Schizophrenia Research 126 (2011) 110–116

Contents lists available at ScienceDirect

Schizophrenia Research
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s c h r e s

Schizophrenia patients at higher risk of diabetes, hypertension and

hyperlipidemia: A population-based study
Chun-Hui Liao a,b, Chen-Shu Chang c, Wan-Ching Wei b, Shih-Ni Chang b,d, Chien-Chang Liao b,d,
Hsien-Yuan Lane a, Fung-Chang Sung b,d,⁎
a
Department of Psychiatry, China Medical University and Hospital, Taichung 404, Taiwan
b
Department of Public Health, China Medical University and Hospital, Taichung 404, Taiwan
c
Department of Neurology, Changhua Christian Hospital, Changhua City 500, Taiwan
d
Management Office for Health Data, China Medical University and Hospital, Taichung 404, Taiwan

a r t i c l e i n f o a b s t r a c t

Article history: Objective: This study investigates risks of developing diabetes mellitus (DM), hypertension,
Received 10 March 2010 and hyperlipidemia in treating schizophrenia with first- and second-generation antipsychotics
Received in revised form 6 December 2010
(FGA and SGA, respectively).
Accepted 8 December 2010
Methods: We established two study sets, each consisting of patients with schizophrenia and
Available online 8 January 2011
without schizophrenia, from the insurance claims from 1997 to 2000. Study set I had 1631
Keywords: patients taking FGA and 6524 non-schizophrenia controls; the other had 1224 patients taking
Antipsychotics SGA and 4896 controls. Controls were selected frequency matched with sex, age and the index
Retrospective cohort study year. All subjects were free of the studied metabolic disorders at the baseline. We measured
Metabolic abnormality incidences of these disorders developed by the end of 2008 in each cohort and their respective
Schizophrenia hazard ratios (HRs) for these disorders.
Results: Schizophrenic patients taking FGA were older than those taking SGA. In the Cox
models, significance adjusted HRs associated with SGA were 1.82 (95% confidence interval (CI)
1.30–2.55) for DM and 1.41 (95% CI 1.09–1.83) for hyperlipidemia. For those on the FGA, the
risk was only significant in developing DM (HR 1.32, 95% CI 1.01–1.75). The age-specific
antipsychotics-associated risks for metabolic disorders were higher in young patients than in
older patients particularly for hypertension; the HRs in 10–19 years of age were 4.52 (95% CI
1.76–11.6) associated with FGA and 3.92 (95% CI 1.83–8.39) associated with SGA.
Conclusions: Patients with schizophrenia on SGA have higher risk of developing metabolic
disorders than those on FGA. It is likely that older patients have already gone through the age of
developing these side-effects and were free of them at the baseline.
© 2010 Elsevier B.V. All rights reserved.

1. Introduction disease are at a higher risk of mortality, with a life expectancy

Schizophrenia is a significant mental illness with preva-
lence rates ranging 1–17 per 1000 worldwide (Chien et al.,
2009; Warner and de Girolamo, 1995). Patients with the
⁎ Corresponding author. China Medical University College of Public Health,
91 Hsueh-Shih Road, Taichung 404, Taiwan. Tel.: + 886 4 2203 5740; fax:
+ 886 4 2201 9901.
E-mail addresses: fcsung@mail.cmu.edu.tw, fcsung1008@yahoo.com
(F.-C. Sung).
approximately 20% shorter than the general population
(Newman and Bland, 1991). Coronary artery disease and
other cardiovascular diseases are also the major disease
associated with schizophrenia (Brown, 1997; Hennekens et
al., 2005). Individuals with schizophrenia commonly require
long-term antipsychotic treatment, and have an unhealthy
lifestyle with poor diet selection. These patients are at an
elevated risk of weight gain, leading to obesity and metabolic
syndrome (Bobes et al., 2007; Cohn et al., 2004; De Hert et al.,
2006; Hagg et al., 2006; Huang et al., 2009; McEvoy et al.,

