Prevalence and risk factors for female sexual dysfunction in women attending a medical clinic in south IndiaJC Singh1, P Tharyan2, NS Kekre1, G Singh3, G Gopalakrishnan1
1 Department of Urology, Christian Medical College, Vellore, India
2 Department of Psychiatry, Christian Medical College, Vellore, India
3 Department of Anesthesiology, Christian Medical College, Vellore, India
Background: Reports from India on the prevalence and determinants of female sexual dysfunction (FSD) are scant. Aims: To determine the prevalence and risk factors for FSD. Settings and Design: A cross-sectional survey in a medical outpatient clinic of a tertiary care hospital. Materials and Methods: We administered a Tamil version of the Female Sexual Function Index (FSFI) to 149 married women. We evaluated putative risk factors for FSD. We elicited participant's attributions for their sexual difficulties. Statistical Analysis: We estimated the prevalence of possible FSD and sexual difficulties from published FSFI total and domain cut-off scores. We used logistic regression to identify risk factors for possible FSD. Results: FSFI total scores suggested FSD in two-thirds of the 149 women (73.2%; 95% confidence intervals [CI] 65.5% to 79.6%). FSFI domain scores suggested difficulties with desire in 77.2%; arousal in 91.3%; lubrication in 96.6%; orgasm in 86.6%, satisfaction in 81.2%, and pain in 64.4%. Age above 40 years (odds ratios [OR] 11.7; 95% CI 3.4 to 40.1) and fewer years of education (OR 1.2; 95% CI 1.0 to 1.3) were identified by logistic regression as contributory. Women attributed FSD to physical illness in participant or partner, relationship problems, and cultural taboos but none had sought professional help. Conclusions: Sexual problems suggestive of dysfunction, as suggested by FSFI total and domain scores, are highly prevalent in the clinic setting, particularly among women above 40 and those less educated, but confirmation using locally validated cut-off scores of the FSFI is needed.
Keywords: Anorgasmia, dyspareunia, female sexual dysfunction, female sexual function index
Though evidence suggests that healthy sexual functioning is an important contributor to women's sense of well-being and quality of life,  women and their clinicians often avoid discussion of this topic. Surveys done in the US and Europe have identified that female sexual dysfunction (FSD) is strikingly prevalent. , However, epidemiological investigations regarding the prevalence and factors associated with FSD from developing countries are limited. , In India, literature on the prevalence of sexual dysfunction among women is particularly scant. ,
Increasing age, urological, gynaecological, surgical and medical conditions, and interpersonal difficulties are the commonly identified risk factors for the development of FSD. ,,,, This study aimed to estimate the prevalence and determinants of FSD in a sample of women attending a general medical outpatient clinic at a teaching hospital in South India.
The study was conducted in the medical outpatient clinic of a teaching hospital in South India that provides primary through tertiary care. A sample size of 150 was aimed at based on the mean prevalence of 30% obtained from published surveys, ,,, and a precision of 5%. Oral informed consent was obtained from participants as many were illiterate. The study, including the informed consent procedure, was approved by the research and ethics committees (Institutional Review Board) of the institution.
Recruitment took place on two outpatient clinic days every week over a three-month period, during the hours when investigators (JCS, GS) were available. Consecutive, Tamil-speaking, married women who had registered in the medical outpatient clinic for unrelated medical illnesses, and married, Tamil-speaking, apparently healthy women accompanying other patients, were identified and referred by the medical records clerk to be screened by the investigators for eligibility. Women who were not living with their husbands in the previous six months were excluded. Eligible women were provided details regarding the survey and that they would be asked about their health and sexual functioning. The Tamil version of the Female Sexual Function Index (FSFI)  that had been previously pilot-tested for cultural appropriateness and linguistic accuracy was administered to consenting participants. Questions were asked in a standard format by a female investigator (GS) and the answers documented by the other investigator (JCS) to ensure uniformity in eliciting responses, since the study population included women who could not read, and to avoid potential social barriers due to a male investigator asking questions related to their sexual life. A minimum of 6-8 women were seen during each clinic in sessions lasting around 30 minutes and privacy was ensured within the limits of what is possible in a busy medical clinic.
