Prevalence and Predictors of Sexual Dysfunction in Patients With Type 1 Diabetes

During a 2-year period, 240 consecutive patients with type 1 diabetes visiting the outpatient diabetes clinic of the University Hospitals of the Catholic University of Leuven were personally invited to participate. Patients were eligible for inclusion if they 1) were ≥18 years of age, 2) had had type 1 diabetes treated with an intensified insulin therapy, 3) did not have other health problems except complications secondary to diabetes, and 4) had had a stable heterosexual relationship for at least 1 year.

The methodology used in this study has been previously described (12). Patients were asked to complete at home a battery of self-report questionnaires to assess psychological adjustment to diabetes, diabetes-related quality of life, marital satisfaction, depression, and relevant aspects of sexual function.

HbA_1c values were determined using the Cobas Integra assay (Roche, Basel, Switzerland) with a normal range of 4.0–6.0%. The medical records were used to obtain data on medication use, BMI, and early-onset microvascular complications (neuropathy, nephropathy, and retinopathy).

The Appraisal of Diabetes Scale (ADS) was used to assess patients’ cognitive appraisal of diabetes, i.e., their thoughts about having diabetes (13). The ADS consists of seven items, for example, “How effective are you in coping with your diabetes?” and “To what degree does your diabetes get in the way of your developing life goals?” Scores can range from 8 to 31, and higher scores mean a more positive appraisal of diabetes. Acceptable 1-week test-retest reliability (r = 0.85), internal consistency (Cronbach’s α = 0.73), and convergent validity are reported (13).

The Diabetes Integration Scale (ATT19) was used to assess patients’ emotional adjustment to diabetes (14). This scale consists of 19 items, such as, “I dislike to be referred to as ‘a diabetic’” and “I try not to let people know about my diabetes.” Scores can range from 19 to 95, and higher scores indicate that patients are accepting of their diabetes, are comfortable with public awareness of their diabetes, have a sense of self-control, and feel well adjusted to their diabetes (14). Internal reliability for this 19-item scale ranged from 0.82 to 0.84 (14).

The Udvalg for Kliniske Undersoegelser is a well-validated questionnaire to assess side effects of psychotropic drugs in clinical studies and clinical practice (15). This questionnaire consists of 48 items; we used only three items that covered sexual functioning (decreased libido, erectile dysfunction/dry vagina, and ejaculatory dysfunction/orgasmic dysfunction). To come closer to the Diagnostic and Statistical Manual of Mental Disorders (4th edition) (DSM-IV) definition of a sexual dysfunction, we added to these items a formulation that enabled us to take into account the severity and to some extent the problematic status of a dysfunction in the scoring of sexual dysfunction. The Udvalg for Kliniske Undersoegelser is a checklist of specific symptoms scored on a 4-point scale, with scores ranging from 0 to 3. Higher scores indicate more severity of the reported problems. A “sexual dysfunction” was indicated if a score of 2 or 3 was mentioned, a score that also reflected marked distress. Good content and concurrent validity was reported for the whole questionnaire (15,16). The validity of the used subscale was based on its face validity assessed by an andrologist, a sexologist, and psychiatrists experienced in working with patients with sexual problems. Reliability analysis of the used subscale revealed a Cronbach’s α of 0.71, reflecting acceptable reliability.

The Beck Depression Inventory (BDI) was used to assess current self-reported symptoms of depression (17). Each item measures the presence and severity of a symptom of depression, and by adding the item scores, a total score is determined. A score of 16 on the BDI was used as a cutoff to indicate “clinical depression” (18). Higher scores indicate a higher number of depressive symptoms. Reliability analysis has shown a Cronbach’s α ranging from 0.92 to 0.93 and a good 1-week test-retest reliability of 0.93, and good content and construct validity were reported (17).

The Dyadic Adjustment Scale (DAS) was used to assess the quality of the marital relationship (19). The DAS consists of 32 items. Scores can range from 0 to 151, and the mean score in the general population is 100. Higher scores indicate better marital quality. The DAS has shown good reliability (Cronbach’s α = 0.94) and construct validity, as shown in high correlations with the Locke-Wallace Marital Adjustment Test (19,20).

HbA_1c was determined on a venous blood sample taken the day patients agreed to participate in this study.

Analyses were performed using the Statistical Package for Social Sciences (SPSS version 10.0; Chicago). Student’s t test, χ² test, Fisher’s exact test (FET), and the Mann-Whitney U test (mwU) were used to calculate differences between groups. Binary logistic regressions were used to study the predictors of sexual functioning for each sex separately, whereas the independent factors were first dichotomized (mid-split procedure) before being entered (forward conditional) in the regression. Where appropriate scores are presented as means ± SD. The level of significance used was P < 0.05.