0920-9964/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.schres.2010.12.007
 C.-H. Liao et al. / Schizophrenia Research 126 (2011) 110–116
2005; Suvisaari et al., 2007). The prevalence of metabolic
syndrome among patients with schizophrenia may range
from 24% to 43% in males and 27%–52% in females. A recent
population-based period-prevalence study in Canada showed
that patients with schizophrenia have an elevated prevalence
of diabetes mellitus (DM), hypertension, hyperlipidemia and
cardiovascular diseases (Bresee et al., 2010). Lin et al. (2008)
even found that patients with schizophrenia have a higher
risk of stroke. Hypertension, DM and hyperlipidemia are
known risk factors of stroke. Chien et al. (2009) also found
that the prevalence of DM in patients with schizophrenia was
around 7.90% in Taiwan. Central obesity, hypertension,
hyperglycemia and dyslipidemia are the core problems of
metabolic syndrome that contribute to the high prevalence of
cardiovascular disease (Grundy et al., 2004; Laaksonen et al.,
2004; Mokdad et al., 2003).
The association between metabolic abnormalities and
schizophrenia has reached new developments in recent years
after the use of second-generation antipsychotics (SGA).
Studies have suggested that the medication may contribute to
metabolic abnormalities, particularly the risk of diabetes
(Hennekens et al., 2005; Newcomer, 2007; Scheen and De
Hert, 2007; Tschoner et al., 2007). Two well-designed multi-
city clinical trials in adults (Lieberman et al., 2005) and youth
(Sikich et al., 2008) with schizophrenia revealed that
olanzapine can most likely increase weight gain and
measures of other metabolic abnormalities. In a meta-
analysis, Smith et al. (2008) found that SGA are associated
with slightly increased risk of diabetes, compared with the
first-generation antipsychotics (FGA). Most of these studies
were limited by small sample sizes. In a retrospective chart
review study for patients with three-year treatment expo-
sure, De Hert et al. (2008) found a more than three-fold
higher risk of metabolism syndrome for patients started with
SGA than those started with FGA. Most studies appeared to
lack of comparison with the general population in measuring
the effectiveness of medications and were limited by short
follow-up time.
We conducted a 12-year follow-up, population-based
cohort study to estimate the risks of developing DM,
hypertension and hyperlipidemia in schizophrenic patients
prescribed with FGA and SGA and compared them with their
corresponding non-schizophrenia controls. The results de-
rived from this study can provide an evidence-based data
plan for a more comprehensive approach in the long-term
care of schizophrenic patients taking into consideration the
prevention of metabolic syndrome and related cardiovascular
diseases.
2. Methods
2.1. Databases of Taiwan National Health Insurance
This study used claims data of the National Health
Insurance (NHI) obtained from the National Health Research
Institute (NHRI) in Taiwan from 1996 to 2008. The Depart-
ment of Health in Taiwan incorporated 13 insurance
programs to launch the universal NHI program in March
1995. This insurance program has covered 99% of all 23
million citizens and contracted with 97% of hospitals and
clinics in Taiwan (Cheng, 2009). The NHI program provides
111
comprehensive medical services, including outpatient and
inpatient cares, Chinese medicine, dental care, childbirth,
physical therapy, preventive cares, home care, and rehabil-
itation for chronic mental illness. Using the electronic claims
database, the NHRI randomly selected from the insured
population to establish a representative reimbursement
claims dataset of a sample of one million insured persons in
2000. We were able to use scrambled patient identifications
to link files, including the information on patient gender,
birth date, and other demographic data. The scrambled
identification system safeguards the insured population
against the violation of personal privacy and confidentiality.
The codes of health care received, codes of health care
providers, and medical costs covered by the insurance
program were also available. The diagnoses were coded
using the International Classification of Diseases 9th Revision
of the Clinical Modification (ICD-9-CM).
2.2. Study samples
From the dataset, the claims data for 837,810 insured aged
10 years and above defined in 1997 were extracted, approx-
imately equivalent to 3.6% of all enrollees in the insurance
system. Based on this sub-dataset, patients with histories of
DM (ICD-9-CM codes 250), hypertension (ICD-9-CM codes
401–405) and/or hyperlipidemia (ICD-9-CM codes 272.0–
272.4) identified were excluded from the study samples
selection. Patients newly diagnosed with schizophrenia (ICD-
9-CM 295) in 1997–2000 that were prescribed with FGA and
SGA were identified and defined separately as two treatment
cohorts, the FGA cohort and SGA cohort, respectively, in two
study sets, set I and set II. For comparison, two cohorts of non-
schizophrenia controls were also established by random
selection from the insured population without the history of
schizophrenia or other psychiatric illness. These controls
were frequency-matched with age, sex and index year of the
cases in the corresponding schizophrenia cohort, with a
control sample size four-fold of the corresponding schizo-
phrenia cohort. Each study subject was followed until a
diagnosis of DM, hypertension or hyperlipidemia was made,
or until the time the subject was censored for loss to follow-
up, death, or termination of insurance, or up to 31 December
2008, the end of the follow-up. We followed these cohorts as
fixed groups for this study. To ensure the validity of the
diagnoses of the three major follow-up outcomes, we
identified the patients through both ICD-9-CM disease
codes and medications prescribed. Medications of FGAs and
SGAs were listed in the Appendix A.
2.3. Statistical analysis
Demographic factors, including age, sex, occupation,
population density, and income between each treatment
cohort and the controls were compared. We categorized the
ages into four groups (10–19, 20–39, 40–59, and 60 or more
years), and occupations into three groups (white collar, blue
collar and others). All the insured were grouped into four-
quartile groups from high to low based on the population
density in the areas the subjects lived. The subjects' incomes
for insurance premium estimation were classified into three
groups with monthly pay of b15,000, 15,000–29,999, and
112 C.-H. Liao et al. / Schizophrenia Research 126 (2011) 110–116

