Understanding the linkages between male circumcision and multiple sexual partnership among married Ghanaian men: Analysis of data from the 2014 Ghana demographic and health survey.

INTRODUCTION: This current study sought to investigate the association between male circumcision status and engaging in multiple sexual partnership among men in Ghana.
METHODS: Data from this study come from the men's file of the 2014 Ghana demographic and health survey. Both descriptive and inferential statistics were conducted among 1, 948 men and the level of statistical significance was pegged at p < 0.05.
RESULTS: Results revealed that men who have been circumcised were more likely to engage in multiple sexual partnership (AOR = 3.36; CI:1.14-9.89), compared to those who have not been circumcised. With the covariates, men with primary level of education were more likely to have multiple sexual partners (AOR = 2.01; CI:1.10-3.69), compared to those with no education. With wealth status, men with richest (AOR = 2.27; CI:1.04-4.97), richer (AOR = 2.05; CI: 1.03-4.08), and middle wealth status (AOR = 1.83; CI:1.01-3.34) had the highest likelihood of having multiple sexual partners, compared to those with poorest wealth status. Conversely, men who professed the Islamic faith were less likely to engage in multiple sexual partnership (AOR = 0.58; CI: 0.36-0.94), compared to Christians. Similarly, men who resided in the Brong Ahafo (AOR = 0.51; CI: 0.26-0.99), Upper East (AOR = 0.41; CI:0.19-0.89), and Ashanti regions (AOR = 0.39; CI: 0.20-0.78) were less likely to engage in multiple sexual partnership.
CONCLUSION: Based on the current findings, educational campaigns by stakeholder groups (e.g., Ministry of Health in collaboration with the National Commission on Civic Education, civil society, educational institutions) should sensitize the sexually active population at the community level to consistently use condoms, especially when they have multiple sexual partners, even when a man is circumcised. Campaign messages must clearly emphasize that male circumcision should not substitute precautionary measures such as delay in the onset of sexual relationships, averting penetrative sex, reducing the number of sexual partners as well as correct and consistent use of male or female condoms regardless one's social standing.

Introduction

Circumcision is one of the ancient surgical procedures concerning the removal of some or all of the penile prepuce (foreskin) performed till date and has remained a highly critical subject (Bossio et al., 2014; Hutcheson, 2004). Today, male circumcision has become a common medical procedure with improved anaesthetic, surgical, and antiseptic techniques (Hutcheson, 2004). Approximately, the global prevalence of male circumcision indicates that one-third of adult men worldwide are circumcised for medical, religious, social, and/ or cultural reasons (UNAIDS, 2007, World Health Organization, 2008). Available evidence reveals that projected 665 million men (30% of all men) over 15 years of age worldwide are circumcised. Specifically, nearly two-thirds (69%) are Muslim (found mostly in Asia, the Middle East, and North Africa), 0.8% are Jewish, and 13% are men in the United States who are not Muslim or Jewish (World Health Organization, 2008). These purported figures have recently been challenged due to limitations (e.g., self-reported, among religious groups) from previous studies (see Morris et al., 2016 for details). Morris et al. (2016) noted that current estimates of male circumcision range between 36.7 and 38.7%, even though the actual percentage could be lower or higher than reported values.

Within Africa, male circumcision is common in many countries and is almost universal in North Africa and most parts of West Africa. Approximately 90% males are circumcised in West Africa and 95% in Ghana (Weiss et al., 2010). Comparatively, it is fewer in southern Africa, where reported prevalence is lowest, about 15% in several countries, including Namibia, Swaziland, Zambia, and Zimbabwe (Connolly et al., 2004; Drain et al., 2006; Langeni et al., 2005; Peltzer et al., 2007; Williams et al., 2006; World Health Organization, 2008), though higher in others such as Malawi (21%), Botswana (25%), South Africa (35%), Lesotho (48%), Mozambique (60%), and Angola as well as Madagascar (> 80%) (Drain et al., 2006; Connolly et al., 2004; Peltzer et al., 2007; Williams et al., 2006; World Health Organization, 2008). Prevalence in Central and East Africa varies from approximately 15% in Burundi and Rwanda to 70% in Tanzania, 84% in Kenya, and 93% in Ethiopia and Somalia (Demographic and health surveys, 2006; Peltzer et al., 2007; Williams et al., 2006; World Health Organization, 2008).

