This research examined sex differences in long-term outcomes after an FEP. Our results suggest that better outcomes for women found after 3 years of treatment at an EIS were underpinned by the presence of more favorable premorbid and baseline characteristics. After an average period of 10 years after treatment initiation, outcomes for women and men converged, with the exception of negative symptoms course. This draws particular attention to the increased dosage of antipsychotic medication for both sexes once they were discharged from the EIS program toward community-based services.

It could be said that women presented better functioning in terms of attaining milestones of adulthood such as completing school, having employment, leaving home, marrying, and having children. However, most of these circumstances were already present at illness onset, which might lead to the suggestion that women have an advantage in coping with difficulties associated with schizophrenia³. More women than men were partnered or married because women, whose illness begins later, often marry before their illness begins. Parenting was more common among women as well⁸. Those findings are consistent with the existing research of Hafner et al.¹⁰, supporting the mentioned advantages, such as higher rates of marriage and children, and lower rates of cannabis use. Seeman et al.³ found that despite women attaining better outcomes, there is a dependence on the social and cultural background of the study population. This is particularly the case in the exhibited maintenance of meaningful occupations, universally seen as positive and more frequent among females. In the same vein, both a higher age at illness onset of psychosis and better premorbid adjustment for women, which are favorable conditions for recovery, have been consistent findings¹¹.

Having a first psychotic break at an older age is a positive prognostic factor because it allows higher premorbid achievements, in terms of greater knowledge and experience in the world before the illness, and time for the development of better social skills and greater resilience¹². This is also confirmed in the present study concerning the higher education achievements and higher premorbid IQs in females, in line with a former study in our group¹³. Our study confirmed as well the replicated finding of equivalent cognitive impairment among male and female FEP patients¹⁴. It has been suggested that gender differences in neuropsychological impairments are evident in bipolar/mania and among depressive psychotic patients but with only minimal effects in schizophrenia¹⁵.

Women showed a marked recovery between the first and third year, but at 10 years the analysis of sex differences in functional recovery did not reach statistical significance. A similar finding was observed in a study carried out across diverse regions of the world¹⁶. Novick et al.¹⁶ found that in Central and Eastern Europe females showed higher rates of clinical remission at 36 months, with no sex differences in functional remission and recovery. Grossman et al.¹², who studied recovery over a span of 20 years, found that women showed significantly better global functioning and higher percentages of recovery in the 2- and 10-year follow-ups. Compared with 41% of the men, 61% of the women with schizophrenia showed a period of recovery at some point during the 20-year period. Dama et al.⁷ suggested that some men may need more than 2 years of EIS to achieve good clinical outcomes and effectively improve their overall functioning. We suggest that patients receiving EIS beyond 3 years may continue to experience further improvements in negative symptoms and functioning, resulting in increased rates of recovery.

Women showed a consistent pattern of less severe and better course of negative symptoms than men. Overall, this result is in agreement with most of the published literature, wherein men with schizophrenia appear to have more negative symptoms and more severe clinical features, particularly in social withdrawal, substance abuse, and blunted or incongruent affects than female patients¹⁷. Hafner et al.¹⁰ proposed that the apparent vulnerability of men to schizoid characteristics might explain, at least partially, the increased negative symptom severity, earlier age of onset, and worse overall social functioning. Younger age at illness onset in men occurs concurrently with more negative symptoms¹⁸. Women have advantages over men in that their illness starts at later ages and their symptoms respond more quickly and more completely to available treatments³. However, even when age of onset was addressed by assessing a uniform sample (younger than 40 years old), and when there was no sex difference in premorbid adjustment, women showed a better course of illness¹⁰. In general, females have lower levels of negative symptom severity, lower rates of alcohol/substance abuse, and higher percentages of having a spouse/partner and independent living¹⁶. This may indicate that a healthy lifestyle, presence of a spouse/partner, and having children are protective roles for females against negative symptoms.