≥30,000 New Taiwan Dollars (1.0 USD is about 31–34 NTD).
The Chi-square test was used to examine the differences.
The sex-specific and age-specific incidence density rates of
the studied disorders were calculated for each cohort. The
case cohort to the control cohort rate ratios with 95%
confidence intervals (CI) were calculated by sex. Cox
proportional hazard regressions were used to measure the
risk of developing DM, hypertension or hyperlipidemia
associated with schizophrenia for each study set. The hazard
ratio (HR) was presented with 95% CI, controlling for socio-
demographic variables used for adjustment in the multivar-
iable Cox analyses. This study used the SAS statistical package
(SAS Institute Inc., Cary, NC. Version 9.1) to perform all
analyses, with the statistical adopted significance level at
0.05.
3. Results
Study set I consisted of 1631 newly diagnosed schizo-
phrenia cases in the FGA cohort and 6524 non-schizophrenia
controls during the period of 1997–2001 (Table 1). The
corresponding sample sizes in study set II were 1224 cases
and 4896 controls. Cases in both schizophrenia cohorts had
lower income, and higher un-employment, low income and
retired rates, and resided in less urbanized areas compared
with the respective controls. However, the two cohorts in
study set I were older than the two cohorts in study set II.
Table 2 shows that the incidence density rates of DM,
hypertension and hyperlipidemia by sex in study set II tended
to be higher in the schizophrenia cohort on SGA than in the
non-schizophrenia controls. The incidence density rates of
DM, hypertension and hyperlipidemia were 1.90, 1.19 and
1.56 times higher in patients taking SGA than in their
corresponding controls, respectively. The incidence of DM
was the highest in the SGA cohort, followed by FGA cohort,
control cohort I, and the lowest in the control cohort II. The
hierarchy was particularly true among female subjects, with
incidence density rates of 6.29, 5.40, 3.94 and 2.65 per 1000
person-years, respectively. Comparing the incidences of DM,
hypertension and hyperlipidemia between the cohorts of
study set I, only the incidence of hyperlipidemia in females
was significantly higher in the FGA cohort than in the
controls.
Table 3 presents the age-specific HRs of DM, hypertension
and hyperlipidemia associated with schizophrenia control-
ling for the socio-demographic variables. The results indicate
that younger schizophrenic patients had higher risk of
developing DM and hypertension for both study sets, and
developing hyperlipidemia for study set II. The HR of
developing hypertension in teens with schizophrenia was
even greater for those taking FGA (HR 4.52, 95% CI 1.76–11.6)
than for SGA (HR 3.92, 95% CI 1.83–8.39).
Table 4 shows that, in the multivariable Cox proportional
hazard analyses, schizophrenic patients taking FGA had a
significant association with developing diabetes (HR 1.32,

Table 1
Comparison of socio-demographic characteristics between schizophrenic patients on first-generation antipsychotics (FGA) and non-schizophrenia controls, and
between schizophrenic patients on second-generation antipsychotics (SGA) and non-schizophrenia controls identified in 1997–2001.

Variables Non-schizophrenia Schizophrenia with FGA Non-schizophrenia Schizophrenia with SGA

N = 6524 N = 1631 N = 4896 N = 1224

n (%) n (%) n (%) n (%)

