This is a broad meta-analysis of the relations of both depression and stres
sors to immunological assays. The number of study samples (greater than 180
) and measures (greater than 40) is much more extensive than any so far. An
alyses are done by both fixed and random effects. By a fixed-effects analys
is, both major depression and naturally occurring acute stressors are assoc
iated with (1) an overall leukocytosis, (2) mild reductions in absolute NK-
cell counts and relative T-cell proportions, (3) marginal increases in CD4/
CD8 ratios, and (4) moderate decreases in T- and NK-cell function. However,
the degree of heterogeneity of the studies' results raises questions about
their robustness. Therefore, we also did the first random effects analysis
to estimate what is likely to appear in future studies. For depression, th
e analysis showed the immunological correlates included (1) an overall leuk
ocytosis, manifesting as a relative neutrophilia and lymphoenia; (2) increa
sed CD4/CD8 ratios; (3) increased circulating haptoglobin, PGE(2), and IL-6
levels; (4) reduced NK-cell cytotoxicity; and (5) reduced lymphocyte proli
ferative response to mitogen. For stressors, the random effects analysis sh
owed that future studies are likely to find the following effects: (1) an o
verall leukocytosis, manifesting as an absolute lymphocytosis; (2) alterati
ons in cytotoxic lymphocyte levels, CD4/CD8 ratios, and natural killer cell
cytotoxicity with the direction of change depending on the chronicity of t
he stressor; (3) a relative reduction of T-cell levels; (3) increased EBV a
ntibody titers; (4) reduced lymphocyte proliferative response and proportio
n of IL-2r bearing cells following mitogenic stimulation; and (5) increased
leukocyte adhesiveness. The random-effects analysis revealed that for both
major depression and naturally occurring stressors the following effects a
re shared: leukocytosis, increased CD4/CD8 ratios, reduced proliferative re
sponse to mitogen, and reduced NK cell cytotoxicity. The implications for t
hese findings for disease susceptibility and the pathophysiology of these c
onditions is discussed. (C) 2001 Academic Press.