Contemporary literature about circumcision is partially skewed (Bossio et al., 2014). According to Bossio et al. (2014), quite significant amount of research has been directed towards the medical benefits of circumcision. Emerging findings suggest that circumcision status has a significant influence on men's sexual partners. For instance, a reduction in HIV and other STD transmission rates among circumcised men on their partners have been reported in some studies (Auvert et al., 2005; Bailey et al., 2005; Gray et al., 2007). Female sexual partners of circumcised men seem to gain protective health benefits over female partners of non-circumcised men against STIs such as bacterial vaginosis (CDCP, 2010; Cherpes et al., 2008; Gray et al., 2009), chlamydia (Castellsagué et al., 2005; Sobngwi-Tambekou et al., 2009), herpes (Cherpes et al., 2003), and HPV (Wawer et al., 2011). Men who have sex with men (MSM) also seem to realize some protecting effects against HIV from having sexual intercourse with circumcised rather than non-circumcised partners (Boyle and Hill, 2012), although the reported impacts appear to be mediated by the nature of sexual intercourse they engaged in (i.e., receptive and insertive sex are related to diverse risks) (Koblin et al., 2006; Sidler et al., 2008; UN Joint Programme on HIV/AIDS, 2009).

Within sub-Saharan Africa, epidemiological evidence has shown a geographical connection between regions of higher prevalence of HIV and lower rates of male circumcision (Drain et al., 2006). Some randomized controlled studies on male circumcision have revealed a reduction of 60%–75% in the risk of female to male transmission of HIV-1 in circumcised men in South Africa (Auvert et al., 2005). Other studies conducted in Kenya and Uganda established a risk reduction of around 58% and 53% respectively (Bailey et al., 2007; Gray et al., 2007). Consequently, some reports (e.g., WHO/UNAIDS, 2007) and research (e.g., Caldwell & Caldwell, 1994; Kebaabetswe et al., 2003; Scott et al., 2005)endorsed male circumcision as a potential strategy to reduce the risk for HIV acquisition, particularly in high HIV prevalence and low male circumcision countries.

Available scholarly information indicates that traditional male role orientations (e.g., masculinity and power distance) and male sexual behaviour may heighten the risk of HIV and other STDs infection among sexually active men in Ghana (e.g., Adomako et al., 2009; Anarfi, 2006; Tsikata, 2007). Specifically, cultural orientation allows men to be sexually domineering (i.e., have multiple sexual partners that increase their risk taking) (Anarfi, 2006; World Health Organization, 2007). Given the widespread sociocultural acceptance of men having multiple sexual partners in most sub-Saharan African societies, studies like this would provide a more comprehensive understanding of multifaceted indicators surrounding the circumcision debate toward public reproductive health policy and implementation of practical interventions. The current study prospectively investigates the association between male circumcision status and the choice of multiple sexual partners in Ghana, where there is limited empirical data on the subject.

Materials and methods

Data

Data for the study was obtained from the 2014 Demographic and Health Survey (DHS) of Ghana. DHS is a nationwide survey conducted every five-year period across low- and middle-income countries in Asia and Africa. It focuses on maternal and child health by interviewing women in their reproductive age (15–49 years) as well as men aged 15–64. The survey follows the same standard procedures—sampling, questionnaires, data collection, cleaning, coding and analysis—that allow for cross-country comparison. It employed a stratified two-stage sampling technique to make it nationally representative (Aliaga & Ren, 2006). The first-stage sampling frame consisted of a list of primary sampling units (PSUs) or enumeration areas (EAs) that covered the entire country and usually were obtained from the latest national census—when available. Each EA was further subdivided into standard size segments of about 100–500 households per segment. At this stage, a sample of predetermined segments was selected randomly with a probability proportional to the EA's measure of size (number of households in EA). For the second stage, the DHS survey personnel selected households systematically from a list of previously enumerated households in each selected EA segment, and in-person interviews were conducted in selected households to target populations: women aged 15–49, men aged 15–64, and children under 5. For the purpose of this study, only married men (15–64) for whom there was complete information on male circumcision and multiple sexual partnership were included for subsequent analysis (N = 1,958).

