Cytomegalovirus Seroprevalence among Parturient Women at Panzi General Referral Hospital, Bukavu, Democratic Republic of the Congo ()
1. Introduction
Cytomegalovirus (CMV), also known as human herpesvirus 5 (HHV-5), is a double-stranded DNA virus of the family Herpesviridae [1]. Like the other members of this family, CMV establishes lifelong latent infection after primary infection, with possible episodes of reactivation, particularly in immunocompromised individuals [2]. In pregnant women, CMV infection is of particular concern because of the risk of mother-to-child transmission. Congenital CMV is the most common infectious cause of congenital infection worldwide and a major public health problem, notably because of its association with sensorineural hearing loss and long-term neurodevelopmental impairment [3].
CMV seroprevalence is closely shaped by socioeconomic factors and is generally higher in low-income, densely populated settings [4]. Worldwide, seroprevalence among women of reproductive age is estimated at about 86%, with wide regional variation, from under 70% in parts of Europe to over 90% in resource-limited countries [5]. These differences reflect variations in early exposure to the virus, hygiene, and living conditions.
In Europe, the highest CMV seroprevalence among women of reproductive age has been estimated in Türkiye, at an average of 96%, and the lowest in Ireland, at an average of 39% [4]; in northern Italy, it reaches 68.3% [6]. In the United Kingdom, seroprevalence is 49% [7]. In the United States, the overall age-adjusted CMV seroprevalence among persons aged 6 - 49 years was 50.4% in 1999-2004, and seropositivity was independently associated with low household income, high household crowding, and low household education [8].
In Asia, the overall CMV seroprevalence among pregnant women in Jiangsu Province, China, was 98.7% [9]. In southern India, prevalence was 75% [10], whereas in Japan it was 69.1% [11].
In the Middle East, high rates have also been reported in several Arab countries, notably Yemen [12] and Saudi Arabia, where prevalence ranges from 95.7% to 100% [13]. In eastern Iran, by contrast, it is reported to be around 72.1% [14].
In Africa, seroprevalence remains high and varies from one country to another [15]: 80.9% in Tunisia [16]; 100.0% among pregnant women in Egypt [17]; 72.2% in Sudan [18]; 73.9% in Mwanza, Tanzania [19]; 77.3% in Thika, Kenya [20]; and 86.4% in Johannesburg, South Africa [21].
In high-income countries, seroprevalence is generally lower and more heterogeneous, whereas in sub-Saharan Africa it remains high, approaching universal levels in some populations. In this context, CMV is recognized as a neglected yet important pathogen in Africa, where its clinical impact is probably underestimated because of limited access to diagnosis and surveillance [3].
HIV/CMV co-infection is also a particularly relevant issue in sub-Saharan Africa, where both infections are endemic. CMV acts as a cofactor in HIV infection and has been associated with increased transmission, faster disease progression, and higher mortality among infected patients [3].
In pregnant women and exposed children, this interaction is all the more worrying because it may increase the risk of congenital infection and affect neonatal outcomes, especially in resource-limited settings [3].
Despite this heavy burden, CMV remains largely underdiagnosed and under-monitored in resource-limited countries. Most African countries have no routine CMV screening program during pregnancy or at birth, which leads to an underestimation of the true burden of congenital CMV and its long-term complications [3].
Against this background, the lack of robust local data on CMV seroprevalence among parturient women in the Democratic Republic of the Congo is a significant gap, both for understanding the epidemiology of the infection and for designing appropriate prevention strategies. The aim of this study was therefore to estimate the seroprevalence of anti-CMV antibodies (IgM and IgG) and to identify associated factors among parturient women seen at Panzi General Referral Hospital in Bukavu.
2. Methods
The study was conducted at Panzi General Referral Hospital in Bukavu, a city in the eastern Democratic Republic of the Congo.