Age(years)
10–19 752 (11.5) 188 (11.5) 740 (15.1) 185 (15.1)
20–39 3176 (48.7) 794 (48.7) 2968 (60.6) 742 (60.6)
40–59 1780 (27.3) 445 (27.3) 1036 (21.2) 259 (21.2)
≥60 816 (12.5) 204 (12.5) 152 (3.1) 38 (3.1)
p value a 1.00 1.00
Sex
Female 3008 (46.1) 752 (46.1) 2168 (44.3) 542 (44.3)
Male 3516 (53.9) 879 (53.9) 2728 (55.7) 682 (55.7)
p-value a 1.00 1.00
Occupation
White collar 3610 (55.3) 629 (38.6) 2909 (59.4) 483 (39.5)
Blue collar 2080 (31.9) 500 (30.7) 1384 (28.3) 281 (23.0)
Others b 834 (12.8) 502 (30.8) 603 (12.3) 460 (37.6)
p value a b0.0001 b 0.0001
Urbanization c
1 1930 (29.6) 436 (26.7) 1570 (32.1) 336 (27.5)
2 1925 (29.5) 425 (26.1) 1414 (28.9) 358 (29.3)
3 1224 (18.8) 316 (19.4) 958 (19.6) 254 (20.8)
4 1444 (22.1) 454 (27.8) 953 (19.5) 276 (22.6)
p value a b0.0001 0.007
Income
b15,000 2163 (33.2) 852 (52.2) 1534 (31.3) 710 (58.0)
15,000–29,999 3218 (49.3) 680 (41.7) 2454 (50.1) 472 (38.6)
≥30,000 1143 (17.5) 99 (6.1) 908 (18.6) 42 (3.4)
p value a b0.0001 b 0.0001
a
Chi-square test.
b
Others: including un-employment, low income and retired.
c
1 indicates the highest degree of urbanization of residential areas and 4 the lowest.
 C.-H. Liao et al. / Schizophrenia Research 126 (2011) 110–116 113

Table 2
Incidences of diabetes mellitus, hypertension and hyperlipidemia occurred during the follow-up period compared between cohorts with and without
schizophrenia by sex and medication.

Variables Non-schizophrenia FGA cohort FGA-to-controls Non-schizophrenia SGA cohort SGA-to-controls

a b a b a b a b
Cases Rate Cases Rate RR (95% CI) Cases Rate Cases Rate RR (95% CI)

Diabetes
Female 116 3.94 35 5.40 1.37 (0.94–2.00) 56 2.65 31 6.29 2.37 (1.53–3.68)
Male 131 3.85 31 4.30 1.12 (0.76–1.65) 73 2.78 27 4.32 1.55 (1.00–2.41)
Total 247 3.89 66 4.82 1.24 (0.94–1.63) 129 2.72 58 5.18 1.90 (1.40–2.60)
Hypertension
Female 664 25.2 145 24.7 0.98 (0.82–1.17) 370 19.1 91 19.3 1.01 (0.80–1.27)
Male 755 24.8 166 25.1 1.01 (0.86–1.20) 387 15.6 124 21.2 1.35 (1.11–1.66)
Total 1419 25.0 311 24.9 1.00 (0.88–1.13) 757 17.2 215 20.3 1.19 (1.02–1.38)
Hyperlipidemia
Female 196 6.76 59 9.39 1.39 (1.04–1.86) 109 5.21 39 7.93 1.52 (1.06–2.19)
Male 241 7.20 42 5.87 0.81 (0.59–1.13) 138 5.32 52 8.42 1.58 (1.15–2.18)
Total 437 7.00 101 7.51 1.07 (0.87–1.33) 247 5.27 91 8.20 1.56 (1.22–1.98)

FGA, first-generation antipsychotics; SGA, second-generation antipsychotics; RR, rate ratio; CI, confidence interval.
a
Cases, incident cases of diabetes, hypertension or hyperlipidemia.
b
Per 1000 person-years.

