Multiple regression models are increasingly being applied to clinical studies. Such models are powerful analytic tools that yield valid statistical inferences and make reliable predictions if various assumptions are satisfied. Two types of assumptions made by regression models concern the distribution of the response variable and the nature or shape of the relationship between the predictors and the response. This paper addresses the latter assumption by applying a direct and flexible approach, cubic spline functions, to two widely used models: the logistic regression model for binary responses and the Cox proportional hazards regression model for survival time data.

Psychotic symptoms in subjects with Alzheimer disease (AD with psychosis, AD þ P) define a phenotype characterized by greater cognitive burden than in AD without psychosis. We have proposed that genes of small effect may contribute to the risk for expression of psychosis in multiple disorders, including AD. Recently, sex-differential association of a three-locus haplotype, including a G-A transition at codon 108/158 of catechol-O-methyltransferase (COMT) resulting in a Val-Met substitution, has been reported to confer an increased risk for schizophrenia. The main objective of the study was to determine if COMT genetic variation is associated with risk of psychosis in AD, and included a case-control study of 373 individuals diagnosed with AD with, or without, psychosis. All subjects were characterized for alleles at the three loci associated with schizophrenia, RS737865, COMT G-A 108/158 (RS4680), and RS165599, and for a C/T transition adjacent to an estrogen response element (ERE6) in the COMT P2 promoter region. Both single locus and haplotype tests of association were conducted. Logit models were used to examine independent and interacting effects of alleles at the associated loci. All analyses were stratified by sex. In female subjects, RS4680 demonstrated a modest association with AD þ P; RS737865 demonstrated a trend towards an association. There was a highly significant association of AD þ P with the four-locus haplotype, which resulted from additive effects of alleles at RS4680 and ERE6 (or RS737865, as this locus was in almost absolute linkage disequilibrium (LD) with ERE6). In male subjects, no single locus test was significant, but there remained a strong association between AD þ P and the four-locus haplotype. This association appeared to result from interaction of the ERE6/ RS737865, RS4680, and RS165599 loci. Genetic variation in COMT is associated with AD þ P, and thus appears to contribute to psychosis risk across disorders. Sex-differential associations of COMT with psychosis may result from variation at, or in LD with, ERE6. Examination of variation at ERE6 in subjects with schizophrenia, and further examination of the independent and additive effects of variations in COMT on gene expression, is warranted.

No abstract

Objective Adolescents, elderly persons, African Americans, and rural residents with bipolar disorder are less likely than their middle-aged, white, urban counterparts to be diagnosed, receive adequate treatment, remain in treatment once identified, and have positive outcomes. The Bipolar Disorder Center for Pennsylvanians (BDCP) study was designed to address these disparities. This report highlights the methods used to recruit, screen, and enroll a cohort of difficult-to-recruit individuals with bipolar disorder. Methods Study sites included three specialty clinics for bipolar disorder in a university setting and a rural behavioral health clinic. Study operations were standardized, and all study personnel were trained in study procedures. Several strategies were used for recruitment. Results It was possible to introduce the identical assessment and screening protocol in settings regardless of whether they had a history of implementing research protocols. This protocol was also able to be used across the age spectrum, in urban and rural areas, and in a racially diverse cohort of participants. Across the four sites 515 individuals with bipolar disorder were enrolled as a result of these methods (69 African Americans and 446 non–African Americans). Although clinical characteristics at study entry did not differ appreciably between African Americans and non–African Americans, the pathways into treatment differed significantly. Conclusions Rigorous recruitment and assessment procedures can be successfully introduced in different settings and with different patient cohorts, thus facilitating access to high-quality treatment for individuals who frequently do not receive appropriate care for bipolar disorder.