This was a descriptive cross-sectional study using convenience (consecutive) sampling, including 551 parturient women. The study ran for 12 months, from 1 October 2024 to 30 September 2025. All parturient women admitted to the maternity ward who gave their consent were included. Of the 723 parturient women admitted during the study period, 58 declined consent and 665 were enrolled; 114 were then excluded ([n] because of an equivocal IgG or IgM result and [n] because of insufficient data), leaving 551 women in the final analytical sample.
The dependent variable was IgM and IgG seropositivity or seronegativity. Explanatory variables were sociodemographic and clinical characteristics: age, place of residence, educational level, marital status, employment status, parity, and gestational age. Parity was categorized by the number of deliveries, including the current delivery, as primiparous, pauciparous, multiparous, or grand multiparous.
Samples consisted of 5 mL of venous blood collected in an anticoagulant (EDTA) tube during labor and sent to the laboratory, where plasma was separated by centrifugation at 3000 ×g for 15 min within 1 h of collection and stored below −20˚C until ELISA testing. Anti-CMV IgG and IgM antibodies were detected using the Cytomegalovirus IgG ELISA (catalogue no. E-CVG-K01) and Cytomegalovirus IgM µ-capture ELISA (catalogue no. E-CVM-K02) kits from RD-RatioDiagnostics (Frankfurt am Main, Germany), according to the manufacturer’s instructions. According to their package inserts, both kits are intended for testing human serum or plasma; however, the specimen-handling instructions specify serum, no anticoagulant is named, and performance data are reported for serum samples only. EDTA plasma was not validated against paired serum samples in our laboratory, which is a limitation of this study. Each run included a substrate blank and the kit calibrators (IgG: calibrators 0 - 4, with calibrator 1 defining the cut-off) or controls (IgM: negative, cut-off, and positive controls). Samples, calibrators, and controls were tested in duplicate. After blank subtraction, a run was accepted only if the optical density of the cut-off calibrator or control was at least 0.2 and that of IgG calibrator 4 or the IgM positive control was at least 3 or 1.5 times, respectively, that of the cut-off.
CMV serology was interpreted using the ratio of the mean optical density of the sample to that of the calibrator, for both IgG and IgM. A ratio above 1.1 was considered positive and a ratio below 0.9 negative, as specified in both package inserts. Results falling between these two values were considered equivocal and excluded from the study without retesting.
Based on the combination of IgG and IgM results, participants were classified into four descriptive antibody profiles: IgG-negative/IgM-negative (no detectable anti-CMV antibodies); IgG-positive/IgM-negative (IgG only, consistent with past infection); IgG-negative/IgM-positive (isolated IgM); and IgG-positive/IgM-positive (dual positivity). These profiles were not used to assign infection stage: isolated IgM may reflect early primary infection, persistent IgM, or nonspecific or false-positive reactivity, and dual positivity may reflect recent primary infection, reactivation, or reinfection, which cannot be distinguished without IgG avidity testing or molecular methods.
Statistical analysis was performed using IBM SPSS Statistics version 26. Descriptive statistics were expressed as proportions. Associations between sociodemographic variables and IgM and IgG seropositivity were first assessed using Pearson’s chi-square test; multivariable logistic regression was then performed to identify factors independently associated with IgM seropositivity. All explanatory variables were entered into the model, with age entered as the four age groups (≤20, 21 - 30, 31 - 40, and >40 years) coded 1 - 4 and treated as an ordinal (linear) term. Because all four widowed women were IgM-positive, maximum-likelihood estimation was affected by quasi-complete separation; the IgM model was therefore fitted using Firth’s penalized-likelihood logistic regression [22] with profile penalized-likelihood confidence intervals. Statistical significance was set at p < 0.05.
All approvals required for this study were obtained from the Provincial Ethics Committee under reference number CNES 001/DPSK/223PP/2024.
3. Results
3.1. Prevalence
In this study, IgG serology was positive in 350 parturient women and negative in 201, giving a seroprevalence of 63.5%. IgM serology, on the other hand, was positive in 203 women and negative in 348, giving a seroprevalence of 36.8% (versus 63.2% seronegative) (Table 1).
