This is one of the first studies to consider the outcome of individuals at clinical high risk of developing psychosis who do not go on to develop a full-blown psychotic illness over a 2½-year period. A recent, small Swiss study demonstrated evidence of transitory attenuated positive symptoms in some individuals who did not convert (13
). However, it was unclear whether these young participants improved or whether the baseline presentation of the nonconverting group was predictive of another disorder. Overall, in our study, the nonconverting group demonstrated significant improvement in attenuated positive symptoms, negative symptoms, and social and role functioning. Although symptoms declined significantly in this group overall, this was driven by a subgroup whose symptoms improved over time. Furthermore, more than 50% of this nonconverting sample no longer presented with any attenuated positive symptoms. However, despite a statistically significant improvement in functioning, this group remained at a lower level of functioning than nonpsychiatric comparison subjects. This suggests that initial prodromal categorization is associated with persistent disability, at least for 2.5 years.
Axis I diagnoses tended to diminish in number over time rather than emerge, although a relatively high proportion of clinical high-risk participants did have a mood and/or anxiety disorder. Only a small proportion demonstrated an emerging axis II disorder.
Strengths of this data set are its large sample size and the well-defined criteria for a psychosis-risk syndrome (8
). There are also limitations to these data. First, the follow-up evaluation was limited to 1 year and 2 years only. It is likely that a longer follow-up assessment period may have identified additional conversions, particularly since the mean age of this sample at the 1-year follow-up assessment was 19 years and the age of onset of a first episode of psychosis is, on average, 20–24 years (14
). Second, the data were originally collected at independently functioning sites not following a common protocol. Third, the outcome measures are constrained by what was common at all sites (8
). Fourth, it is possible that some individuals had episodes of transient psychotic-like experiences as reported in the general population (15
), but our clinical sample consisted of all help seekers and outcome is mixed in terms of risk status among individuals who do not seek help and are identified in population-based studies as having had psychotic-like experiences (16
). Fifth, other medications (e.g., antidepressants) may have had an effect on attenuated positive symptoms. We did not assess this because details of medication status were not universally collected, and for those who were receiving treatment with other medications, dosage and compliance were unknown. However, a study of the entire North American Prodrome Longitudinal Study sample demonstrated that with the same limitations, antidepressant use was not significantly associated with a decline in attenuated positive symptoms (17
). Finally, we excluded those individuals who were receiving treatment with antipsychotics, and thus we cannot say how they may have been distributed among the three groups (remission, symptomatic, converted).
In summary, we found that help-seeking individuals who met attenuated positive symptom risk criteria appeared to cluster into several groups over a 2.5-year period. Among the 255 participants with 1 year of follow-up data in the North American Prodrome Longitudinal Study, approximately 35% developed a psychotic illness, 24% remitted their attenuated symptoms, 20% retained attenuated positive symptoms at lower levels of severity than those present at baseline, and 21% received an antipsychotic drug while in the prodromal phase and therefore could not be used to represent the natural course. Thus, prodromal attenuated positive symptoms may predict a more severe condition in some but by no means all cases. Thus, better prediction requires a range of both clinical and biological marker-predictors in the future
Overall, our results suggest that persons who meet symptom and functional criteria for a psychosis-risk syndrome represent a collection of the following groups: 1) those who are truly at risk for psychosis and are showing the first signs of disorder, 2) those who remit in terms of the symptoms used to index clinical high-risk status, and 3) those who continue to have attenuated positive symptoms. Future work is needed not only to replicate these findings but to extend them over much longer follow-up periods and with more comprehensive assessments that were beyond the scope of this project.