Measurement of variables

The outcome variable used for this study was multiple sexual partnership. To derive this variable, married men were asked about the number of sex partners they have had in the last 12 months, excluding their spouses. The responses ranged from 0 to 7. Those who gave 0 as a response were considered as those without multiple sexual partners and were coded as 0 = No, while those who said they have had 1 or more sex partners in the last 12 months preceding the survey were considered as having multiple sexual partners and were coded as 1 = Yes.

The key explanatory variable considered in this study was self-report of male circumcision status, based on the following question: ‘‘Some men are circumcised, that is, the foreskin is completely removed from the penis. Are you circumcised?’’ The responses were Yes = 1, No = 2, 8 = Don't know. For the purpose of the analysis, only men who provided confirmatory responses (either “Yes” or “No”) were included in the study, and those who did not know the answer were excluded. Since this question alone does not differentiate traditional circumcision from medical circumcision conducted in a clinical facility, the survey in Ghana collected additional information. For example, the questionnaires asked the age the respondent was circumcised, who performed the circumcision, and where was it done (Ghana Statistical Service [GSS], Ghana Health Service [GHS], and ICF International, 2015). These additional detailed questions revealed valuable information about the phenomenon, since traditional circumcision may be ‘‘partial’’ compared to medical circumcision performed by trained clinicians (Glasier et al., 2006). Despite the difference between traditional and medical circumcision, we included all men who were circumcised irrespective of the type of circumcision. The reason being that, the sample size for men who had undergone medical circumcision was small in the dataset. Moreover, our interest was to find out if circumcision could lead to multiple sexual partnership, irrespective of the type. Previous studies on the association between male circumcision and risky sexual behaviour among men in sub-Saharan Africa did not differentiate between medical and traditional circumcision (e.g., Balekang & Dintwa, 2016; Friedrich, 2014; Mapoma & Bwalya, 2017).

Apart from the key explanatory variable, other variables were included in the analysis as covariates. These variables were age, employment status, level of education, place of residence, religious affiliation, wealth status, and region. They were extracted from the males' recode file of the 2014 Ghana DHS, and some of them were recoded. Residence was coded as urban = 1 and rural = 2; age was categorised into 15–24 = 1, 25–34 = 2, 35–44 = 3, 45–54 = 4, 55–64 = 5. Employment status was also categorised into two: unemployed = 1 and employed = 2. Level of education was classified into four categories: no formal education = 1, primary = 2, secondary = 3, and tertiary = 4. Wealth status was categorised as poorest = 1, poorer = 2, middle = 3, richer = 4, and richest = 5. Religious affiliation was recoded as Christian = 1, Islam = 2, 3 = Traditional religion, and No religion = 4. The choice of these explanatory variables was based on their association with multiple sexual partnership in previous studies (Dorina et al., 2014; Exavery et al., 2015; Maher et al., 2011).

Analytical procedure

Descriptive and inferential statistics were conducted. Descriptive analysis mainly focused on the presentation of frequencies and percentages for male circumcision and socio-demographic characteristics of the respondents. Again, the descriptive statistics looked at the distribution of multiple sexual partnership across male circumcision and socio-demographic characteristics of the respondents, using chi-square test of independence (see Table 1). The inferential statistics focused on a multivariable analysis of the influence of male circumcision on multiple sexual partnership while controlling for the effect of all the covariates that were significant at the chi-square level (see Table 2). The results were reported using adjusted odds ratios (AORs) and confidence intervals (CI). The statistical significance level was set at p < 0.05. All frequency distributions were weighted while the survey command (svy) in Stata was used to adjust for the complex sampling structure of the data in the regression analyses. Stata 14.1 (College Station, TX) statistical analysis tool was used for the analysis. This software has the advantage of directly including robust standard errors that account for the complex sample design (two-stage sample design). To test uncertainties of the multivariable logistic regression models, a further test was conducted using Poisson regression, with results showing no evidence of uncertainties since the significance of the predictors in the multivariable logistic regression models did not change (see Table 3).

