Is Pinpointing the Prevalence of Sexual Dysfunction in Schizophrenia an Exercise in Impossibility?

Sexual Dysfunction in Schizophrenia

Schizophrenia commonly develops in early adult life and commonly before the individual finds a lifetime partner. Given the longstanding nature of the illness, physical factors ranging from poor hygiene to obesity, and the common persistence of negative symptoms and social impairments, schizophrenia patients can be expected to experience difficulties in finding a life partner or a sexual partner, and even within a marriage, schizophrenia patients may experience difficulties in initiating sexual activity.

The availability of opportunity solves few problems. Schizophrenia patients commonly experience sexual difficulties related to low libido and erectile dysfunction arising from illness domains that are refractory to treatment, as well as arising from predisposing factors that include comorbid depression, anhedonia, relationship problems, smoking, obesity, atherosclerosis, diabetes mellitus, drug-induced sedation, drug-induced hyperprolactinemia, drug-related anticholinergic action, drug-related alpha-1 receptor blockade, and others.

Prevalence of Sexual Dysfunction in Schizophrenia

So, how common is sexual dysfunction in schizophrenia? This question was examined by Korchia et al. ³ in a recent systematic review and meta-analysis. These authors searched 5 electronic databases from inception to June 2022 for observational studies on the prevalence of sexual dysfunction in schizophrenia. They excluded interventional studies, inpatient studies, and studies on treatment-resistant patients because the samples in such studies would likely be biased; rather, they sought to determine the prevalence of sexual dysfunction in “more ordinary” patients.

The authors ³ identified 72 studies (pooled N = 21,076) conducted in 33 countries across 6 continents; Africa and Polynesia, however, were underrepresented. These studies had been published between 1979 and 2021, that is, spanning 4 decades. There were 35 studies that examined the prevalence of orgasmic dysfunction, 34 on decreased libido, 33 on erectile dysfunction, 19 on ejaculatory dysfunction, and 10 on genital pain. Related conditions examined included amenorrhea (6 studies) and galactorrhea (5 studies). Most studies (82%) examined the prevalence of sexual dysfunction using standardized instruments such as the Arizona Sexual Experience Scale.

The study ³ found high values for the pooled prevalence of sexual dysfunction. The figures were 56.4% (95% confidence interval [CI], 50.5 to 62.2) for overall sexual dysfunction, 40.6% (95% CI, 30.7 to 51.4) for decreased libido, 28% (95% CI, 18.4 to 40.2) for orgasmic dysfunction, and 6.1% (95% CI, 2.8 to 12.7) for genital pain.

In men, prevalences were 55.7% (95% CI, 48.1 to 63.1) for overall sexual dysfunction, 44% (95% CI, 33.5 to 55.2) for erectile dysfunction, and 38.6% (95% CI, 26.8 to 51.8) for ejaculatory dysfunction.

In women, the prevalence was 60.0% (95% CI, 48.0 to 70.8) for overall sexual dysfunction. Values were 25.1% (95% CI, 17.3 to 35.0) for amenorrhea and 7.7% (95% CI, 3.7 to 15.3) for galactorrhea.

The Elephant in the Room

Astonishingly, heterogeneity in this study ³ was very high in all analyses. For example, the I ² value for heterogeneity was 98% for overall sexual dysfunction in the whole sample; it was 98% for overall sexual dysfunction in men and 96% for overall sexual dysfunction in women. An examination of the forest plots for individual sexual dysfunctions, available in the supplementary materials, showed that heterogeneity was very high for all these analyses as well.

The heterogeneity did not arise from a few outlying study results. The forest plots, with study results arranged in order of effect size, showed linear patterns for increasing effect size. However, the minimum and maximum values varied very widely. For example, the prevalence of overall sexual dysfunction ranged from 3% to 97%; this range was 5% to 97% in men and 1% to 96% in women. Values for loss of libido ranged from 6% to 100%; for orgasmic dysfunction, it was 0% to 93%; for erectile dysfunction, it was 9% to 97%; for ejaculatory disorder, it was 6% to 89%; and so on. In other words, a wide range of possible values for prevalence was represented.

As a result of the wide range and the high heterogeneity, the 95% CIs were very wide for the pooled prevalence data. For example, as already stated in the previous section, the 95% CIs were 30.7 to 51.4 for decreased libido and 18.4 to 40.2 for orgasmic dysfunction. The implication is that the pooled estimates were imprecise.

Given the high I ² values, exploration of sources of heterogeneity, through subgroup and meta-regression analyses, was inevitable. The study found that a large number of variables influenced sexual dysfunction prevalences. These included the year of publication, study design, manner of recruitment, whether the study of sexual dysfunction was a primary objective of the study, whether dysfunction was assessed using a standardized instrument, whether dysfunction was assessed through patient reports, and others. Prevalences also varied by sex, employment status, illness severity and duration, presence of comorbid alcohol use disorders, medication use, and others. Important moderators not explored included anticholinergic drug use and presence of medical comorbidities.

Implications of the Findings

A key take-home message is that the prevalence of sexual dysfunction in schizophrenia varies very widely, depending on how patients are sampled, how patients are assessed, and what the sample characteristics are; the number of moderating variables is very large. In other words, it is not possible to easily generalize the pooled estimates, obtained in this meta-analysis, ³ to any specific population; there is likely to be a wide margin of error, as exemplified by the wide CIs around the pooled estimates. This leads to the unusual conclusion that pinpointing the prevalence of sexual dysfunction in schizophrenia may be an exercise in impossibility.

The above notwithstanding, the meta-analysis ³ was not an exercise in futility. It did draw attention to the fact that sexual dysfunctions are very common in schizophrenia and that there are identifiable risk factors for dysfunction, some of which are modifiable. Therefore, all patients with schizophrenia should be screened for sexual dysfunction as part of routine clinical assessment, and interventions should be offered, where appropriate. Sexual dysfunction in schizophrenia is a field that is wide open for intervention, and this is where future research should go.