Several previous studies have identified a continuity between childhood anxiety/withdrawal and anxiety disorder (AD) in later life. However, not all children with anxiety/withdrawal problems will experience an AD in later life. Previous studies have shown that the severity of childhood anxiety/withdrawal accounts for some of the variability in AD outcomes. However, no studies to date have investigated how variation in features of anxiety/withdrawal may relate to continuity prognoses. The present research addresses this gap.

Data were gathered as part of the Christchurch Health and Development Study, a 40-year population birth cohort of 1265 children born in Christchurch, New Zealand. Fifteen childhood anxiety/withdrawal items were measured at 7–9 years and AD outcomes were measured at various interviews from 15 to 40 years. Six network models were estimated. Two models estimated the network structure of childhood anxiety/withdrawal items independently for males and females. Four models estimated childhood anxiety/withdrawal items predicting adolescent AD (14–21 years) and adult AD (21–40 years) in both males and females.

Approximately 40% of participants met the diagnostic criteria for an AD during both the adolescent (14–21 years) and adult (21–40 years) outcome periods. Outcome networks showed that items measuring social and emotional anxious/withdrawn behaviours most frequently predicted AD outcomes. Items measuring situation-based fears and authority figure-specific anxious/withdrawn behaviour did not consistently predict AD outcomes. This applied across both the male and female subsamples.

Social and emotional anxious/withdrawn behaviours in middle childhood appear to carry increased risk for AD outcomes in both adolescence and adulthood.

The aim of this study was to develop empirically validated criteria for the diagnoses of clinically relevant somatization.

This study was performed in a population-representative cohort consisting of 461 males (47.8%) and 503 females (52.2%), with an average age of 55.8 years (s.d.=11.1). Somatization, anxiety and depression were derived from the Composite International Diagnostic Interview. Mplus was used to perform confirmative factor analyses on the current DSM-IV symptom groups; on alternative symptom clusters previously suggested; and to perform latent class analysis in order to define an empirically derived cut-off for somatization.

The existence of symptom groups as described in DSM-IV was not supported by our data, whereas a differentiation between cardiopulmonary, musculoskeletal, gastrointestinal and general somatic symptoms did fit our data. Latent class analysis revealed two classes characterized by few (n=859) and many (n=105) symptoms. The class of subjects could be approached by a simple cut-off of four functional symptoms (sensitivity 79%, specificity 98%, positive predictive value 82%, negative predictive value 97%) regardless of the number of organ systems involved.

This study in a large population-representative cohort suggests that a simple symptom count can be used as a dimensional diagnosis of somatization. In those instances in which a categorical diagnosis is preferred, a simple cut-off of four out of 43 functional symptoms best fitted our data. We did not find any added value for incorporating the number of symptom clusters into the diagnostic criteria.