Table 1. Distribution of CMV serology results.
CMV serology |
N = 551 |
% |
IgG |
|
|
Positive |
350 |
63.5 |
Negative |
201 |
36.5 |
IgM |
|
|
Positive |
203 |
36.8 |
Negative |
348 |
63.2 |
Overall, 63.5% of parturient women were CMV IgG-positive. IgG seroprevalence was highest among women aged 21 - 30 years (65.1%, p = 0.145), and IgM seroprevalence among those aged 20 years or younger (42.3%, p = 0.465) (Table 2).
Table 2. CMV seroprevalence (IgM and IgG) by sociodemographic characteristics.
Variables |
IgM |
IgG |
Age |
Positive n/N |
Seroprevalence (%) |
p |
Positive n/N |
Seroprevalence (%) |
p |
≤20 |
22/52 |
42.3 |
0.465 |
31/52 |
59.6 |
0.145 |
21 - 30 |
101/298 |
33.9 |
194/298 |
65.1 |
31 - 40 |
75/189 |
39.7 |
121/189 |
64.0 |
>40 |
5/12 |
41.7 |
4/12 |
33.3 |
Residence |
|
|
|
|
|
|
Urban |
186/500 |
37.2 |
0.586 |
314/500 |
62.8 |
0.271 |
Rural |
17/51 |
33.3 |
36/51 |
70.6 |
Educational level |
|
|
|
|
|
|
None |
9/14 |
64.3 |
0.001 |
12/14 |
85.7 |
0.001 |
Primary |
43/153 |
28.1 |
80/153 |
52.3 |
Secondary |
103/289 |
35.6 |
188/289 |
65.1 |
University |
48/95 |
50.5 |
70/95 |
73.7 |
Marital status |
|
|
|
|
|
|
Single |
18/35 |
51.4 |
0.005 |
24/35 |
68.6 |
0.250 |
Married |
181/512 |
35.4 |
322/512 |
62.9 |
Widowed |
4/4 |
100 |
4/4 |
100 |
Employed |
|
|
|
|
|
|
No |
196/524 |
37.4 |
0.228 |
326/524 |
62.2 |
0.005 |
Yes |
7/27 |
25.9 |
24/27 |
88.9 |
Parity |
|
|
|
|
|
|
Primiparous |
57/138 |
41.3 |
0.015 |
97/138 |
70.3 |
0.018 |
Pauciparous |
31/83 |
37.3 |
53/83 |
63.9 |
Multiparous |
52/186 |
28.0 |
102/186 |
54.8 |
Grand multiparous |
63/144 |
43.8 |
98/144 |
68.1 |
Gestational age (weeks) |
|
|
|
|
|
|
28 - 36 |
31/87 |
35.6 |
0.795 |
63/87 |
72.4 |
0.059 |
37 - 40 |
159/424 |
37.5 |
258/424 |
60.8 |
>40 |
13/40 |
32.5 |
29/40 |
72.5 |
3.2. Sociodemographic Characteristics
As shown in Table 2, educational level and employment status were significantly associated with IgG positivity (p = 0.001 and p = 0.005, respectively). Parity was also significantly associated with both IgM seropositivity (p = 0.015) and IgG seropositivity (p = 0.018). Educational level (p = 0.001) and marital status (p = 0.005) were likewise associated with IgM positivity. In contrast, age, place of residence, and gestational age were associated with neither IgM nor IgG.
Table 3. Multivariable logistic regression: association between sociodemographic factors and maternal IgM serology.