The female sexual functioning index
The FSFI consists of 19 questions covering six domains - desire (two questions), arousal (four questions), lubrication (four questions), orgasm, satisfaction, and pain (three questions each).  Responses to each question relate to the previous month and are scored either from 0 (no sexual activity) or 1 (suggestive of dysfunction) to 5 (suggestive of normal sexual activity). Individual domain scores are obtained by adding the scores of the individual questions that comprise the domain and multiplying the sum by the domain factor provided in the FSFI for each domain.  The full scale score is obtained by adding the six domain scores. The minimum score possible is 2 and the maximum is 36.  The cut-off score used to demarcate sexual dysfunction on the total FSFI score was obtained from a validation study that compared FSFI scores in women with documented sexual dysfunction with those of dysfunction-free volunteers and determined a total score below 26.55 to denote sexual dysfunction.  The cut off scores to determine the presence of difficulties on the six domains of the FSFI were obtained from published sources; ,, accordingly scores less than 4.28 on the desire domain, less than 5.08 on the arousal domain, less than 5.45 on the lubrication domain; less than 5.05 on the orgasm domain; less than 5.04 on the satisfaction domain; and less than 5.51 on the pain domain were used to classify participants as having difficulties in that domain. The FSFI has been shown to discriminate reliably between women with and without female sexual arousal disorder and with or without female orgasmic disorder on each of the six domains: desire, arousal, lubrication, orgasm, satisfaction, and pain and has validated psychometric properties. ,
In addition to the FSFI, participants were asked about their attributions for their sexual difficulties using an unstructured free-flowing format.
All statistical analyses were performed using the statistical program Statistical Package for the Social Sciences, version 11.0. Univariate analyses were initially used to investigate the significance of putative risk factors for possible FSD; continuous variables were compared between women with and without possible FSD using the Student's t test or the Mann Whitney U test (if data were skewed). Chi squared tests or Fischer's exact test (as appropriate) were used to assess the effects of chance in differences in categorical variables, supplemented by odds ratios (OR) and 95% confidence intervals (CI) that were used to estimate the strength of associations. Measures that emerged statistically significant in univariate analyses were entered into unconditional, stepwise, forward, binary logistic regression with the presence or absence of sexual dysfunction (FSD) as the dependant variable and age less or more than 40 years, years of education, monthly income, and years of marriage as covariates. The robustness of the model was confirmed by repeating the analysis using unconditional, stepwise, backward logistic regression. One-way Analysis of Variance (ANOVA) was used in post hoc sensitivity analyses of FSFI scores to test the validity of age as a risk factor for FSD, as suggested by logistic regression, given the lack of validated cut off scores to define FSD or disorders on individual domains in the Indian population.
One hundred and forty-nine women who fulfilled intake criteria and consented to discuss issues related to sexual function were recruited. None of the women who were invited to participate refused consent, though 6 women were excluded due to lack of a husband living with them in the preceding six months. The mean age of patients in the study was 38.2 years (Standard Deviation (SD) 10.7 years; range 17-75 years). Of the 149 women, 29 (19.5%) were between 17 and 19 years, 64 (43%) were between 30 and 39 years, 35 (23.5%) were between 40 and 49 years, 14 (9.4%) were between 50 and 59 years, 6 (4%) were between 60 and 69 years and one woman (0.7%) was older than 70 years.
The mean number of years of education was 8.9 years (SD 3.8; range 0-16 years). Of these, 32 (22%) had less than six years of education, 78 (52%) had completed high school and 39 (26%) had some form of college education. The mean income of the sample was 5664 Rupees per month (SD 4771 Rupees; range 1000-35,000 Rupees). Of the sample, 72 (48%) earned less than 5000 Rupees per month.
