Female genital cutting in Kilimanjaro, Tanzania: changing attitudes?


  • Sia E. Msuya,

    1. Department of Obstetrics and Gynaecology, The National Hospital, University of Oslo, Oslo, Norway,
    2. Department of International Health, Institute of General Practice and Community Medicine, University of Oslo, Oslo, Norway,
    3. Kilimanjaro Christian Medical Centre, Moshi, Tanzania
    Search for more papers by this author
  • Elizabeth Mbizvo,

    1. Department of Obstetrics and Gynaecology, The National Hospital, University of Oslo, Oslo, Norway,
    2. Department of International Health, Institute of General Practice and Community Medicine, University of Oslo, Oslo, Norway,
    Search for more papers by this author
  • Akhtar Hussain,

    1. Department of International Health, Institute of General Practice and Community Medicine, University of Oslo, Oslo, Norway,
    Search for more papers by this author
  • Johanne Sundby,

    1. Department of International Health, Institute of General Practice and Community Medicine, University of Oslo, Oslo, Norway,
    Search for more papers by this author
  • Noel E. Sam,

    1. Kilimanjaro Christian Medical Centre, Moshi, Tanzania
    Search for more papers by this author
  • Babill Stray-Pedersen

    1. Department of Obstetrics and Gynaecology, The National Hospital, University of Oslo, Oslo, Norway,
    Search for more papers by this author

Dr Sia E. Msuya Department of International Health, University of Oslo, Postboks 1130, Blindern, N-0317 Oslo, Norway. E-mail: siamsuya@hotmail.com


OBJECTIVES To study the prevalence, type, social correlates and attitudes towards female genital cutting (FGC) among urban women in Kilimanjaro, Tanzania; and to examine the association between FGC and gynaecological problems, reproductive tract infections (RTIs) and HIV.

METHODS In 1999, 379 women attending reproductive health care clinics were interviewed and underwent pelvic examination. Specimens for RTI/HIV diagnosis were taken.

RESULTS Seventeen per cent had undergone FGC, mostly clitoridectomy (97%). Female genital cutting prevalence was significantly lower among educated, Christian and Chagga women. Women aged ≥35 were twice as likely to be cut as those < 25 years. Seventy-six per cent of those who had undergone FGC intend not to perform the procedure on their daughters. Age < 25 years (P < 0.0001) and low parity (P < 0.01) were predictors of that intention. There was no association between RTIs, HIV or hepatitis B and FGC.

CONCLUSION FGC is still fairly common but there is evidence of a change of attitude towards the practice, especially among young women. The opportunity to educate women who attend reproductive health care facilities on FGC should be taken.


Female genital cutting (FGC), also known as female genital mutilation, comprises all procedures involving total or partial removal of the external female genitalia or other injury to the female genital organs for cultural or non-therapeutic reasons (WHO 1995). The practice is most prevalent in Africa, where it is estimated that more than 100 million girls and women have been subjected to some form of FGC and each year at least 2 million girls undergo the procedure (WHO 1995). The practice is found across socio-economic classes and among many ethnic and cultural groups in Africa. Muslims, Christians and followers of indigenous African religions practice FGC (Toubia 1993; Obermeyer 1999). Immediate and long-term complications have been reported in women who have undergone all types of FGC (Koso-Thomas 1987; MYWO 1991; Carr 1997; Jones et al. 1999; Shell-Duncan & Herlund 2000). The likelihood of experiencing long-term obstetric, gynaecological and genital-urinary tract problems increases with the severity of genital cutting (DeSilva 1989; MYWO 1991; Jones et al. 1999; Shell-Duncan 2001).

In Tanzania, 18% of all women have undergone FGC (Bureau of Statistics 1997). The practice is most common in the northern and central regions, where prevalence ranges from 20 to 81%. The age at which girls undergo FGC varies among ethnic groups, but according to the Bureau of Statistics (1997), nearly 70% are cut by the age of 15. The typical reasons for supporting FGC are numerous, including tradition and custom, religious requirement, rite of passage, cleanliness, better marriage prospects, prevention of promiscuity, preservation of virginity and increased sexual pleasure for men (Bureau of Statistics 1997; Chugulu 1998).

Kilimanjaro region, where we conducted our study, is in the northeast of Tanzania. It ranked fourth (37%) in the number of women who have undergone genital cutting among the 20 regions in the country (Bureau of Statistics 1997). Chugulu (1998) studied the reasons for perpetuation of the practice in rural Kilimanjaro. There is, however, a paucity of information regarding the prevalence, types and attitudes towards FGC among urban women.

The purpose of this study was to determine by clinical examination the current prevalence and types of FGC among urban women of childbearing age in Kilimanjaro. The intention to perpetuate the practice, the relationship between current reproductive tract infections (RTIs), hepatitis B and HIV and the relationship between current gynaecological problems and FGC were also examined.