Table 1

Multiple sexual partnership across men's circumcision status and their background characteristics (weighted).

Variables	Frequency	Percentage	Multiple sexual partnership
Male circumcision (χ2 = 14.07, p < 0.001)			Yes	No
No	101	5.1	2.5	97.6
Yes	1857	94.9	12.2	87.8
Age (χ2 = 9.52, p < 0.001)
15–24	29	1.5	18.9	81.1
25–34	483	24.7	13.5	86.5
35–44	725	37.0	12.1	87.9
45–54	538	27.5	9.6	90.4
55–64	184	9.4	7.3	92.7
Employment status (χ2 = 0.197, p > 0.05)
Unemployed	16	0.8	12.5	87.5
Employed	1942	99.2	11.4	88.6
Level of education (χ2 = 31.63, p < 0.001)
No education	370	18.9	4.7	95.3
Primary	236	12.1	12.8	87.2
Secondary	1086	55.5	13.9	86.1
Higher	266	13.6	14.7	85.3
Place of residence (χ2 = 13.08, p < 0.001)
Urban	957	48.9	14.3	85.7
Rural	1001	51.1	9.1	90.9
Religious affiliation (χ2 = 21.6820, p < 0.001)
Christianity	1323	67.6	14.0	86.0
Islam	413	21.1	6.3	93.7
Traditional religion	118	6.0	9.1	90.9
No religion	104	5.3	9.4	90.7
Wealth status (χ2 = 43.06, p < 0.001)
Poorest	395	20.2	5.2	94.9
Poorer	356	18.2	11.1	88.9
Middle	322	16.4	12.5	87.5
Richer	369	18.8	14.2	85.8
Richest	516	26.4	18.5	81.5
Region (χ2 = 82.40, p < 0.001)
Western	198	10.1	16.6	83.4
Central	201	10.3	11.7	88.3
Greater Accra	383	19.6	26.0	74.1
Volta	158	8.1	19.1	80.9
Eastern	175	8.9	13.4	86.6
Ashanti	335	17.1	7.4	92.6
Brong Ahafo	157	8.0	7.3	92.7
Northern	209	10.7	6.6	93.4
Upper East	90	4.6	5.2	94.8
Upper West	52	2.7	4.9	95.1

Source: Computed from 2014 GDHS.

Multivariable logistic regression results of the association between male circumcision, socio-demographic characteristics and multiple sexual partnership.

Table 2

Multivariable logistic regression results on the association between male circumcision and multiple sexual partnership among married men in Ghana.

Variables	COR	AOR 95% CI
Male circumcision
No	Ref	Ref
Yes	5.52***[2.03–15.05]	3.36*[1.14–9.89]
Age
15–24		Ref
25–34		0.69[0.261.83]
35–44		0.57[0.22–1.48]
45–54		0.43[0.16–1.13]
55–64		0.38[0.13–1.12]
Level of education
No education		Ref
Primary		2.01*[1.10–3.69]
Secondary		1.53[0.85–2.74]
Higher		1.39[0.67–2.90]
Place of residence
Urban		Ref
Rural		0.95[0.63–1.46]
Religion
Christianity		Ref
Islam		0.58*[0.36–0.94]
Traditionalist		1.59[0.83–3.05]
No religion		0.95[0.48–1.87]
Wealth status
Poorest		Ref
Poorer		1.75[0.98–3.12]
Middle		1.83*[1.01–3.34]
Richer		2.05*[1.03–4.08]
Richest		2.27*[1.04–4.97]
Region
Western		Ref
Central		0.65[0.35–1.20]
Greater Accra		1.52[0.89–2.60]
Volta		1.15[0.64–2.06]
Eastern		0.81[0.45–1.46]
Ashanti		0.39**[0.20–0.78]
Brong Ahafo		0.51*[0.26–0.99]
Northern		0.82[0.38–1.77]
Upper East		0.41*[0.19–0.89]
Upper West		0.52[0.22–1.22]
N		1,958
Psuedo R2		0.086

Exponentiated coefficients; 95% confidence intervals in brackets.

