappa	0.167	—
McNemar’s test	—	p = 0.289
No Information rate	0.670	p = 0.958

AFC: antral follicle count; AMH: anti-Müllerian hormone; CI: confidence interval; DOR: diminished ovarian reserve; NOR: normal ovarian reserve; PPV: positive predictive value; NPV: negative predictive value.

Table 3.

Principal findings of the study.

No:	Principal findings
1	Almost 4 out of 5 attendants at the fertility clinic were either overweight (37%) or obese (40%)
2	A moderately strong positive linear correlation exists between serum AMH and transvaginal AFC as biomarkers
3	Transvaginal AFC identifies DOR 2 years earlier at 38 years than serum AMH, which does so at 40 years
4	A poor level of agreement exists between the two biomarkers, serum AMH and transvaginal AFC, in classifying attendants into NOR and DOR

AFC: antral follicle count; AMH: anti-Müllerian hormone; DOR: diminished ovarian reserve; NOR: normal ovarian reserve.

Discussion

Socio-demographic characteristics

The median age of treatment dropouts was 34.0 (30.00, 37.00) years, which was closer to the median age of the participants in the self-cycle group (33.00 (29.50, 35.50) years) than that of the oocyte recipients (40 years). Given the similarity in age, the treatment dropouts could be assumed to have been mostly self-cycle participants. The reasons for discontinuation were not examined in this study. However, treatment cost prohibition, failure to fully appreciate the impact of the age-dependent decline in fertility and its implication for the limited reproductive lifespan, or the absence of partner approval could account for the dropout rate. In a recent study on barriers to ART treatment in Ghana, Damalie et al.²⁰ found that prohibitive treatment costs and partner non-availability were attributable to ART treatment dropout.²¹ Moreover, evidence from Nigeria, a country with social and cultural characteristics similar to Ghana, suggest that women who reported adequate awareness of the age-dependent decline in OR tended to overestimate the age at which female fertility declines.²² Lack of awareness of age-related fertility decline has been associated with involuntary childlessness.²³

The oocyte donors in this cohort were relatively younger than those in other studies. Although the median age of oocyte donors in this cohort was 23 years, Ben-Nagi et al.²⁴ reported a median age of 26 years. Differences in the social and economic contexts of the populations may account for this age disparity. Although this study did not explore motivations for voluntary and anonymous oocyte donation, financial reward was likely a significant factor owing to the donors’ youth and unemployment status. This assumption may not be far from the truth because, although financial reward has been reported to be secondary to altruism as a motivation for oocyte donation in other regions,^25,26 Okafor et al.²⁷ found that 93.4% of young Nigerian oocyte donors donated primarily for financial gain.

The small number of unmarried women seeking pregnancy through ART was notable. However, this was not due to legal restrictions, as seen in some other regions. Although single women in different parts of the world face legal hurdles in achieving pregnancy through IVF using donor sperm,²⁸ many women in Ghana pursue this option only within the context of a familial relationship.²¹ Four of the five single women who underwent ART treatment in this cohort were oocyte recipients and older women who had likely become less hopeful of finding a husband or had experienced divorce from a previous relationship because of childlessness.

Notably, only a limited number of fertility clinic participants (one in five) had a normal weight. There was an almost equal likelihood that a woman with infertility was overweight (37%) or obese (40%). These findings are consistent with the global trend of increasing obesity associated with infertility, particularly anovulatory infertility.²⁹

Nomogram for AMH and AFCs over age

The patterns of variation in AMH and AFC with age were nonlinear or polynomial, with an initial rise in both AMH levels and AFC from the early 20s, peaking in the mid-20s, and declining toward 0 by 45 years of age. The trajectory of decline was similar for both OR markers. However, the trajectory of decline in AMH and AFC with age differed from findings reported in other studies. For example, Jain et al.³⁰ and Jain et al.³¹ reported a linear decline in AFC with age, whereas Jain et al.³⁰ reported a nonlinear (polynomial) decline for AMH, similar to that observed in the present study. The disparity in AFC findings may be attributable to differences in transvaginal ultrasound expertise. Similarly, Loy et al.⁷ reported a linear decline pattern for both AFC and AMH among Chinese women in Singapore. Some authors have also described a quadratic function as the best fit for age-dependent variation in AMH.³² Most notably, the ages at which DOR was observed in this study, as measured by AFC and AMH, were 38 and 40 years, respectively. This finding contrasts sharply with reports describing a rapid decline in AMH to a nadir at 35–36 years of age.³³ These findings are highly significant and may have implications for clinical practice and reproductive health counseling. However, the results should be interpreted with caution because they were derived from a single hospital. Differences in operator expertise may also have contributed to the observed findings.