The data were collected as part of a broader study to provide information about the prevalence and risk factors for sexually transmitted infections among women of childbearing age in Moshi urban district, the administrative capital of Kilimanjaro region. The study was conducted within the three largest government primary health care clinics between September and December 1999. Subjects were attending the clinics primarily for antenatal care, family planning or immunization of their children. In this area, 99% of all women attend the clinics for antenatal care, 90% attend for immunization and 20% for family planning services (Ministry of Health 1999); hence, they are considered fairly representative of the general population of childbearing women. The Tanzanian Ministry of Health and the Norwegian Ethical committee gave ethical approval for this study.

We invited 392 women to participate, and 379 (97%) agreed while 13 declined. Ten women declined the invitation citing time as a limiting factor. Three women agreed to undergo the interview but refused gynaecological examination. In total 13 women were excluded from the analysis. Trained interviewers then interviewed all consenting women to obtain information on demographic, behavioural, obstetric and gynaecological data and information about FGC. Those who had undergone genital cutting were further asked at what age, and whether they intended to continue the practice with their daughters. Reasons for their intention to continue or stop the practice were elicited.

In a separate private room, a female physician performed a pelvic examination on all participants and noted whether the woman had undergone genital cutting and the type, according to the World Health Organisation classification (WHO 1995). Any visible gynaecological complications that might have been the result of cutting were noted. Specimens for diagnosis of different RTIs were collected as described elsewhere (Msuya 2000).

Descriptive statistics (frequencies and percentages) were used to describe the data. Where relevant the χ2 test was used to test for group differences. All P-values were two-tailed, and a value of ≤ 0.05 was considered significant. Analysis was performed using the statistical software SPSS 8.0 for windows (SPSS, Chicago, IL, USA).


Of 379 women 63 had undergone genital cutting, giving a prevalence of 16.6%. Table 1 shows demographic, behavioural and clinical characteristics of the women. The likelihood of having undergone genital cutting was significantly higher among women who were not of the Chagga ethnic group, were of Muslim religious affiliation, were illiterate and had more than three children. Women who had undergone genital cutting were slightly older than those who had not (mean age 28 ± 6.7 vs. 26 ± 5.8 years ). Only 12% of women aged < 25 years had undergone cutting while 18.6 and 24.5%, respectively, of women between 25 and 34 and ≥ 35 years had undergone the procedure.

Table 1.   Demographic, behavioural and clinical characteristics among women who have undergone genital cutting Thumbnail image of

The age of sexual debut and number of sexual partners did not differ between women who had undergone FGC and those who had not, nor did the history of infertility. Women from ethnic groups in Arusha region had the highest prevalence of FGC (67%), followed by those from Dodoma (47%) and Kilimanjaro (36%) regions. Among the two major ethnic groups in Kilimanjaro region, Pare women had significantly higher FGC prevalence (35.8%) than Chagga women (4.8%) (Table 2).

Table 2.   Prevalence of FGC by ethnic groups and regions Thumbnail image of

Type 1, either partial (n=18) or total (n=43) clitoridectomy was the most common type of genital cutting among all the ethnic groups, occurring in 97% of FGC cases. There was no case of infibulation (Figure 1). Two women had undergone Type II cutting, i.e. clitoridectomy with total excision of labia minora; one from the Nyaturu ethnic group (central Tanzania) and the other from the Sambaa ethnic group (northern Tanzania).

Figure 1.

 Types of genital cutting among 63 women.

Age at which genital cutting was performed ranged from 6 to 18 years, with median age being 10 years (Table 3). By the age of 10, 59% of the affected women had undergone genital cutting, and the percentage rose to 84% by the age of 14. Chagga women underwent cutting when they were older (≥ 15 years) than those from other ethnic groups.

Table 3.   Age at genital cutting by ethnic groups Thumbnail image of

When asked if they intended to perform genital cutting on their daughters, 76% (48 of 63) of the cut women said they did not intend to do so, while 24% did. One woman had already performed genital cutting on her first daughter by the time of the interview. Perpetuation of tradition (67%) and the opportunity for teaching about marriage and life during the ritual (40%) were the most common reasons given for the intention to continue with the practice (Table 4). Fifty-three per cent of those intending to continue the practice on their daughters were of the Pare ethnic group. When all cut subjects were grouped by age, 5% of those under 25 years, 16% of those between 25 and 34 years and 75% of those ≥ 35 years intended to continue the practice on their daughters.