95% CI 1.01–1.755). The effects seemed greater with patients et al., 1989; Thakore et al., 2002). Although the study design
taking SGA for developing DM (HR 1.82, 95% CI 1.30–2.55), could not control and adjust all these potential confounding
hypertension (HR 1.14, (95% CI 0.97–1.34), and hyperlipid- factors, we also found that the occurrence of DM is higher
emia (HR 1.41, 95% CI 1.09–1.83). The overall risks of among schizophrenic patients than those without schizo-
developing DM, hypertension and hyperlipidemia increased phrenia, from our insurance claims data. The schizophrenia
with age (p for trend b0.0001). There were strong relation- association is stronger for DM than for hypertension and
ships among these three disorders. hyperlipidemia. Antipsychotics are associated with the
increased risk for metabolic abnormalities.
4. Discussion This is the first nature course study using a population-
based retrospective cohort design to measure the risk of
Schizophrenia has been associated with metabolic abnor- metabolic abnormalities in schizophrenic patients between
malities such as DM. Previous prevalence studies estimated taking typical and atypical antipsychotics, using the general
the ratios of DM among those with schizophrenia to be 1.5 to population as controls. The major findings in our present
2-fold higher than in the general population (Bresee et al., study indicate that, compared with the non-schizophrenia
2010; Chien et al., 2009; Mukherjee et al., 1996). Factors controls, schizophrenic patients prescribed with SGA had a
associated with the increased incidence of diabetes among higher risk of developing DM, hypertension, and hyperlipid-
schizophrenic patients are obesity, genetic factors, lifestyle emia after adjusting for demographic factors and comorbid-
factors, and medication effects (Dixon et al., 2000; Mukherjee ities. We further analyzed whether schizophrenic patients
prescribed with FGA also had higher risks for these effects.
Table 3 Interestingly, schizophrenic patients taking FGA were older,
Age-specific Cox's proportional regression estimated hazard ratio for with 12.5% of patients aged 60 and above; whereas only 3.1%
diabetes, hypertension and hyperlipidemia in association with schizophrenia
taking first- and second-generation antipsychotics controlling for sex,
of patients prescribed with the SGA were aged 60 and above.
occupation, urbanization level and income. This is why the incidence rates of hypertension were higher
in the study set I cohorts than in the study set II cohorts.
Variables Diabetes Hypertension Hyperlipidemia
Additional analysis was focused on comparing the incidence
HR (95% CI) HR (95% CI) HR (95% CI) of developing DM, hypertension, and hyperlipidemia for
FGA patients taking specific SGA. Most patients (as high as 65.3%)
Age (years) were prescribed with risperidone (data not shown). How-
10–19 4.86 (0.64–37.1) 4.52 (1.76–11.6)** 2.15 (0.52–8.87) ever, quetiapine had higher risk than risperidone associated
20–39 2.28 (1.31–3.98)** 1.48 (1.16–1.88)** 1.34 (0.89–2.01) with developing DM (9.3 vs. 5.7 per 1000 person-years),
40–59 0.98 (0.66–1.47) 0.99 (0.81–1.20) 1.15 (0.85–1.57)
hypertension (23.6 vs. 17.4 per 1000 person-years) and
≥60 0.97 (0.52–1.81) 0.67 (0.52–0.86)** 0.73 (0.41–1.29)
hyperlipidemia (9.0 vs. 6.9 per 1000 person-years). Cloza-
SGA pine also had high risk in association with DM (6.8 per 1000
Age (years) person-years) and hyperlipidemia (14.2 per 1000 person-
10–19 – – 3.92 (1.83–8.39)** 3.02 (0.59–15.5)
years).
20–39 1.98 (1.23–3.17)** 1.63 (1.28–2.06)*** 2.38 (1.69–3.36)***
40–59 1.68 (0.99–2.85) 0.83 (0.63–1.09) 1.02 (0.65–1.60) Studies have also found a higher prevalence of hypertension
≥60 2.41 (0.76–7.64) 0.78 (0.48–1.26) 0.17 (0.02–1.27) in schizophrenic patients than in controls (Dixon et al., 1999;
HR, hazard ratio; CI, confidence interval; FGA, first-generation antipsychotics;
Nasrallah et al., 2006; Osborn et al., 2008). This study found a
SGA, second-generation antipsychotics. moderate risk of developing hypertension in schizophrenic
*p b 0.05, **p b 0.01, ***p b 0.0001. patients after adjusting for socioeconomic status and
114 C.-H. Liao et al. / Schizophrenia Research 126 (2011) 110–116

Table 4
Multivariable Cox proportional model measuring hazard ratios of developing diabetes mellitus, hypertension and hyperlipidemia in schizophrenic patients taking
first- and second-generation antipsychotics controlling for sex, occupation, urbanization and income.

Diabetes Hypertension Hyperlipidemia

HR (95% CI) HR (95% CI) HR (95% CI)

1st generation antipsychotics

Age, years
10–19 1.00 (Reference) 1.00 (Reference) 1.00 (Reference)
20–39 2.72 (0.97–7.65) 6.11 (3.78–9.88)*** 3.13 (1.57–6.26)**
40–59 6.81 (2.45–18.9)** 20.2 (12.6–32.5)*** 5.36 (2.69–10.7)***
≥60 6.30 (2.25–17.6)** 44.7 (27.9–71.8)*** 4.30 (2.13–8.69)***
p for trend b 0.0001 b0.0001 0.0001
Schizophrenia 1.32 (1.00–1.75)* 0.98 (0.86–1.11) 1.16 (0.93–1.45)
Diabetes – – 1.85 (1.57–2.17)*** 5.65 (4.62–6.90)***
Hypertension 2.88 (2.19–3.78)*** – – 3.90 (3.19–4.78)***
Hyperlipidemia 7.11 (5.58–9.05)*** 2.59 (2.27–2.96)*** – –

