Am J Psychiatry 163:10, October 2006
This article is featured in this month’s AJP Audio.
Hyperlipidemia Following Treatment
With Antipsychotic Medications
Mark Olfson, M.D., M.P.H.
Steven C. Marcus, Ph.D.
Patricia Corey-Lisle, Ph.D., R.N.
A.V. Tuomari, M.S.
Patricia Hines, A.S.
Gilbert J. L’Italien, Ph.D.
Objective: This study attempted to esti-
mate the relative risk of developing hy-
perlipidemia after treatment with anti-
psychotics in relation to no antipsychotic
Method: A matched case-control analy-
sis was performed with pharmacy and
claims data from California Medicaid
(Medi-Cal). Patients were excluded if they
were treated for medical disorders or pre-
scribed medications known to increase
their risk of hyperlipidemia. Cases were
ages 18 to 64 years with schizophrenia,
major depression, bipolar disorder, or
other affective psychoses and incident hy-
perlipidemia. Cases were matched to up
to six control subjects by age, sex, race,
and psychiatric diagnosis. Both groups
were prescribed either no antipsychotic
medication or had two or more prescrip-
tions for one and only one antipsychotic
medication during the 60 days prior to
the first indication of hyperlipidemia
(cases) or matched index date (controls) in
the billing record. Conditional logistic re-
gressions were used to derive odds ratios
and 95% confidence intervals (95% CIs) of
each antipsychotic medication in relation
to no antipsychotic medication.
Results: A total of 13,133 incident cases
of hyperlipidemia were matched to
72,140 control subjects. As compared
with no antipsychotic medication, treat-
ment with clozapine (odds ratio: 1.82,
95% CI: 1.61–2.05), risperidone (odds ra-
tio: 1.53, 95% CI: 1.43–1.64), quetiapine
(odds ratio: 1.52, 95% CI: 1.40–1.65), olan-
zapine (odds ratio: 1.56, 95% CI: 1.47–
1.67), ziprasidone (odds ratio: 1.40, 95%
CI: 1.19–1.65), and first-generation antip-
sychotics (odds ratio: 1.26, 95% CI: 1.14–
1.39), but not aripiprazole (odds ratio:
1.19, 95% CI: 0.94–1.52) was associated
with a significant increase in risk of inci-
Conclusions: These findings suggest that
most commonly prescribed antipsychotic
medications increase the risk of develop-
ing hyperlipidemia in patients with
schizophrenia or mood disorders.
(Am J Psychiatry 2006; 163:1821–1825)
T here is accumulating empirical evidence and growing
clinical concern that some of the newer antipsychotic
medications may increase the risk of hyperlipidemia. Case
reports have linked treatment with clozapine and olanza-
pine to hyperlipidemia that disappears when antipsy-
chotic medications are discontinued (1–5). Medical record
reviews further support a connection between clozapine
and olanzapine and the increased risk of hypertriglyceri-
demia (6–9). A small prospective observational study
demonstrated that most patients developed hyperlipi-
demia during the first few months of olanzapine treat-
Clinical epidemiological studies provide a second line
of evidence linking treatment with antipsychotic medica-
tions to an increased risk of hyperlipidemia (11, 12). In one
case-control study, olanzapine and clozapine, but not ris-
peridone or combination therapy, were associated with a
significantly increased risk of hyperlipidemia (11). A sec-
ond study reported no overall differences in the risk of hy-
perlipidemia between patients treated with clozapine or
first-generation antipsychotic medications, although
clozapine was associated with an increased risk of hyper-
lipidemia in the young adult cohort (12).
Prospective research further suggests that antipsychotic
medications may adversely affect serum lipids. In one ran-
domized double-blind controlled trial, olanzapine and
clozapine resulted in significant increases in total serum
cholesterol (13). A second 3-week prospective randomized
trial reported that treatment with olanzapine or quetiap-
ine was associated with a significant increase from base-
line in fasting triglyceride levels (14). In a 4-week trial,
olanzapine (+13.8%) and risperidone (+5.1%) in combina-
tion with divalproex were associated with statistically
nonsignificant increases in total serum cholesterol (15).
In relation to the accumulation of evidence linking cloz-
apine and olanzapine to a worsening of the lipid profile,
considerably less is known about the metabolic effects of
the other second-generation antipsychotic medications.
In the current study, we estimated the relative risk of hy-
perlipidemia after initial treatment with first-generation