Family and biologic variables in the same etiologic studies of schizophrenia: a proposal

Abstract

The authors recommend combining biologic and family variables as follows: multifactor designs wherever possible, with at least one biological and one family variable among the set of two or more independent variables. The more frequent use of elementary psychological variables as dependent variables in addition to the symptoms of schizophrenia per se. These may include attentional variables, stimulus representational variables, short term memory variables, and cognitive variables. Biologic variables need not be modified from present studies and can certainly include genotypes, enzyme levels, and peripheral indicators of neural functioning such as the evoked response. The authors have formulated a number of requirements for the family variables: There should be evidence that they have a specific toxic effect likely to produce or to contribute, along with other variables, to a schizophrenic outcome or, even more desirably, to the elementary psychological variables whose deficit might constitute the core of schizophrenia. There should be evidence that the individual has had sustained exposure to the specific, toxic variable during a vulnerable period in his development. A variable must be adequately specified by a clear operational definition: it must be distinguished from similar variables and defined by its functional relationship with other family variables. Wherever possible, variables should not be measured by content analysis; rather, they should be assessed through their functional impact on other family behavior (e.g., the referential communication task).