The mean duration of marriage was 19.2 years (SD 12.1 years; range 1-50 years). Fifty eight (38.9%) lived in joint families. Five of the women (3.4%) were post-menopausal. Seventy women (49.7%) had genitourinary complaints such as burning on urination, stress incontinence, vaginal discharge, heavy bleeding during periods, etc. Of the 149 women, 102 (69%) had come for their own medical complaints such as hypertension, diabetes, headache, gastrointestinal complaints, joint pains, etc; the remainder were attendants of other patients.
Prevalence of FSD
Using the pre-determined cut off scores, 73.2% (109/149; 95% CI 65.5% to 79.6%) of women in this sample had total FSFI scores suggestive of FSD. Domain scores suggestive of difficulties related to desire was prevalent in 77.2% (115/149; 95% CI 69.8% to 83.2%); arousal in 136 (91.3%; 95% CI 85.7% to 94.8%); lubrication in 144 (96.6%; 95% CI 92.4% to 98.6%); orgasm in 129 (86.6%; 95% CI 80.2% to 91.1%); poor satisfaction in 121 (81.2%; 95% CI 74.2% to 86.7%); and pain in 96 (64.4%; 95% CI 56.5% to 71.7%).
[Table 1] displays the numbers and proportion of women with FSFI scores suggesting FSD and difficulties in the various domains in those above and below 40 years of age. All categories of difficulties were more common in older than in younger women with the prevalence increasing dramatically with each decade of age (data available on request). Women older than 40 years were significantly more likely to have scores suggesting difficulties with desire than those younger than 40 years (93% vs. 68%; OR 6.2; 95% CI 2.1 to 18.7). They were also more likely to have scores suggestive of difficulties with arousal (98% vs. 87%; OR 8.2; 95% CI 1.0 to 64.5); orgasm (98% vs. 80%; OR 14.1; 95% CI 1.8 to 108.7); satisfaction (97% vs. 72%; OR 10.5; 95% CI 2.4 to 46.1) and experiencing pain (84% vs. 53%; OR 4.7; 95% CI 2.1 to 10.7). Older and younger women did not differ significantly in proportions with scores suggesting difficulties with lubrication, though in both age groups the prevalence was very high (100% and 95%, respectively). Overall FSD, as indicated by the total FSFI scores, was more common in women older than 40 years than in those younger than 40 years (95% vs. 60%; OR 11.7; 95% CI 3.4 to 40.1).
Risk factors for FSD
On univariate analysis, age, years of education, monthly income, and years since marriage significantly differed between those with or without FSFI total scores suggestive of FSD, while menopausal status, living in a nuclear or joint family and presence of genitourinary complications did not [Table 2]. [Table 3] displays the results of forward and backward logistic regression that independently identified age above 40 years (adjusted OR 8.2; 95% CI 2.3 to 28.9) and fewer years of education (adjusted OR 1.2; 95% CI 1.0 to 1.3) as risk factors for the presence of possible FSD, correctly identifying 92.7% of those thus classified. Monthly income and years of marriage did not emerge significant in logistic regression as contributory.
The cut off values to define FSD and the six categories of sexual disorders were derived from reported validation exercises conducted elsewhere; these cut off values have not been validated in Indian women. To additionally check the validity of age below 40 years emerging as the single most important risk factor for the development of possible FSD and problems in its component domains, the mean scores of the FSFI and the domain scores (without application of any cut points to define dysfunction, problems or disorder) were compared post-hoc between those above and below 40 years of age. We used one way ANOVA to avoid the possibility of chance significance in FSFI domain scores occurring due to multiple comparisons [Table 4]. Women aged less than forty years consistently had significantly lower scores on all FSFI domains and in FSFI total scores than did women older than 40 years ( P = 0.000 for all comparisons). Also evident was the clear lack of overlap in the 95% CI for the FSFI means for the domain and total scores in the two age groups [Table 4]. This suggests that notwithstanding the lack of local validation of the cut-off scores, sexual difficulties in this sample were clearly associated with increasing age, as were difficulties in the individual sexual domains that are assessed by the FSFI and these mirror the results obtained using the published cut off scores.