2nd generation antipsychotics

Age, years
10–19 1.00 (Reference) 1.00 (Reference) 1.00 (Reference)
20–39 16.9 (2.31–123)** 3.21 (2.16–4.77)*** 1.41 (1.09–1.83)**
40–59 26.9 (3.67–198)** 10.8 ( 7.30–16.1)*** 4.82 (2.09–11.1)**
≥60 28.3 (3.68–218)** 27.5 (18.1–41.8)*** 7.03 (3.01–16.4)***
p for trend b 0.0001 b0.0001 0.0001
Schizophrenia 1.82 (1.30–2.55)** 1.14 (0.97–1.34) 1.41 (1.09–1.83)**
Diabetes – – 1.90 (1.53–2.37)*** 5.29 (4.05–6.92)***
Hypertension 3.16 (2.26–4.43)*** – – 4.57 (3.59–5.83)***
Hyperlipidemia 6.90 (5.01–9.49)*** 3.14 (2.64–3.74)*** – –

*p b 0.05, **p b 0.01, ***p b 0.0001; HR, hazard ratio; CI, confidence interval.

comorbidities. Our definition of hypertension developed in the
follow-up period was based on the antihypertensive drug
therapy and treatment for hypertension in patients with
schizophrenia. The hypertension risk might be underestimated.
Schizophrenia-related hypertension has been associated with
body weight gain and metabolic syndrome because of the
sedentary life style in patients. (Goff et al., 2005; Nasrallah et al.,
2006; Susce et al., 2005).
Our study also showed a higher risk of developing
hyperlipidemia associated with schizophrenia. Obesity in
people with an unhealthy lifestyle with reduced physical
activity and poor diet selection is more common in schizo-
phrenic patients than in the general population, and this
could also be related with hyperlipidemia. Studies on both
adults and children have indicated the important role of
antipsychotics in the risk for obesity and DM in schizophrenic
patients (Hennekens et al., 2005; Newcomer, 2007; Scheen
and De Hert, 2007; Tschoner et al., 2007) The findings could
be underestimated because of inequalities in recognizing and
managing cholesterol in schizophrenic patients (Hippisley-
Cox et al., 2007).
Sex and age differences associated with schizophrenia and
the risk of DM, hypertension and hyperlipidemia deserve
attention. Previous studies have found women with schizo-
phrenia are at a higher risk for metabolic syndrome (Pilote et
al., 2007; Saari et al., 2005; Suvisaari et al., 2007). We also
found in the sex-specific analysis that the schizophrenia
cohort of SGA are at a greater risk than the non-schizophrenia
controls for these three disorders. The incident DM was even
greater for females than for males (6.29 vs. 4.32 per 1000
person-years) in the schizophrenia cohort taking SGA. On the
other hand, in the SGA cohort, males were at a slightly greater
risk than females in developing hypertension and hyperlip-
idemia. Since our findings were from an observational study
and limited by some unavailable data, further studies are
needed to examine the sex-specific issues in schizophrenic
patients.
Obesity and older age are highly related to DM, hyper-
tension and hyperlipidemia, (Flegal et al., 2002). However,
studies regarding schizophrenia with age-specific analysis
found that younger patients had higher incidences of
developing these metabolic comorbidities (Kato et al., 2005;
Suvisaari et al., 2007). Bresee et al. (2010) also found that the
adjusted risk of DM in schizophrenic patients was greater in
women than in men for all age groups, with the greatest risk
in women ages 20–29 (aOR 2.49; 95% CI: 1.78–3.49). Our age-
specific analysis also demonstrated, the hazards of develop-
ing DM and hypertension or hyperlipidemia in schizophrenic
patients taking either FGA or SGA were significantly elevated
in the 20–39 year old group, but not in the older age group.
The patients on FGA aged 10–19 had the highest risk of
developing hypertension among all age groups in this study.
Our findings reflect the greater vulnerability of younger
patients in developing DM, hypertension and hyperlipidemia.
Patients on FGA in the oldest group had a 33% less risk
compared with their controls. It is likely that the older
patients have already gone through the age of developing
these metabolic disorders and free of them at the baseline.
The elevated risk for younger patients and the protective
effect of age deserve further investigation.
The longitudinal design is the strength of the present