Of 240 subjects, 222 patients (response rate 92.5%) agreed to participate and took the questionnaires home (50% women; 10 men and 8 women refused). Of those, 97 women (response rate 87.4%; 10 did not send back, 4 sent back a blank questionnaire) and 95 men (response rate 85.6%; 13 did not send back, 3 sent back a blank questionnaire) filled out the questionnaires properly and sent them back. Patients’ characteristics are shown in Table 1.

A total of 27% of women and 22% of men with type 1 diabetes reported sexual dysfunction [χ² (1) = 0.61, P = 0.49]. No significant differences were found in the distribution of the kind of sexual dysfunction between sexes. However, when taking into account the presence of complications, more women without complications reported sexual dysfunction [χ² (1) = 4.9, P = 0.037], and especially decreased sexual desire [χ² (1) = 4.3, P = 0.048], than men without complications (Table 2).

Women who reported sexual dysfunction were not significantly different for age, BMI, duration of diabetes, or HbA_1c compared with those not reporting sexual dysfunction. In men, no significant difference for HbA_1c was found between individuals with and without sexual dysfunction. Men with sexual dysfunction, however, were significantly older (46.9 vs. 34.5 years of age), had a higher BMI (25.6 vs. 23.9 kg/m²), and had a longer diabetes duration (17.5 vs. 11.7 years) (Table 3).

Women with and without sexual dysfunction reported more depressive symptoms than men with and men without sexual dysfunction, respectively (Table 3). Both women and men with sexual dysfunction reported more depressive symptoms than their respective counterparts without sexual dysfunction (Table 3). Based on a cutoff score of 16 on the BDI, four times more women with sexual dysfunction (37.7%) had scores suggestive of clinical depression (BDI score ≥16) than women without sexual dysfunction (8.3%) [FET: χ² (1) = 25.3, P < 0.001]. When using this cutoff (BDI score ≥16) in men, no significant difference for clinical depression was found between men with (15.0%) and without sexual dysfunction (4.2%) [FET: χ² (1) = 2.9, P = 0.12].

In women, an association was found between depression and sexual dysfunction. This association was found for sexual dysfunction in general [FET: χ² (1) = 11.7; P = 0.002], for libido decrease [FET: χ² = 14.9, df = 1, P = 0.001], and for arousal [FET: χ² (1) = 11.3, P = 0.003], but not for orgasm [FET: χ² (1) = 1.2, P = 0.28]. In men, however, no such association was found for the presence of sexual dysfunction in general [FET: χ² (1) = 2.9, P = 0.12], for libido decrease [FET: χ² (1) = 3.3, P = 0.13], for arousal [FET: χ² (1) = 4.7, P = 0.07], or for orgasm [FET: χ² (1) = 0.2, P = 0.51].

Women with sexual dysfunction reported a significantly lower overall quality of marital relationship than women without sexual dysfunction (mwU: z = −4.6, P < 0.001), a difference not found in men (mwU: z = −1.3, P = 0.19).

Women with diabetic complications did not report more sexual dysfunction (33%) than women without complications (22%) [FET: χ² (1) = 1.3, P = 0.34]. Men with diabetic complications were more likely to report sexual dysfunction (40.5%) than men without complications (6.1%) [FET: χ² (1) = 15.6, P < 0.001]. There was, however, for both men and women an association between the number of complications and the occurrence of sexual dysfunction: subjects with more complications were more likely to report more sexual dysfunction [χ² (12) = 30.9, P = 0.002; χ² (12) = 40.1, P < 0.001]. A comparison of men with and without complications revealed a significant association between complications and decreased desire [FET: χ² (1) = 5.2, P = 0.04], arousal dysfunction [FET: χ² (1) = 9.6, P = 0.002], and orgasmic dysfunction [FET: χ² (1) = 8.7, P = 0.005]. In women, however, no such association was found between complications and decreased desire [FET: χ² (1) = 0.01, P = 1.0], arousal dysfunction [FET: χ² (1) = 1.9, P = 0.22], or orgasmic dysfunction [FET: χ² (1) = 1.9, P = 0.22].

To find out which factors predict the presence of sexual dysfunction, several binary logistic regressions were performed for each sex. The independent factors that were taken into account after dichotomization comprised longer duration of diabetes, presence of complications, poor emotional and cognitive adjustment to diabetes, low-quality partner relationships, high age, poor glycemic control, and depression.

In men, the significant predictors for sexual dysfunction were presence of complications [95% CI for Exp(B) = 2.0–31.8] and high age [95% CI for Exp(B) = 2.0–48.8] (Table 4). In women, the significant predictor for sexual dysfunction was depression [95% CI for Exp(B) = 2.0–18.0] (Table 4). The results of the logistic regressions performed per sexual dysfunction for women and men separately are shown in Table 4.

