Rd. Gibbons et Jv. Lavigne, EMERGENCE OF CHILDHOOD PSYCHIATRIC-DISORDERS - A MULTIVARIATE PROBIT ANALYSIS, Statistics in medicine, 17(21), 1998, pp. 2487-2499
We applied a computationally practical form of probit analysis for mul
tiple response variables to data on early childhood development of fou
r psychiatric disorders: disruptive disorders (DD - attention deficit
disorders, oppositional defiant disorder, conduct disorder); adjustmen
t disorders (ADJ); emotional disorders (ED - all anxiety disorders, de
pression); and other DSM-III-R Axis I disorders (OTHER). In addition t
o estimating the intercept slope and higher order polynomial terms for
each age versus diagnosis regression, we estimated simultaneously the
correlation among the four diagnostic categories. We then took into a
ccount the correlation found among these four diagnostic categories wh
en testing the hypothesis of no age effect, which would have been igno
red in a piecemeal univariate approach. Regression lines for diagnosti
c prevalence indicate a linear increase for OTHER disorders, and a cur
vilinear increase for ED. We then used expected frequencies of individ
ual response patterns (that is, the 2(4) = 16 possible diagnostic comb
inations) in obtaining more precise estimates of diagnostic comorbidit
y and its relation to age. We further generalize the Beck and Gibbons
model to alternative specification of the random-effects distribution
(that is, they assumed multivariate normality), illustrate how one can
estimate the random-effects distribution empirically, and study the r
obustness of parameter estimates to specification of the random-effect
s distribution. (C) 1998 John Wiley & Sons, Ltd.