Variables |
B |
df |
p |
Exp(B) |
95% CI for Exp(B) |
Lower |
Upper |
Constant |
22.266 |
1 |
0.999 |
4676064139.0 |
|
|
Gestational age (weeks) |
|
2 |
0.756 |
|
|
|
28 - 36 |
−0.018 |
1 |
0.966 |
0.982 |
0.426 |
2.263 |
37 - 40 |
0.156 |
1 |
0.666 |
1.169 |
0.574 |
2.381 |
Employed |
−0.445 |
1 |
0.337 |
0.641 |
0.258 |
1.589 |
Educational level |
|
3 |
0.000 |
|
|
|
Primary |
−1.549 |
1 |
0.013 |
0.212 |
0.062 |
0.726 |
Secondary |
−1.064 |
1 |
0.090 |
0.345 |
0.101 |
1.182 |
University |
−0.345 |
1 |
0.600 |
0.708 |
0.195 |
2.573 |
Parity |
|
3 |
0.018 |
|
|
|
Primiparous |
−0.481 |
1 |
0.182 |
0.618 |
0.305 |
1.254 |
Pauciparous |
−0.528 |
1 |
0.133 |
0.590 |
0.297 |
1.174 |
Multiparous |
−0.871 |
1 |
0.002 |
0.419 |
0.240 |
0.729 |
Marital status |
|
2 |
0.099 |
|
|
|
Single |
−20.445 |
1 |
0.999 |
0.000 |
0.000 |
. |
Married |
−21.464 |
1 |
0.999 |
0.000 |
0.000 |
. |
Residence |
−0.327 |
1 |
0.350 |
0.721 |
0.363 |
1.431 |
Age |
0.008 |
1 |
0.969 |
1.008 |
0.668 |
1.521 |
Table 3 shows that, after adjustment, educational level (p < 0.001) and parity (p = 0.018) remained significantly associated with IgM positivity: women with primary education (OR = 0.212; 95% CI: 0.062 - 0.726) and multiparous women (OR = 0.419; 95% CI: 0.240 - 0.729) had a lower risk than women with no formal education and grand multiparous women, respectively.
4. Discussion
This study was conducted to determine the seroprevalence of CMV infection among parturient women at the maternity unit of Panzi General Referral Hospital in Bukavu, DR Congo. IgG seroprevalence was 63.5% and IgM seroprevalence 36.8%; this high IgM rate, which would point to recent infection or reactivation, should be interpreted with caution in the absence of avidity testing. To our knowledge, this is the first study of its kind in the region. Across the African continent, studies have reported high rates in East Africa (72% - 89%) and even higher rates in North, Southern, and West Africa (100%) [23]. In Kenya, for example, CMV seroprevalence was 77.3% [20], compared with 72.2% in western Sudan [18].
CMV seroprevalence in our population is lower than that reported in most developing countries, both African and non-African [4] [9] [10], and closer to that observed in some developed countries [5] [7] [8] [24]. Low educational and socioeconomic status, together with the higher prevalence of poor hygiene conditions in developing countries, may help explain the higher CMV prevalence in these settings [24].
IgM seroprevalence was highest among women aged 20 years or younger (42.3%), with no statistically significant difference by age (p = 0.465). This is consistent with the observation by Radoi et al. in south-western Romania, where active infection was more frequent among the youngest women [25], whereas Vilibić-Čavlek et al. in Croatia identified age as a significant determinant of both IgM and IgG positivity, with IgG seroprevalence rising steadily with age [26]. The predominance of infection among younger women—a period of intense reproductive activity and close contact with young children—is consistent with the main routes of CMV transmission (saliva, urine, and cervicovaginal secretions).
Educational level was significantly associated with both IgM and IgG positivity (p = 0.001 for both), and employment status with IgG positivity (p = 0.005). Several authors have reported similar links between socioeconomic or educational status and CMV serology; Zhang et al., in Jiangsu Province, identified low maternal education as a risk factor for adverse pregnancy outcomes in the context of CMV infection [9]. In our series, women with no formal education had the highest IgG seroprevalence (85.7%), which may reflect socioeconomic, hygiene, and crowding conditions that favor early and repeated exposure to the virus, although the relationship was not linear (52.3% among women with primary education versus 73.7% among those with university education). Employed women were more often IgG-seropositive (88.9% vs. 62.2%), possibly reflecting greater occupational exposure, particularly contact with young children.