Though the prevalence of scores indicative of FSD progressively increased with age of the women in the sample, the causes attributed by the women were multiple. Some attributed it to medical illness in their husbands (HIV infection, tuberculosis, paraplegia, chronic renal failure, diabetes, etc); others attributed it to the lack of desire, either for sex or for their spouses, while some considered it a taboo to consider sexual activity after the age of forty, though they still had the desire. Relationship problems, particularly husband's past infidelity in one case, were cited by some. A few felt it was not right to involve in sexual activity when they have adolescent children at home and alluded to lack of privacy and a sense of guilt. The prevalence of low scores on the FSFI was significantly higher among those less educated and women older than 40 years had significantly less years of formal education than did women younger than 40 years (mean 7 years (SD 3.7) versus mean 10.1 years (SD 3.4); P = 0.000). Of the 149 women, 72% had not attended college and the higher overall prevalence of SD among all ages of women in this sample may likely to be due to this.
None of the women in the sample had sought professional help for sexual problems and none seemed inclined to do as a result of the survey. Although not formally assessed with a validated instrument, many women seemed distressed by their lack of sexual fulfillment and none expressed reluctance to talk about sexual matters.
The salient findings of this survey were that FSD, suggested by published FSFI cut-off scores, was prevalent in two-thirds of this sample (73.2%, 95% CI 65.5 to 79.6%), of whom none had sought help. Difficulties in individual domains of sexual functioning, again as determined by published FSFI domain scores, were also highly prevalent (Desire 77.2%; 95% CI 69.8% to 83.2%; Arousal 91.3%; 95% CI 85.7% to 94.8%; Lubrication 96.6%; 95% CI 92.4% to 98.6%; Orgasmic 86.6%; 95% CI 80.2% to 91.1%; Satisfaction 81.2%; 95% CI 74,2% to 86.7%; Pain 64.4%; 95% CI 56.5% to 71.7%).
The prevalence of FSD increased from 60% in those below 40 years to 90% in those above 40 years of age. This dramatic increase in an already high prevalence was also seen in difficulties with individual domains of sexual functioning assessed by the FSFI.
Prevalence of FSD in developed and less developed countries
[Table 5] summarizes the results of 25 relevant studies from 1999 to 2009 that have systematically assessed the prevalence of FSD in clinic, community, and population-based samples of women. A direct comparison between these studies is hampered by the lack of a uniform, validated questionnaire that was used, characteristics of the study population, the method of assessment (self-rated questionnaire, personal interview, phone interview, mailed surveys), definitions of FSD, and other methodological issues such as sample size and acquisition, response rate, recall bias, inclusion criteria and age groups included. These result in widely differing prevalence rates across studies, particularly between community and population based studies and hospital based studies.
However, the results of hospital based studies from the developing world, , especially those that used the FSFI, ,, reveal prevalence rates for FSD ranging from 43 to 69%. The 95% CI for the prevalence of FSD from the present study of 65.5 to 79.6% indicate that while there is an overlap in prevalence rates across these studies, rates of FSD in this study are higher than those previously reported. The reasons for this could be methodological or socio-cultural.
Risk factors for FSD: Age and education
Apart from age that was strongly associated with the prevalence of possible FSD in this study, (older age is the commonest risk factor for FSD in studies that have reported this), less strong associations were seen for lower levels of education and FSD in our sample. A higher prevalence of FSD in less educated women was borne out by numerous other enquires [Table 5] across the developed and less developed world, ,,,,, though two studies, a hospital based enquiry from Nigeria  and a primary care population from Malaysia  both reported an association between higher education and the prevalence of FSD; the Nigerian study also reported that younger age was a risk factor for FSD [Table 5]. Thus, socio-cultural and methodological issues color our understanding of the true prevalence and risk factors for FSD worldwide.