*p < 0.05, **p < 0.01, ***p < 0.001.

COR= Crude Odds Ratio AOR = Adjusted Odds Ratio CI=Confidence Interval; Ref = reference category.

Table 3

Poisson regression results on the association between male circumcision and multiple sexual partnership among married men in Ghana.

Variables	IRR 95% CI
Male circumcision
No	Ref
Yes	3.41*[1.20–9.74]
Age
15–24	Ref
25–34	0.82[0.39–1.72]
35–44	0.61[0.29–1.28]
45–54	0.50[0.24–1.06]
55–64	0.60[0.26–1.39]
Level of education
No education	Ref
Primary	2.25*[1.33–3.82]
Secondary	1.67*[1.01–2.75]
Higher	1.27[0.68–2.32]
Place of residence
Urban	Ref
Rural	1.00[0.70–1.44]
Religion
Christianity	Ref
Islam	0.59*[0.39–0.90]
Traditionalist	1.44[0.81–2.58]
No religion	1.18[0.68–2.06]
Wealth status
Poorest	Ref
Poorer	1.74[1.03–2.93]
Middle	2.19**[1.26–3.80]
Richer	2.34**[1.27–4.31]
Richest	2.60**[1.31–5.15]
Region
Western	Ref
Central	0.72[0.44–1.19]
Greater Accra	1.43[0.93–2.19]
Volta	1.06[0.65–1.75]
Eastern	0.76[0.46–1.25]
Ashanti	0.42**[0.24–0.76]
Brong Ahafo	0.51*[0.29–0.93]
Northern	0.94[0.50–1.77]
Upper East	0.44*[0.22–0.90]
Upper West	0.61[0.28–1.31]
N	1,958
Psuedo R2	0.089

Exponentiated coefficients; 95% confidence intervals in brackets.

*p < 0.05, **p < 0.01, ***p < 0.001.

IRR = Incident Rate Ratios; CI=Confidence Interval; Ref = reference category.

Ethics approval

The men gave oral and written consent. Ethical approval was given by individual national institutional review boards and by ICF International institutional review board. Permission to use the data set was requested from MEASURE DHS. The data set is available to the public at www.measuredhs.org.

Results

Descriptive statistics on multiple sexual partnership across Men's status circumcision and their background characteristics

Table 1 shows results on multiple sexual partnership across men's circumcision status and their background characteristics. The results show that most men who are circumcised engage in multiple sexual partnership (12.2%), compared to those who are not circumcised. Multiple sexual partnership was high among men aged 15–24 (18.9%), those who are unemployed (12.5%), those with higher education (14.7%), those in urban areas (14.3%), Christians (14.0%), those with richest wealth status (18.5%), and men who live in the Greater Accra region (26.0%). The results of the chi-square test showed that all the independent variables, except employment status, had statistically significant associations with multiple sexual partnership among men in Ghana.

Table 3 shows results of the multivariable logistic regression analysis of the association between socio-demographic characteristics and multiple sexual partnership among married men in Ghana. The results indicate that men who have been circumcised are more likely to engage in multiple sexual partnership (AOR = 3.36; 95% CI: 1.14–9.89), compared to those who have not been circumcised. With the covariates, men with primary level of education were more likely to have multiple sexual partners (AOR = 2.01; 95% CI: 1.10–3.69), compared to those with no education. With wealth status, men with richest (AOR = 2.27; 95% CI: 1.04–4.97), richer (AOR = 2.05; 95% CI: 1.03–4.08), and middle wealth status (AOR = 1.83; 95% CI: 1.01–3.34) had the greatest likelihood of having multiple sexual partners, compared to those with poorest wealth status. Conversely, men who professed the Islamic faith were less likely to engage in multiple sexual partnership (AOR = 0.58; 95% CI: 0.36–0.94), compared to Christians. Similarly, men who resided in the Brong Ahafo (AOR = 0.51; 95% CI: 0.26–0.99), Upper East (AOR = 0.41; 95% CI: 0.19–0.89) and Ashanti regions (AOR = 0.39; 95% CI: 0.20–0.78) were less likely to engage in multiple sexual partnership.

