A prospective study of stability and change over 2 years of affective temperaments in 14-18 year-old Italian high...

It is generally accepted that temperament is not entirely stable, and that it changes with development, particularly in juvenile subjects; also, some temperaments are believed to be inherently more unstable. There is a great deal of current interest in Kraepelin's thesis that temperamental dysregulation in juvenile subjects represents the constitutional foundation from which the more florid episodes of manic-depressive illness arise; the cyclothymic, hyperthymic, depressive and irritable temperaments under consideration might represent the first observable phenotypes of the genetic diathesis for bipolarity. The analyses on the temperamental attributes in juvenile subjects were undertaken within this theoretical framework.

We evaluated 206 Italian high school students (14-18 years old) by means of a semi-structured affective temperament interview (TEMPS-I) at T0 and T1 two years later. Age, sex and psychometric properties of TEMPS-I raw scale score and weighted cut-off (as specially weighted linear combination of items) were used as predictive variables of stability.

Affective temperaments had a low to moderate level of stability, reaching 60% in the case of subjects with dominant cyclothymic temperament. The stability of the depressive temperament was primarily related to its weighted cut-off. The stability of the hyperthymic temperament appeared related to male sex, young age, and total scale score. Male sex represented the best stability predictor for the cyclothymic temperament as well. The group of subjects with an unstable depressive temperament showed a change toward the dominant cyclothymic temperament, whereas individuals with unstable hyperthymic temperamental traits moved on towards the dominant cyclothymic and depressive temperaments. The irritable construct was the least stable.

The infeasibility of a multiwave design represents the main limitation in evaluating the predictors of stability. Furthermore, in the present analyses, the size of the cyclothymic subsample was small.

Our data indicate considerable fluctuation and instability in depressive and hyperthymic temperaments in mid-adolescence. The cyclothymic temperament appears to be the most stable. Interestingly, cyclothymic moodiness appears more persistent in juvenile males; likewise persistent hyperthymic traits appear more of a "male" attribute. CLINICAL AND PUBLIC HEALTH IMPLICATIONS: We submit that these sex-relevant traits could be important in the risk of developing juvenile bipolarity. Literature review indicates that clinical studies, albeit on small samples, have already provided some support for this thesis. Larger studies on epidemiological samples could be more informative from a public health perspective. A user-friendly affective temperament questionnaire, which is under development, is critical for the methodology of such studies. Our study indicates that the present version of the Akiskal-Malya questionnaire can be easily used post-pubertally. Age adjustment must be considered for younger subjects.