Sensitivity of AMH and AFC as OR markers

The pattern of age-related variation in OR measured by AMH and AFC was similar, although the slope of decline was slightly steeper for AFC. Consequently, DOR was identified at 38 years of age using AFC, 2 years earlier than when determined using AMH. In the same cohort, AFC identified more women with DOR (38%) than AMH (33%). Moreover, AFC demonstrated a slightly stronger negative nonlinear correlation with age (r = −0.34, p = 0.00) than AMH (r = −0.30, p = 0.00) and a moderate positive correlation with AMH (r = 0.42, p = 0.000). Therefore, the evidence from this study suggests that AFC is a more sensitive marker of OR than AMH. These findings contrast with those of Arvis et al.,³⁴ who reported that AMH demonstrated a stronger negative correlation with age than AFC. However, their study included a smaller sample size. Another study also concluded that AMH is a more accurate predictor of OR.³⁵ Meanwhile, Arvis et al.³⁴ found very poor correlation between AMH and AFC among poor responders. The findings of the current study imply that transvaginal AFC can be considered a credible alternative to AMH. These findings are consistent with those of Anuradha et al.,³⁶ who reported a strong correlation between AFC and AMH and advocated the use of AFC as a first-line biomarker for OR testing in poor responders.

Recommendations

Reproductive health professionals should provide targeted education on the high prevalence of overweight and obesity in Kumasi, their effects on fertility, and strategies for maintaining a normal BMI. Because transvaginal AFC is cheaper, more readily available, noninvasive, and potentially more sensitive than AMH, with which it has a moderate correlation, it should be considered a sufficient first-line option in the evaluation of women with infertility, particularly in low-resource settings. Combining AFC with AMH may not be necessary for evaluating fertility potential. Further studies are needed to determine IVF/ICSI success rates among women in their late 30s to strengthen the findings of this study that women in this age group may retain a good chance of achieving a biological child through ART.

Limitations of the study

Data were collected from a single private fertility hospital. Additionally, study participants were conveniently selected and categorized into oocyte donor, self-cycle, and oocyte recipient groups. This sampling approach is prone to selection bias. Therefore, these participant groups may not adequately reflect age-related OR patterns in the general population. Moreover, the results are subject to an increased risk of confounding and potential over- or underestimation of outcomes. This limits the generalizability of the findings to the broader population of patients undergoing ART treatment in Kumasi or Ghana. Consequently, the results should be interpreted with caution.

Conclusion

Many of the study findings differed markedly from conventional knowledge and may serve as important references to guide ART practice in Ghana. The study provides local data for a population that remains understudied in the ART context. These data will serve as an important resource for comparing findings from Ghana with those from other populations to identify similarities and differences.

In this cohort, oocyte donors were predominantly single, unemployed Christians in their early 20s. Although very few single women presented to the clinic with the desire to have a child, potential self-cycle patients appeared more likely to withdraw from treatment than oocyte recipients. Overweight and obesity emerged as important demographic characteristics among women with infertility in Kumasi.

Particularly, the study established novel age-specific nomograms for AMH and transvaginal AFC and evaluated the correlation between these OR markers in Kumasi.

Among women attending fertility consultations in Kumasi, OR peaked in the mid-20s and declined progressively toward 0 by 45 years of age. The decline in OR was more gradual or geometric, contrasting with the sharper decline described in conventional models from other regions. Notably, the age at DOR among women in Kumasi was 38–40 years, at least 3 years later than that predicted by conventional models from other parts of the world. AFC also showed a stronger correlation with age than AMH and a moderate correlation with AMH for determining OR. In addition, AFC identified more women with DOR and did so at an earlier age than AMH. These findings suggest that transvaginal ultrasound measurement of AFC may be a more sensitive OR test than AMH in this cohort.

These novel findings provide a new perspective on the trajectory of OR and contribute to our understanding of fertility patterns in Ghana. The findings suggest that transvaginal ultrasound assessment of AFC may suffice for OR evaluation in Kumasi, a low-resource setting. Combining multiple OR markers to improve prediction of ovarian response may not always be necessary.