Discussion

The purpose of the study was to establish the association between male circumcision and multiple sexual partnership among married men in Ghana. We found that men who have been circumcised are more likely to engage in multiple sexual partnership, compared to those who have not been circumcised, a finding that corroborates previous studies (e.g., Balekang & Dintwa, 2016; Kibira et al., 2015). For varied sexual experience from multiple partners, circumcised men could derive penile sensation caused by innervations from nerve endings around the glans penis (Halata & Munger, 1986). According to Cox et al. (2015), penile sexual sensation or sensitivity is higher in circumcised men because of greater access of genital corpuscles to sexual stimuli after removal of the prepuce. This sensation is centered on the distribution of these corpuscles located in the glans.

Further, in many indigenous societies like Ghana, typical socio-cultural norms (e.g., power distance and masculinity orientation) could indirectly influence circumcised men to sexually explore with multiple partners to prove their prowess (i.e., virility), a characteristic of one's manliness could also account for the findings noted. Other cultural and social factors such as the stigma associated with uncircumcision may contribute to the current finding. Women may not be willing to engage in sexual intercourse with uncircumcised men due to health issues associated with the condition; hence, this perception might increase their vulnerability towards circumcised men (Grund et al., 2019; Morris et al., 2019). For some religions (e.g., Christians/Muslims) and ethnic groups (e.g., Akan, Mole-Dagbon) in Ghana, male circumcision is compulsory for the male child and, therefore, women are more likely to see sexual intercourse with such men as “safe” (Adu-Gyamfi & Adjei, 2014). The proportionally reduced number of men who are uncircumcised among the Ghanaian populace may also contribute to this very finding. Future research could investigate risk compensation among circumcised men in ecological setting (e.g., using ecological momentary assessment) instead of the randomized control trials and/or other artificial contexts that currently dominate circumcision literature (Eaton & Kalichman, 2009). For example, further studies could target alterations in beliefs and perceptions among circumcised men on long-term sexual behaviour changes.

With the covariates, circumcised men with primary level of education were more likely to have multiple sexual partners, compared to those with no education. The finding is in line with the findings of previous studies, which established that men with some level of education are more likely to have multiple sexual partners (Dolcini et al., 1993, pp. 208–214; Ho-Foster et al., 2010). The authors explained that men with some level of education are more likely to have more wealth through employment and hence are able to afford additional sexual partners. Men with richest, richer, and middle wealth status had the highest likelihood of having multiple sexual partners, compared to those with poorest wealth status. Madise et al. (2007) confirm that poverty may prevent men from engaging in multiple sexual partners because they may not be able to take care of their partners' financial and health needs. Additionally, the fear of pregnancy might be a factor which could deter men from engaging in multiple sexual behaviours (Doku, 2012). Additionally, linking the ability of men to afford contraceptive usage in Ghana may increase their chances of engaging in multiple sexual behaviours. These propositions confirm why men with richest, richer, and middle wealth status are more likely to have multiple sexual partners in Ghana.

Men who professed the Islamic faith were less likely to engage in multiple sexual partnership, compared to Christians. Despite both religions holding strong practices regarding sexuality, Agyei-Asabere (2016) indicates that the Islamic religion upholds stronger views, compared to their counterparts. Therefore, Islamic men are likely to defray from engaging in unaccustomed multiple sexual relations for the fear of ostracism. Another possible explanation is that heteronormative Christian sexual norms may really encourage Christian men to have multiple sexual partners, compared to their Muslim counterparts (Pan et al., 2016). Similarly, the current trend regarding moral degradation (e.g., casual sexual behaviour) across many developing societies has eroded the somewhat strong moral arguments on doctrinal position of some Christian denominations on sexuality. Therefore, it is unsurprising to see many Christian men having multiple sexual partners. Therefore, religious doctrines on sexual relations may play a major role in regulating the extent to which members of such affiliations may be predisposed to risky sexual behaviours (Malamba et al., 1994; Slaymaker et al., 2004). We admit that some of these explanations on the role of religion in multiple sexual partnership from previous studies could be biased by differences in study settings as well as issues related to over and under-reporting (i.e., social desirability).