Table 4.   Reasons cited for intentions to continue or to stop the FGC practice Thumbnail image of

Of the 48 women who said they would not have their daughters undergo cutting (Table 4), more than 80% cited medical complications as the reason, while 73% of the women reported that its time had passed, and another 25% said it does not have any benefit. In univariate analysis, age < 25 years and low parity were significantly associated with the intention not to continue the procedure of FGC on their daughters (Table 5). However, when these factors were controlled for confounders by multivariate analysis using a model containing age, religion, parity, education and ethnic group, only age < 25 years (P=0.009) remained significantly associated with the intention of not performing FGC on their daughters.

Table 5.   Univariate analysis of significant factors associated with the intention not to perform FGC on the daughters Thumbnail image of

Table 6 shows the occurrence of sexually transmitted/RTIs among women by genital cutting status. Genito-urinary tract infections and other sexually transmitted infections were equally prevalent among women who had undergone genital cutting and those who had not. HIV-1 and current hepatitis B infections were not significantly different between the two groups.

Table 6.   Sexually transmitted/reproductive tract infections according to genital cut status among women in Kilimanjaro Thumbnail image of

Keloid scar was observed in 30% of the women who had undergone FGC. None had dermoid cysts, abscess or vesicovaginal fistulae. One of the women who had undergone excision had vaginal narrowing (stenosis) and reported painful intercourse secondary to difficult penetration.


Prevalence and social correlates

The results show that FGC is still prevalent among women of childbearing age in urban Kilimanjaro. The observed prevalence (16.6%) was considerably lower than the 41% reported among women in rural Kilimanjaro (Chugulu 1998). This difference may be because of a selection bias. Most of the women attending the clinics were young: 70% were < 30. Therefore, this clinic-based study may have missed older women who are no longer attending reproductive health care clinics and might influence the overall prevalence rate. The difference observed, however, might reflect a true lower prevalence of FGC in urban than in rural women, as was found in Mali ([7]Jones et al. 1999).

Nearly all (97%) of the cut women had undergone Type I cutting. Types II and III were rare in this population. This differs from the Tanzanian Demographic Health Survey data (1997), which showed that 57% of the women had undergone Type I cutting, 36% Type II and 5% Type III. Data in that survey, however, were based on self-reporting. Women's self-reporting of FGC status is less accurate than direct observation during pelvic examination, as was shown in Nigeria (Adinma 1997).

A significant inverse association was found between education and FGC. The prevalence of FGC decreased from 50% among illiterate women to 16.5 and 10%, respectively, among those with primary and secondary education. In Egypt, Sayed et al. (1996) reported that 92% of the mothers of circumcised daughters were illiterate, compared with 69% of mothers of uncircumcised daughters, while in Sierra Leone and Ghana lower rates of FGC among those with higher education were reported (Koso-Thomas 1987; Mbackéet al. 1998). In Kenya, secondary education was associated with a fourfold increase in disapproval of FGC (Njeru & PATH/Kenya Staff 1996). This may suggest the fact that education is one way of empowering women towards improved reproductive health outcomes. But the association between education and FGC is more complex than it appears. Parental background in terms of education, urban ethnic mix with groups that do not perform the procedure and information in the media may play a larger role in FGC than education per se. Further research is needed to address this issue.

The prevalence of 16.5 and 10%, respectively, among primary and secondary school leavers is still high, considering the recommendation to end the practice (WHO 1995). More than 90% of urban Tanzanian children attend primary school; therefore, integration of FGC education offers an excellent opportunity to target both boys and girls, tomorrow's parents, giving them preventive messages. For example, in this study it would have reached some still uncut (i.e. the Chagga, cut at 15 years or more), and might influence the attitude of those already cut. Positive attitudes formed at this younger age can give a foundation later for better reproductive health outcomes and a positive attitude towards their own sexuality and eventually that of their children.

The fact that age at sexual debut and number of partners did not differ between the women who had undergone cutting and those who had not refutes the belief that genital cutting reduces promiscuity and preserves virginity (Chugulu 1998).

Ethnic difference

Tribal difference regarding the age of cutting and prevalence was noted. Genital cutting for non-Chagga women was performed when they were children, at the median age of 9 years. The age for genital cutting in Egypt was reported to be 5–9 years and 9 years in Burkina Faso (Sayed et al. 1996; Jones et al. 1999). However, women of the Chagga ethnic group underwent cutting when they were older, as was documented in rural Kilimanjaro (Chugulu 1998). Interventions to eliminate the practice should, therefore, be tailored to the ethnic group. Where cutting is performed at a young age, parents should be the main target group for educational campaigns, whereas with the Chagga, both parents and adolescents should be targeted.

Kilimanjaro region has two major ethnic groups, the Chagga and Pare. The Bureau of Statistics (1997) reported that 37% of the women in the region had undergone FGC. However, we have noted that this proportion represents the prevalence only in the Pare ethnic group (36%), and is an overestimation for the Chagga, who had a 5% prevalence. Ethnic group should be taken into account when generating regional or national data for FGC and in devising intervention strategies.