Parity was also significantly associated with IgM (p = 0.015) and IgG (p = 0.018) seropositivity. Multivariable logistic regression confirmed a statistically significant association between parity and IgM positivity (p = 0.018), with only multiparous women having a significantly lower risk than the reference category of grand multiparous women (OR = 0.419; 95% CI: 0.240 - 0.729; p = 0.002). One possible explanation is repeated contact with young children—the main reservoirs and shedders of CMV—over successive pregnancies [27]; however, contact with children was not measured in this study, and IgM positivity did not vary monotonically with parity. Other studies, however, such as that by Mhandire et al. in Harare, found no significant association between CMV seropositivity and parity [28], which underscores how much this factor varies across study populations.
In contrast, no significant association was found between place of residence (urban or rural) or gestational age and CMV serology. Marital status was associated only with IgM (p = 0.005), an association that did not persist after adjustment (p = 0.099) and rested on very small numbers (4 widowed women). This contrasts with the findings of Vilibić-Čavlek et al. in Croatia, where region and place of residence were significant determinants of seropositivity [26], and of Radoi et al. in Romania, who observed variation in seroprevalence by area of residence [25]. The homogeneity of our population, which was overwhelmingly urban (500 of 551 parturient women), may explain why no difference was seen by place of residence.
5. Strengths and Limitations
This study has several strengths. As the first study of CMV seroprevalence among parturient women in DR Congo, it provides new data. The large sample (551 parturient women) and consecutive recruitment over twelve months strengthen the robustness of the estimates, while dual IgM/IgG serology, multivariable logistic regression analysis, and prior ethical approval add to its methodological rigor.
The main limitations of this study are its single-center design and the use of convenience sampling, which restrict the representativeness and generalizability of the findings. The cross-sectional design does not allow causal inference, and diagnosis relied solely on ELISA serology (IgM/IgG), without avidity testing or PCR; primary infection therefore cannot be reliably distinguished from reactivation or a false-positive result. In addition, antibodies were measured in EDTA plasma, whereas the specimen-handling instructions of both kits specify serum; because EDTA plasma was not validated against paired serum samples in our laboratory, some misclassification of antibody status cannot be excluded. Finally, the small size of some subgroups (widowed women) reduces statistical power.
Larger multicenter prospective studies combining IgG avidity testing, PCR, and follow-up of newborns would help refine estimates of seroprevalence and its risk factors, and clarify the risk of congenital CMV infection.
6. Conclusion
This study shows a high CMV seroprevalence among parturient women at Panzi General Referral Hospital, with IgG positivity of 63.5% and IgM positivity of 36.8%. Because IgM positivity alone does not establish recent infection, the IgM estimate should be confirmed with IgG avidity testing or PCR. In multivariable analysis, educational level and parity remained independently associated with IgM seropositivity, and employment status was associated with IgG seropositivity in bivariate analysis. Together, these results suggest that sociodemographic factors are associated with CMV serostatus in this population; whether contact with young children, which was not measured in this study, contributes to these associations remains to be determined. They support stronger awareness and prevention of CMV infection in pregnant women, as well as further studies to assess the risk of congenital infection. Routine antenatal CMV screening may merit evaluation in this setting, but its introduction would first require evidence on confirmatory testing pathways, such as IgG avidity testing and PCR, and on clinical benefit.
Author Contributions
All authors made a significant contribution to the work reported, whether in the conception, study design, and execution or in all these areas, took part in drafting, revising, or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
Acknowledgements
We express our gratitude to all the patients who participated in this study.
Funding
The study was not funded.
Data Availability Statement
The datasets utilized in this study can be obtained from the corresponding author upon request. Please contact Dr. Julien Bwama via email at [email protected].