Regarding residential location, men who resided in the Brong Ahafo, Upper East, and Ashanti regions were less likely to engage in multiple sexual partnership. This finding partially confirms the observation by Addai (1999) that the Ashantis, Bonos, and Ahafo as well as the Dagombas of Brong Ahafo and Upper East respectively, are less likely to engage in culturally untoward sexual behaviours such as having multiple sexual partnerships. These ethnic groups are more likely to hold sexual mischief as taboos. Therefore, this finding is consistent with the documentation regarding the role culture plays in shaping sexual behaviours among Ghanaian men (Anarfi & Awusabo, 1993). Future studies could explore reasons using qualitative designs to better understand the diverse roles that culture and ethnicity may play in multiple sexual partnership in Ghana.

Strength and limitations

The strengths of the study lies in the relatively large sample size and the multistage sampling technique employed (i.e., two stage), which makes the data nationally representative, hence the possibility to generalise these findings to men in Ghana. Admittedly, this study has various limitations. These limitations include our inability to claim cause-effect relationship between the explanatory and outcome variables in the study due to the cross-sectional nature of the data analysed. Similarly, the use of secondary data restricts the scope of this current study to measured variables within the dataset. The Ghana Demographic and Health Survey did not collect enough information on circumcision to allow in-depth analysis. Similar to most demographic surveys, the lack of qualitative data limits this in-depth understanding and explanation of patterns found within the quantitative analysis of this study. We also admit that collecting data on sexual behaviour sometimes is difficult, because it is considered private and intimate and has a link with self-image and personality. As most of the variables were also self-reported, the possibility of social desirability bias is high. Additionally, research evidence suggests that, in some instances, men tend to either under or over-report their number of sexual partners (e.g., Clark et al., 2011; Smith, 1992).

Conclusion and practical implications

Summarily, we found that men who have been circumcised are more likely to engage in multiple sexual relationships, compared to those who have not been circumcised. Additionally, men with primary level of education and those who belong to the middle, richer, and richest wealth are more likely to have multiple sexual partners. Likewise, men who professed the Islamic faith and men in Brong Ahafo, Upper East, and Ashanti regions are less likely to engage in multiple sexual partnership. Findings suggest that circumcised men in Ghana may continue to be exposed to high risky sexual behaviours (i.e., having multiple sexual partners and perhaps, unprotected sex). Based on the current findings, educational campaigns by stakeholder groups (e.g., Ministry of Health, in collaboration with the National Commission on Civic Education, educational institutions) should sensitize the sexually active population at the community level to consistently use condoms, especially when having multiple sexual partners, even when a man is circumcised. Although there are benefits of male circumcision (e.g., prevention method), educational messages must clearly emphasize that male circumcision should not substitute precautionary measures such as delay in the onset of sexual relationships, averting penetrative sex, reducing the number of sexual partners as well as correct and consistent use of male or female condoms regardless of one's social standing. Women as sexual partners should also strongly negotiate for condom use with circumcised men with the clear understanding that they are not fully protected against sexually transmitted diseases, including HIV.

CRediT authorship contribution statement

Bright Opoku Ahinkorah: Conceptualization, Methodology, Formal analysis, Writing - original draft, Supervision, Methodology, Writing - review & editing. John Elvis Hagan: Conceptualization, Methodology, Formal analysis, Writing - original draft, Supervision, Methodology, Writing - review & editing. Abdul-Aziz Seidu: Conceptualization, Methodology, Formal analysis, Writing - original draft, Supervision, Methodology, Writing - review & editing. Eric Torgbenu: Conceptualization, Methodology, Formal analysis, Writing - original draft, Supervision, Methodology, Writing - review & editing. Eugene Budu: Data curation, Writing - original draft, Writing - review & editing. Thomas Schack: Conceptualization, Methodology, Formal analysis, Writing - original draft, Supervision, Methodology, Writing - review & editing.

Declaration of competing interest

The authors declare that they have no competing interests.