Age and FGC

In this study, the prevalence of FGC consistently decreased with age. The highest prevalence was observed in those ≥ 35 years (24.5%), followed by the 25–34 age group (18.6%) with the lowest in those < 25 (12%). As the median age for genital cutting was 10 years, the majority of those who were going to be cut would have undergone the procedure by 15 years. This observation may imply that in this population a decline of the practice has occurred over time.

Intention to perpetuate

More than three-quarters of the women (76%) who had undergone cutting were in favour of not performing the procedure on their daughters. This is in contrast to the reports of Egyptian and Sudanese women where 98 and 83%, respectively, favour its continuation (DeSilva 1989; Sayed et al. 1996). In this study, young age was positively associated with the intention not to continue the practice. Only 5% of those < 25 years intended to continue FGC, while 16% of those 25–34 years old and a dramatic 75% of those ≥ 35 years intended to continue. This trend also supports the contention that attitudes toward FGC have changed over time. There may be several explanations for this observation. Most of the women (98%) had primary or secondary education, which may greatly contribute to the change of attitude towards the practice. Modernization, with its resulting change in lifestyle, urban ethnic mix, exposure to mass media, easier access to clinics and increased FGC information given at the clinics (indicated by the majority citing health complications as being the major reason to stop the practice), might have all contributed to that change. This observation needs to be validated by population-based surveys.

RTIs, hepatitis B, HIV and FGC

Infertility, endogenous and sexually transmitted RTIs were equally prevalent in genitally cut women and those who were not. Our findings are in contrast to those documented by DeSilva (1989) where a higher prevalence of candidiasis was found among cut women and to those by Morison et al. (2001) where a significantly higher prevalence of bacterial vaginosis and herpes simplex type 2 virus was found among cut women.

HIV and current hepatitis B infections were not associated with genital cutting, contrary to other reports which postulated that FGC may play a significant role in facilitating the transmission of bloodborne diseases such as hepatitis B and HIV (Brady 1999; Morison et al. 2001). Using the same non-sterilized instrument on several girls at a time (Koso-Thomas 1987) may not be the practice in Kilimanjaro. As shown in different settings the performers of the procedure are increasingly using new razors and surgical blades on each woman; hence, the lower risk for spreading these infections (Chugulu 1998; Shell-Duncan & Herlund 2000).

Keloid formation was the most observed gynaecological complication. Other authors have noted the same finding among women who have undergone clitoridectomy, 62% in Burkina Faso and 78% in Kenya (Jones et al. 1999; Shell-Duncan 2001).

Limitations of the study

The results of this study cannot be generalized to other parts of Tanzania where FGC is practised, as there are more than 100 ethnic groups in the country. The results can only be related to the ethnic groups represented in the study, especially the Chagga and the Pare.

Obstetric complications, psychological and sexual problems associated with FGC, were not addressed in this study. More information regarding these issues is needed.

A larger sample would have been desirable but pelvic analysis is time-consuming. With a prevalence of 16.6%, nearly 400 women had to be examined to find the 63 who had been exposed to FGC. We may thus have failed to show some associations that do exist and especially to elicit the more complex interrelationships between FGC and the other variables examined, because of the small sample size.


Based on these results, several strategies are suggested to help eliminate FGC:

• Ethnic differences should be taken into account.

• Information and education regarding the elimination of FGC should be integrated into reproductive health care clinics because these clinics are highly attended by women of childbearing age in this area. This offers an opportunity to target mothers/parents of young girls.

• Concentrated efforts should be made to reach the vulnerable/minority population in the community, e.g. those who do not attend the reproductive health care clinics or illiterate women. Community health care workers and women's groups can be targeted as principal educators.

• Education about the possible consequences and the need to prevent FGC should be integrated into the school health programs both in primary and secondary schools.

• A course concerning FGC consequences and revention should be developed for integration into the training of health care workers at all levels so as to increase their understanding of the health risks and social/moral issues and how to counsel affected women.

• Further research should include determining men's attitudes towards FGC.

In conclusion, a change of attitude towards the practice has been documented with declining prevalence by age. Public information campaigns targeting both men and women regarding FGC and its hazards should be strengthened to reinforce this positive behavioural change.


The authors thank the administration and staff of Majengo, Pasua and Mawenzi clinics where this study took place. We thank Rose Mushi, Tefonia Mbugu, Caroline Sululu and Candida Mafoi for data collection, and Robert Keig for critically reviewing the manuscript. We also thank the Tanzanian Ministry of Health, Regional medical officer and District medical officer for allowing us to conduct this study in MCH/FP clinics in Moshi. This study was supported by a grant from the Norwegian Cancer Society (Project no.: 68110/001) and the Letten F. Saugstads Fund.