study, which aims to evaluate the relationship between
schizophrenia with antipsychotics and DM, hypertension, and
hyperlipidemia using population-based data. The technique
 C.-H. Liao et al. / Schizophrenia Research 126 (2011) 110–116
to define both exposure and outcomes by using medication
prescriptions, instead of only clinical diagnoses, is the other
strength. We only included patients who have received the
FGA or SGA therapy in the two schizophrenia cohorts. The
sample size and the power for measuring the risk of DM,
hypertension and hyperlipidemia were accordingly reduced.
However, we were still able to observe the effects of
schizophrenia associated with developing the three disor-
ders. Another limitation of this study is that we were unable
to measure the lifestyle association with the outcome because
the information was unavailable in the claims data. However,
we were able to use socio-demographic factors to control for
as major covariates in the risk estimation.
5. Conclusions
This population-based study demonstrates that patients
taking antipsychotics have an elevated risk of developing DM,
hypertension and hyperlipidemia. These three disorders may
occur among patients early in the life, indicating impact of
schizophrenia is greater for younger patients than for older
patients. In our study, the selected samples were newly
diagnosed and under medication with antipsychotics. Risk
estimations directed at the effects of medications in relation
to sex and age are of particularly important. Results derived
from this study are of potential value in planning compre-
hensive care plans for schizophrenic patients who are more
susceptible to developing hypertension and metabolic dis-
orders leading to cardiovascular disorders. In addition to
pharmacotherapy, providing patients with psychological
intervention and family interventions to establish a healthy
life style is necessary.
Role of funding source
This study was supported partly by the National Science Council,
Executive Yuan, Taiwan, Republic of China (NSC 97-2625-M-039-003),
China Medical University Hospital (1MS1, DMR-98-067 and DMR-98-091)
and Taiwan Department of Health Clinical Trial and Research Center
for Excellence (DOH99-TD-B-111-004) and Cancer Research Center of
Excellence (DOH99-TD-C-111-005).
Contributors
All authors participated in designing the study, drafting and revising the
manuscript, approving the submissions. Liao CH and Chang CS contribute equally.
Conflict of interest
The authors report no confliction of interest. The authors alone are
responsible for the content and writing of the paper.
Acknowledgement
The database was provided by the National Health Research Institute
(NHRI) of Taiwan. The interpretation and conclusions contained in this
article do not represent those of the NHRI.
Appendix A. Antipsychotics are available for patients in
the National Health Insurance program of Taiwan
First-generation antipsychotics
Haloperidol, flupentixol, sulpiride, chlorpromazine, tri-
fluoperazine, thioridazine HCl, chloroprothixene HCl,
thiothixene, clothiapine, loxapine, droperidol, and pimozide.
115
Second-generation antipsychotics
Risperidone, ziprasidone, clozapine, olanzapine, quetia-
pine, amisulpride, aripiprazole, and paliperidone.
References
Bobes, J., Arango, C., Aranda, P., Carmena, R., Garcia-Garcia, M., Rejas, J., 2007.
Cardiovascular and metabolic risk in outpatients with schizophrenia
treated with antipsychotics: results of the CLAMORS Study. Schizophr.
Res. 90, 162–173.
Bresee, L.C., Majumdar, S.R., Patten, S.B., Johnson, J.A., 2010. Prevalence of
cardiovascular risk factors and disease in people with schizophrenia: a
population-based study. Schizophr. Res. 117, 75–82.
Brown, S., 1997. Excess mortality of schizophrenia: a meta-analysis. Br. J.
Psychiatry 171, 502–508.
Cheng, T.M., 2009. Taiwan's National Health Insurance system: high value for
the dollar. In: Okma, K.G.H., Crivelli, L. (Eds.), Six Countries, Six Reform
Models: The Health Reform Experience of Israel, the Netherlands, New
Zealand, Singapore, Switzerland and Taiwan. World Scientific, New
