A research label such as “persistent psychotic symptoms” needs its operational definition. In McKetin and colleagues’ 2016 study, it meant symptoms reported during at least one observed month of methamphetamine use and at least one observed month of abstinence. It did not mean that researchers continuously documented psychosis throughout the intervening years.
How was the cohort observed?
The prospective Australian study analyzed 278 people with methamphetamine dependence who did not meet lifetime DSM-IV criteria for schizophrenia or mania. Interviews occurred at baseline, three months, one year and three years. Each interview assessed the preceding month.
These were four noncontiguous observation windows, not thirty-six months of continuous symptom surveillance. The dataset contained 1,064 observed person-months because not everyone contributed the final interview. Most participants had injected methamphetamine, and the selected cohort should not be treated as representative of all people using the drug.
Why were only 232 people classified?
The symptom grouping excluded 35 participants who reported methamphetamine use at every observed follow-up, because symptom status during abstinence could not be evaluated. Eleven more had symptoms only during abstinent months and were excluded from these three groups.
The remaining 232 comprised 110 with no observed psychotic symptoms, 85 with symptoms only during use months and 37 with symptoms during both use and abstinent months. The reported 16% persistent-symptom figure was 37/232. It was not 16% of a community sample or a lifetime risk estimate for everyone who uses methamphetamine.
What counted as a psychotic symptom?
Trained interviewers used the Brief Psychiatric Rating Scale. The research threshold was a score of at least four on suspiciousness, unusual thought content or hallucinations. That criterion identifies clinically significant symptoms under the instrument’s rules; it does not independently establish a particular psychotic-disorder diagnosis.
Use and abstinence were largely assessed by report, with validation of abstinence in a subset of the larger cohort. No standardized methamphetamine dose or purity was assigned. A month classified as abstinent is also not equivalent to an experimentally controlled drug-free period for every participant.
Which associations distinguished the groups?
Conduct disorder and anxiety disorders were associated with both transient and persistent categories relative to the no-symptom group. Persistent symptoms were additionally associated with family history of primary psychotic disorder and major depression.
After adjustment for dependence severity, the reported odds ratios for persistent symptoms were 2.7 for family history (95% confidence interval 1.1–6.4) and 2.8 for depression (1.2–6.8). These describe group associations. They are not validated probabilities that classify an individual or prove which factor caused symptoms.
What could the observation windows miss?
Someone could have experienced symptoms between interviews and been classified as having no observed symptoms. Someone labeled transient could have experienced symptoms during an unobserved abstinent period. The authors specifically identified these possible misclassifications.
Only 37 people occupied the persistent group, limiting more detailed modeling. Depression could precede symptoms, accompany them or follow their consequences; the design did not resolve every temporal pathway. Several apparently predictive characteristics could also reflect related comorbidities.
The study is useful for distinguishing observed symptom patterns and identifying research questions. It cannot determine a particular person’s mental state at an unobserved incident, attribute every symptom to methamphetamine, or establish uninterrupted illness from four sampled months.
Source
McKetin R, Gardner J, Baker AL, et al. Correlates of transient versus persistent psychotic symptoms among dependent methamphetamine users. Psychiatry Research. 2016;166–171. doi:10.1016/j.psychres.2016.02.038.
Leave a Reply