To our knowledge, this the first study to comprehensively examine gender differences in illness onset and course in a treated epidemiological cohort of patients with FEPM. A further strength of the study is that we included all patients who first presented for treatment and not just those patients with more severe or chronic illness or those who have had their first inpatient admission. Similar to previous studies into patients with chronic illness and bipolar I disorder, we found an array of gender differences in patients with FEPM between the ages of 15 to 29 years. Key differences were found with respect to premorbid history, severity of illness at entry, treatment and discharge characteristics.
Univariate analyses indicated that males with FEPM were more likely to have a history of substance and forensic issues. It has previously been shown that males with bipolar disorder are more likely to have conduct problems or antisocial tendencies in childhood [7
] which may increase the likelihood of violence [31
] and problems with the law [14
]. Substance use (especially, alcohol, cannabis and nicotine) also tends to be more common in males than females with bipolar disorder [9
], and might explain the increased prevalence of forensic issues [34
] in our cohort.
Gender differences in the prevalence of substance use disorders have been reported in the general population, with males being more likely to abuse or be dependent on drugs [35
]. At service entry, 74.6% of males and 42.6% of females were using substances; much higher than rates in the general population. The persistent substance use in males, particularly prior to, and at the onset of the psychotic mania, might in part explain their more severe levels of psychopathology and poorer functioning (as measured by the GAF). This is supported by findings that substance use in patients with bipolar disorder is associated with poorer social adjustment and outcome [36
Males were more likely to use cannabis than females with FEPM; but no such gender differences were found with alcohol or polysubstance use. Cannabis use is common in patients with bipolar disorder, particularly in younger patients (< 30 years) [37
]. Cannabis use may be associated with an earlier age of onset of bipolar disorder, particularly in vulnerable individuals [38
]. However, no such gender differences in age of onset were found in the current study cohort.
During treatment, males were more likely than females to decrease or cease their substance use. Increased familial support in males might have contributed to the reduction of substance use. Males who are ill with a first episode of psychosis are more likely than females to receive understanding and support from family members [39
]. This may relate to cultural and societal differences in that underperformance and behavioural issues are tolerated more in males than females [39
]. Further, females with psychotic disorders such as schizophrenia are more likely to have experienced childhood adversities [40
], and are hence likely to leave home at an early age. Family involvement in early treatment can be pertinent for minimising the chances of substance use becoming a more serious disorder [41
]. An alternative explanation might be that the percentage of males who live with their families increases when they stop using substances. The decrease in substance use in males may also account for why there are no differences between the genders in terms of severity of illness and functioning at discharge.
Females were seven times more likely to have experienced sexual abuse than males with FEPM. A recent meta analysis indicated that the prevalence of sexual abuse is higher in females (19.7%) than males (7.9%) in the general population [44
]. However, in our sample approximately 30% of females with FEPM had been exposed to sexual abuse; a rate of trauma that is of concern. A relationship between sexual abuse or traumatic experiences and the onset of psychotic disorders has been proposed [45
], and this association may be stronger in females than males [49
]. Exposure to such trauma, may negatively affect brain development, through chronic hyper-reactivity of the hypothalamic-pituitary-adrenal (HPA) axis. This hyper-reactivity may then lead to disruption in the functions of dopaminergic and serotonergic systems, and may worsen prognosis. Such events sequelae have been described in the “traumagenic neurodevelopmental” model. Although the ultimate alterations to functions in dopaminergic and serotonergic may precipitate the onset of FEPM, there may be broader long-term effects. In females with chronic bipolar disorder, a history of sexual and physical abuse/trauma may relate to poorer mood outcomes such as more shifts in polarity and mixed episodes [50
]. Exposure to sexual abuse might also place females with bipolar disorder at a disproportionately greater risk of comorbid anxiety disorders such as post-traumatic stress disorder (PTSD) [14
]. Because our cohort was young, the possibility of the development of such (generally) longer-term sequelae might not have been captured in this group.
A history of sexual abuse may also explain why females with FEPM were less likely to be living with their families at discharge from the service. If the abuse occurred within the context of the family networks, then this is a unsurprising finding. However, it should also be noted, a history of sexual abuse and/or trauma is linked with markedly poorer social functioning especially with respect to building constructive and trusting relationships [52
With regards to the characteristics of the current study, a number of factors need to be kept in mind in interpreting the data. FEPM cohorts may offer a unique window into the disorder, as the effects of illness progression and treatment are controlled for. The primary limitation of this study is the use of a retrospective medical file audit. There can be numerous problems with such an approach including: (i) poor quality of documented information; (ii) clinical experience of raters; (iii) lack of inter-rater reliability; and (iv) questionable data validity. Numerous strategies were adopted in the current study to minimise the impact of such limitations including: (i) medical files at EPPIC were systematised according to FEP guidelines [53
]; (ii) medical files were rated by two consultant psychiatrists with expertise knowledge of EPPIC and the treatment of FEP; (iii) sound inter-rater reliability was determined for clinical and functioning measures; and (iv) concurrent validity of psychoses and baseline SUD was established for a sub-sample of patients [20
]. The study population was a catchment area sample, and is likely to be broadly representative of the population of FEPM (16).
A further study limitation was the lack of information recorded regarding psychopathology (e.g., mood episodes, psychotic symptoms, relapse) and type and dosage of pharmacological treatments; these factors could have had an effect on the degree of gender difference at discharge.