This study provides new evidence to what happens with antipsychotic treatment when patients are discharged from EIS programs towards routine outpatient mental healthcare services. Females particularly showed an increase in dosage. Effective doses of antipsychotics in women may have to be lower than recommended for men, especially with regards clozapine and olanzapine, and it is evident that some antipsychotic side effects, such as weight gain, are more worrying for women¹⁹. Women may be more vulnerable than men to adverse reactions because the recommended doses in the market are calculated for a 70 kg (154 lb) man²⁰. The ability of women to respond to lower doses has been attributed to the effects of female hormones on the absorption and metabolism of antipsychotic medications, and to their relatively greater blood flow in the brain bringing more drug to receptors²¹. Brain structure, genetic and hormonal differences play a role on metabolic, cardiovascular, inflammatory, and immune parameters explaining sex differences in response to antipsychotics²⁰. Seeman et al.²² suggested that the dosage intervals for women should be longer than for men. Furthermore, a recent review suggests that adjuvant estrogen may make treatment more effective and may reduce the dose of antipsychotics and, therefore, help to prevent side effects²³. This applies to both sexes, but is especially applicable to women. Women require more thorough follow-up, including bone density scans, electrocardiograms, and mammograms. Meanwhile, the minimal effective dose of antipsychotics should always be selected. We found no differences in extrapyrimidal symptoms between male and female patients, which makes medication adjustments very important. Effective treatment means eliminating symptoms and improving function, but also the absence of side effects.

As expected, based on literature and in previous research in our group, females were older at illness onset and admission to the PAFIP program^4,14. Mean age at onset for females was 4 years higher than that for males. This finding is consistent with the nationwide assessment of life time risks for mental disorders conducted in Denmark. Paying attention to sex- and age-specific incidence rates, and sex- and age-specific cumulative incidences of mental disorders treated in secondary care during a full life time, Pedersen et al.²⁴ found that both sexes had almost identical incidence rates and cumulative incidences for schizophrenia during childhood and adolescence, but from 20 to 50 years of age, males had higher incidence rates and cumulative incidences. However, when Thorup et al.²⁵, also in Denmark, extended the age range up to 71 years, they found that males had the highest incidence rates from age 17 to 40, whereas females had the highest incidence rates from age 50 to 68, confirming that male sex is a major risk factor for the development of schizophrenia but showing that risk for women increases with increasing age. They suggest that traditions in the diagnosis system (the narrowing of the diagnostic criteria leads to more women being excluded from a diagnosis of schizophrenia), and reluctance in some clinicians to use the diagnosis of FEP in elderly people lead to biases that should be avoided in the future.

To our knowledge, there are no other published longitudinal studies that have explored sex differences in such long-term outcomes among FEP patients after being in an EIS program. Most of the studies have been conducted in one of three ways, such that patients were sampled from chronic patients, were matched for sex, or were statistically controlled for sex in the analyses. To overcome this limitation, we focused our analyses on sex differences in FEP patients. Regretfully, the current analyses did not explore the relative contributions of biological (i.e., plasma level of estrogen), psychosocial (i.e., resilience) and environmental factors to the sex differences and their interplay, which clearly warrants further research. The choice of better measures to assess function, particularly concerning DAS, is not without its critics. DAS is a quite an outdated measure, consisting of a criteria that does not necessarily reflect the real functioning observed in outpatients. Meanwhile, the follow-up in the PAFIP cohort was scheduled for 3 years, at which point, the patients were discharged from our EIS to community mental health services. For the purpose of the present study, they were contacted within a time window ranging from 8 to 16 years after enrolling in the program. We cannot rule out the influence of information that was not recorded but could have affected outcomes. Finally, our findings concerning exitus corresponds with the results from a recent published nationwide cohort study that included patients admitted for the first time to hospital during 2000–2014 for schizophrenia or schizoaffective disorder in Finland²⁶. Sommer et al.²⁶ found that during the 10-year follow-up, mortality was lower in women, with fewer suicide and cardiovascular deaths, but more cancer deaths. A detailed exploration on exitus was not an aim of the present study, although it is an important issue to research in the future.

Neither sex has the monopoly on good outcomes. However, women have certain advantages over men in that their illness starts at a later age, providing them with more time to achieve goals in life, which in turn results in better outcomes over the first 3 years of treatment in specialized programs. Yet, these benefits for women appear to dissipate, particularly once they are discharged from an EIS towards community-based services, a transition that is accompanied by an increase in the doses of antipsychotic medication. This pattern helps to pose the question of whether it is wise to target the sexes and lengthen EIS interventions. Exploring long-term sex differences in outcomes among FEP patients could provide useful information to better understand and to improve sex specific approaches to treat psychosis.