Ethics approval and informed consent

The guidelines and regulations outlined in the Declaration of Helsinki for data collection were adhered to. Ethical approval (CHRPE/AP/468/23) was obtained for this study from the Committee on Human Research and Publication Ethics (CHRPE) of the Kwame Nkrumah University of Science and Technology (KNUST), Kumasi. A patient information leaflet was used to explain the study’s purpose, risks, and benefits to each potential participant in a language they understood. Participants were assured of anonymity and confidentiality and informed of their right to withdraw from the study, available alternatives for obtaining quality care, and appropriate channels for complaints and redress. Only participants who voluntarily provided written informed consent by signature or thumbprint were recruited. A signed informed consent obtained from at least the female partner of a couple sufficed.

Footnotes

Acknowledgments

The authors express their gratitude to all study participants and to the management and staff of Hallmark Medicals, Ahensan Estate, Kumasi, Ghana, for their support in the successful conduct of this study.

Author contributions

Conception of the Research Proposal: Francis Jojo Moses Kodzo Damalie, Kwaku Bedu-Addo, Charles Mawunyo Senaya, Kofi Owusu-Daaku, and Patrick Addo-Fordjour

Project Development and Data Collection, Analysis and Interpretation: Francis Jojo Moses Kodzo Damalie, Kwaku Bedu-Addo, Edward Tieru Dassah, Alexander Anning, Phillip Kweku Baidoo, Kofi Owusu-Daaku, and Patrick Addo-Fordjour

Manuscript Writing: Francis Jojo Moses Kodzo Damalie, Kwaku Bedu-Addo, Charles Mawunyo Senaya, Edward Tieru Dassah, Alexander Anning, Phillip Kweku Baidoo, Kofi Owusu-Daaku, and Patrick Addo-Fordjour

All authors read and approved the final manuscript.

Consent for publication

Not applicable.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declaration of conflict of interests

The authors report no conflicts of interest in this work.

Funding

No external funding was either sought or obtained for this study.

ORCID iD

Francis Jojo Moses Kwadzo Damalie

References

Practice Committee of the American Society for Reproductive Medicine. Testing and interpreting measures of ovarian reserve: a committee opinion. Fertil Steril 2020; 114: 1151–1157.

Salemi

Jambarsang

Kheirkhah

, et al. The best ovarian reserve marker to predict ovarian response following controlled ovarian hyperstimulation: a systematic review and meta-analysis. Syst Rev 2024; 13: 303.

Scheffer

JAB

Scheffer

, et al. Are age and anti-Müllerian hormone good predictors of ovarian reserve and response in women undergoing IVF? JBRA Assist Reprod 2018; 22: 215–220.

Maged

Amin

Askalani

, et al. Comparative study between ovarian response prediction index versus anti-Müllerian hormone, antral follicular count as predictors of ovarian response in women undergoing IVF/ICSI cycles. Evid Based Womens Health J 2021; 146–156.

Humaidan

Alviggi

Fischer

, et al. The novel POSEIDON stratification of ‘Low prognosis patients in Assisted Reproductive Technology’ and its proposed marker of successful outcome. F1000Res 2016; 5: 2911.

Gunasheela

Murali

Appaneravanda

, , et al. Age-specific distribution of serum anti-Müllerian hormone and antral follicle count in Indian infertile women. J Hum Reprod Sci 2021; 14: 372–379.

Loy

Cheung

Fortier

, et al. Age-related nomograms for antral follicle count and anti-Mullerian hormone for subfertile Chinese women in Singapore. PLOS ONE 2017; 12: e0189830.

Aslan

Kasapoglu

Kosan

, et al. Age-stratified anti-Müllerian hormone (AMH) nomogram: a comprehensive cohort study including 22.920 women. Front Endocrinol (Lausanne) 2025; 16: 1612194.

Uncu

Aslan

Cakir

, et al. Do we overlook predictive factors in POSEIDON 1 patients? J Ovarian Res 2024; 17: 1.

10.

Naasan

Harrity

Pentony

, et al. Anti-Mullerian hormone normogram in an Irish subfertile population. Ir J Med Sci 2015; 184: 213–218.

11.

Mawusi

Yakass

Abaidoo

, et al. Use of anti-Mullerian hormone (AMH) for testing of ovarian reserve: a survey of fifteen (15) fertility centres in Ghana. Advances in Reproductive Sciences 2021; 09: 81–96.

12.

Razafintsalama-Bourdet

Bah

Amand

, et al. Random antral follicle count performed on any day of the menstrual cycle has the same predictive value as AMH for good ovarian response in IVF cycles. J Gynecol Obstet Hum Reprod 2022; 51: 102233.

13.

Liu

Pan

, et al. Comparison of anti-Müllerian hormone and antral follicle count in the prediction of ovarian response: a systematic review and meta-analysis. J Ovarian Res 2023; 16: 117.

14.

La Marca

Sunkara

SK.