In the present study, prevalence of sexual dysfunction in type 1 diabetic women (27%) and men (22%) was high (taking into account the young age of this sample) but comparable to previously reported data. In a smaller study, Jensen (21) found that 27.5% of women and 44% of men reported sexual dysfunction. The doubling of men reporting sexual dysfunction in Jensen’s study cannot be explained by a difference in mean age between his sample (35.8 years) and our sample (37.5 years). A possible explanation might be that Jensen did not take into account the severity and/or problematic status of sexual dysfunction. The prevalence of sexual dysfunction in patients with diabetes is higher than in the general population. Recently, Diemont et al. (22) reported a prevalence of sexual dysfunction of 14.9% in women and 8.7% in men in a community-based sample in the Netherlands. The findings of the present study thus confirmed our hypothesis that sexual functioning is affected not only in men with type 1 diabetes, but also in women with type 1 diabetes (11,12).

The comparison of the type of sexual dysfunction by presence of complications revealed that significantly more women without complications reported sexual dysfunction than men without complications, which was due to a higher number of women reporting decreased desire. This observation, however, parallels the sex difference in desire disorders found in community and clinical samples (23). A comparison of the type of sexual dysfunction in patients with complications revealed no significant differences between sexes, which suggests that the sexual response cycle of women and men with diabetes can equally be affected, as shown previously (3,11,24–26). Within-sex comparisons of patients with and without complications revealed a statistically significant association between complications and sexual dysfunction in men but not in women. This was remarkable because of the doubling of the percentage of women with desire or orgasm problems when comparing those with and those without complications.

Interesting sex differences were found concerning the predictors of sexual dysfunction in type 1 diabetic women and men. First, differences were found in the relation between sexual dysfunction and descriptive variables. In men, but not in women, evidence was found that higher age, higher BMI, poor glycemic control, longer duration of diabetes, and the presence of complications were associated with sexual dysfunction. These results confirm what is already known about men with diabetes, and they moreover suggest that in women, as opposed to men, sexual dysfunction is not related to diabetic complications.

Second, using binary logistic regressions, sex differences were found in the relation between sexual dysfunction and somatic and psychological variables. In men, the predictors of different sexual dysfunctions were higher age, worse glycemic control, complications, low-quality partner relationships, and poor emotional and cognitive adjustment to diabetes. This finding reveals that, in men, sexual dysfunction is related to both somatic and psychological variables. In women, however, the predictors of different sexual dysfunctions were all psychological variables (depression and poor cognitive adjustment to diabetes). This study is the first to confirm that, in men with diabetes, sexual dysfunction is linked to both somatic and psychological factors and that, in women with diabetes, sexual dysfunction is predominantly linked to psychological factors. The fact that in women no relation was found among diabetes-related factors, however, could be because we just relied on self-reported data. Schreiner-Engel (24) suggested that even if a complication such as peripheral neuropathy interferes with genital vasocongestion, a woman would probably not be aware of a decrease in lubrication and thus not report it. Therefore, it is important to use both subjective and objective methodology in future research in this field (11). Only two objective studies have been performed in this field and yielded contradictory results (27,28).

Furthermore, this study showed that more women than men with diabetes reported depressive symptomatology and that more women reached a BDI score suggestive of clinical depression (BDI score ≥16). These results, however, reflect the female-to-male ratio of depression in the general population (28,29) and confirm the results of a meta-analysis on comorbidity of depression in diabetes (30). The present study also revealed that patients with sexual dysfunction reported twice as much depressive symptomatology as patients without sexual dysfunction. The mean score of women with sexual dysfunction was almost as high as the cutoff score for clinical depression. This study further questions the relation between sexual dysfunction and depressive symptoms because, in women, depression was related not only to a decreased desire, but also to arousal dysfunction. Moreover, in men, no relation was found between depression and any sexual dysfunction. To unravel the relation between these interrelated factors in both sexes, future prospective research should focus on the longitudinal interaction between sexual dysfunction and depression in patients with diabetes.

Finally, this study has several limitations. First, because of the cross-sectional design, the temporality of the relation between the different studied variables and sexual dysfunction could not be evaluated. Therefore, future studies should use a longitudinal design to clear up sex differences in the temporal evolution of sexual dysfunction in diabetes. Second, caution should be exercised in interpreting these results because some of the analyses were based on small groups. Although our study is the largest ever done in the field of sexual function in diabetes, large numbers could not always be included in every analysis. Third, no control groups were included because the focus was on the comparison of the etiological factors related to sexual dysfunction in diabetes. Therefore, we just relied on general population data to interpret the frequencies of sexual dysfunction.

In conclusion, this study confirms that women with diabetes are at as much of an increased risk for sexual dysfunction as men with diabetes and that the predictors of sexual dysfunction are different for each sex. Therefore, sexual problems of women with diabetes should gain more attention in both clinical practice and research. Such increased attention could confirm the hypothesis that, in men, sexual dysfunction is related to somatic and psychological factors, whereas, in women, sexual dysfunction is predominantly related to psychological variables.