Biological versus psychological risk factors for FSD
The "classic" risk factors for the development of erectile dysfunction in males such as hypertension, diabetes mellitus, hyper-lipidemia, and a history of cardiac diseases were generally not strongly correlated to the presence of FSD, underscoring the multi-causality of this disorders. ,, Hence we did not include analysis of these factors in our study. While diabetes has been associated with FSD in some studies, the confounding effect of age and other age-related physical problems need to be considered. It is suggested  that in men with diabetes, sexual dysfunction is related to somatic and psychological factors, whereas in women psychological factors are more predominant.
As the present enquiry was questionnaire-based, a formal gynaecological evaluation with assessment of vaginal atrophy and other factors likely to cause sexual dysfunction was not performed. The role of vaginal atrophy caused by peri- or postmenopausal oestrogen deficit for the development of FSD in our population therefore remains speculative.
The role of psychological factors is underscored by the ELIXIR study that reported high rates of sexual dysfunction in 4557 depressed patients,  and in other studies reviewed in [Table 5].
Lower urinary tract symptoms and urinary incontinence are important risk factors for FSD as well. , This study did not find an association between uro-gynaecological symptoms and FSD.
Socio-cultural contexts and FSD
The use of cutoff scores to define FSD and its components in epidemiological surveys may be confounded by complex cultural expectations and taboos that influence women's perceptions and expectations of their sexuality, thus rendering judgments of disorder and dysfunction subject to debate. In India, there are other issues that negatively affect the sexual health of women. As a woman matures, she is expected to maintain a culture of silence when confronted with issues of her own sexuality. This suppresses a women's ability to access information about her sexual health for fear that her reputation will be damaged. In many Indian languages, there is no precise word to translate 'orgasm.' In fact, one difficulty in culturally adapting the FSFI was to find a suitable word for orgasm since none exist in Tamil that is in wide use and the one finally used was chosen after seeking the opinion of 15 clinicians. Lack of privacy, foreplay, sexual openness and freedom, expressions of physical or emotional affection, coupled with ignorance of issues pertaining to sex and sexuality and poor communication regarding sexual matters, result in many women in India portrayed their experiences with sexual intercourse as furtive, shameful, acts in cramped, crowded rooms, lasting barely a few minutes and considered a duty; an experience to be submitted to, often from a fear of beating. 
In our survey, though what constitutes FSD in Indian context cannot be clearly defined, since we did not formally assess distress in a structured manner, it must be noted that many women studied did not have sexual satisfaction and orgasm and were dissatisfied, though fatalistic in their attributions.
The other limitations of this study are that a) it was conducted on a cohort of women attending the hospital for unrelated ailments, and these ailments may have influenced sexual functions. The community prevalence may also not be reflected by this study; b) only a limited number of risk factors were assessed due to concerns about the length of interviews; c) the FSFI has not been validated in India though we culturally adapted it for this study and did a sensitivity analysis that did not use predetermined, and un-validated cut off scores to define disorder and that clearly demonstrated differences in domain scores according to age, and d) the qualitative part of this study was unstructured.
However, this is the first detailed enquiry into FSD in India using standard assessment methods and it is noteworthy that Indian women were willing to talk about sexual issues freely. A community-based survey with appropriate sample size to identify the prevalence and risk factors accompanied by qualitative methods to better understand possible methods of intervention is warranted. The cut-off for classification needs to be determined based on FSFI scores obtained from women with and without documented FSD in our population. Population based studies including women of all age groups and unmarried women as well will better capture the range of normal sexual functioning and dysfunction and provide much needed data on a hitherto neglected area of enquiry in India.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]