Individualization of controlled ovarian stimulation in IVF using ovarian reserve markers: from theory to practice. Hum Reprod Update 2014; 20: 124–140.

15.

Sun

Xiong

Liu

, et al. Comparison of the predictive capability of antral follicle count vs. the anti-Müllerian hormone for ovarian response in infertile women. Front Endocrinol (Lausanne) 2022; 13: 862733.

16.

American Society of Reproductive Medicine, Practice Committee. Fertility evaluation of infertile women: a committee opinion. Fertil Steril 2021; 116: 1255–1265.

17.

Donkor

Sandall

Coping strategies of women seeking infertility treatment in Southern Ghana. Afr J Reprod Health 2009; 13: 81–93.

18.

Volk-Jesussek

Cohen’s Kappa: measuring inter-rater agreement, https://numiqo.com/tutorial/cohens-kappa (2026, accessed 28 April 2026).

19.

Von Elm

Altman

Egger

; STROBE Initiativeet al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147: 573–577.

20.

Damalie

FJMK

Senaya

Damalie

, et al. Barriers to assisted reproductive technology (ART) services in Ghana: a countrywide cross-sectional quantitative survey of fertility health workers and women with infertility. BMC Health Serv Res 2025; 25: 1228.

21.

Osaah

Damalie

FJMK

Amoah

VMK

, et al. Fertility knowledge, childbearing intentions and attitudes towards social egg freezing: a study of female students of a private university in Kumasi, Ghana. BMC Womens Health 2024; 24: 553.

22.

Abiodun

Alausa

Olasehinde

Ignorance could hurt: an assessment of fertility awareness, childbirth intentions and parenting attitudes among university students. Int J Adolesc Ment Health 2016; 30: 2.

23.

Knowing and not knowing about fertility: childless women and age-related fertility decline. Anthropol Aging 2021; 42: 30–48.

24.

Ben-Nagi

Kasaven

Jones

, et al. Oocyte yield in social, medical and donor oocyte cryopreservation cycles. Hum Fertil (Camb) 2022; 25: 508–515.

25.

Borgstrøm

Nygaard

Danielsen

, et al. Exploring motivations, attitudes and experiences of oocytedonors: a qualitative study. Acta Obstet Gynecol Scand 2019; 98: 1055–1062.

26.

Acharya

Bryant

Twiddy

Altruism or obligation? The motivations and experience of women who donate oocytes to known recipients in assisted conception treatment: an interpretative phenomenological analysis study. J Psychosom Obstet Gynecol 2017; 38: 4–11.

27.

Okafor

Ikechebelu

Joe-Ikechebelu

, et al. ‘ Gift with a price tag’: Nigerian egg donors’ knowledge, experiences and motivations. Afr J Reprod Health 2022; 26: 64.

28.

Zamora-Martínez

Pedelini

Llanos

, , et al. A systematic review on the demographics, motivations, and experiences of single women seeking fertility treatment. Reprod Biomed Online 2025; 50: 102345.

29.

Practice Committee of the American Society for Reproductive Medicine. Practice Committee of the American Society for Reproductive Medicine. Obesity and reproduction: a committee opinion. Fertil Steril 2021; 116: 1266–1285.

30.

Jain

Shukla

Jain

, et al. A correlation between antral follicle count and anti-Müllerian hormone in healthy Indian women of reproductive age. J South Asian Fed Obs Gynaecol 2022; 14: 1–5.

31.

Jain

Shukla

Jain

, et al. Age-specific nomogram for antral follicle count in fertile and infertile Indian women: a comparative study. Indian J Radiol Imaging 2023; 33: 463–470.

32.

Kelsey

Wright

Nelson

, et al. A validated model of serum anti-Müllerian hormone from conception to menopause. PLOS One 2011; 6: e22024.

33.

Iyer

Kallathikumar

Sinkar

, et al. Anti-Mullerian hormone (AMH) and age–an Indian laboratory retrospective analysis. Asian J Health Sci 2019; 5: 7.

34.

Arvis

Rongières

Pirrello

, et al. Reliability of AMH and AFC measurements and their correlation: a large multicenter study. J Assist Reprod Genet 2022; 39: 1045–1053.

35.

Joo

Lee

HJ.

Which is more reliable ovarian reserve marker between AMH and AFC in patients undergoing IVF?

Maturitas 2023; 173: 104.

36.

Anuradha

Partha Majumder

Shiffin

. Correlation of anti Mullerian hormone and antral follicular count in ovarian reserve testing. Obstet Gynecol Res 2022; 5: 170–174.