Jersey, pp. 71–204.
Chien, I.C., Hsu, J.H., Lin, C.H., Bih, S.H., Chou, Y.J., Chou, P., 2009. Prevalence of
diabetes in patients with schizophrenia in Taiwan: a population-based
National Health Insurance study. Schizophr. Res. 111, 17–22.
Cohn, T., Prud_homme, D., Streiner, D., Kameh, H., Remington, G., 2004.
Characterizing coronary heart disease risk in chronic schizophrenia: high
prevalence of the metabolic syndrome. Can. J. Psychiatry 49, 753–760.
De Hert, M.A., van Winkel, R., Van Eyck, D., Hanssens, L., Wampers, M.,
Scheen, A., Peuskens, J., 2006. Prevalence of the metabolic syndrome in
patients with schizophrenia treated with antipsychotic medication.
Schizophr. Res. 83, 87–93.
De Hert, M., Schreurs, V., Sweers, K., Van Eyck, D., Hanssens, L., Sinko, S.,
Wampers, M., Scheen, A., Peuskens, J., van Winkel, R., 2008. Typical and
atypical antipsychotics differentially affect long-term incidence rates of
the metabolic syndrome in first-episode patients with schizophrenia: a
retrospective chart review. Schizophr. Res. 101, 295–303.
Dixon, L., Postrado, L., Delahanty, J., Fischer, P.J., Lehman, A., 1999. The
association of medical comorbidity in schizophrenia with poor physical
and mental health. J. Nerv. Ment. Dis. 187, 496–502.
Dixon, L., Weiden, P., Delahanty, J., Goldberg, R., Postrado, L., Lucksted, A.,
Lehman, A., 2000. Prevalence and correlates of diabetes in national
schizophrenia samples. Schizophr. Bull. 26, 903–912.
Flegal, K.M., Carroll, M.D., Ogden, C.L., Johnson, C.L., 2002. Prevalence and
trends in obesity among US adults: 1999–2000. JAMA 288, 1723–1727.
Goff, D.C., Cather, C., Evins, A.E., Henderson, D.C., Freudenreich, O., Copeland,
P.M., Bierer, M., Duckworth, K., Sacks, F.M., 2005. Medical morbidity and
mortality in schizophrenia: guidelines for psychiatrists. J. Clin. Psychiatry
66, 183–194.
Grundy, S.M., Brewer Jr., H.B., Cleeman, J.I., Smith Jr., S.C., Lenfant, C.,
American Heart Association; National Heart, Lung, and Blood Institute,
2004. Definition of metabolic syndrome: report of the National Heart,
Lung, and Blood Institute/American Heart Association conference on
scientific issues related to definition. Circulation 109, 433–438.
Hagg, S., Lindblom, Y., Mjorndal, T., Adolfsson, R., 2006. High prevalence of
the metabolic syndrome among a Swedish cohort of patients with
schizophrenia. Int. Clin. Psychopharmacol. 21, 93–98.
Hennekens, C.H., Hennekens, A.R., Hollar, D., Casey, D.E., 2005. Schizophrenia
and increased risks of cardiovascular disease. Am. Heart J. 150, 1115–1121.
Hippisley-Cox, J., Parker, C., Coupland, C., Vinogradova, Y., 2007. Inequalities
in the primary care of patients with coronary heart disease and serious
mental health problems: a cross sectional study. Heart 93, 1256–1262.
Huang, M.C., Lu, M.L., Tsai, C.J., Chen, P.Y., Chiu, C.C., Jian, D.L., Lin, K.M., Chen,
C.H., 2009. Prevalence of metabolic syndrome among patients with
schizophrenia or schizoaffective disorder in Taiwan. Acta Psychiatr.
Scand. 120, 274–280.
Kato, M.M., Currier, M.B., Villaverde, O., Gonzalez-Blanco, M., 2005. The
relation between body fat distribution and cardiovascular risk factors in
patients with schizophrenia: a cross-sectional pilot study. Prim. Care
Companion J. Clin Psychiatry 7, 115–118 quiz 119-120.
Laaksonen, D.E., Niskanen, L., Lakka, H.M., Lakka, T.A., Uusitupa, M., 2004.
Epidemiology and treatment of the metabolic syndrome. Ann. Med. 36,
332–346.
Lieberman, J.A., Stroup, T.S., McEvoy, J.P., Swartz, M.S., Rosenheck, R.A.,
Rosenheck, R.A., Perkins, D.O., Keefe, R.S., Davis, S.M., Davis, C.E.,
Lebowitz, B.D., Severe, J., Hsiao, J.K., Clinical Antipsychotic Trials of
Intervention Effectiveness (CATIE) Investigators, 2005. Effectiveness of
116 C.-H. Liao et al. / Schizophrenia Research 126 (2011) 110–116
antipsychotic drugs in patients with chronic schizophrenia. N. Engl. J.
Med. 353, 1209–1223.
Lin, H.C., Hsiao, F.H., Pfeiffer, S., Hwang, Y.T., Lee, H.C., 2008. An increased risk
of stroke among young schizophrenia patients. Schizophr. Res. 101,
234–241.
McEvoy, J.P., Meyer, J.M., Goff, D.C., Nasrallah, H.A., Davis, S.M., Sullivan, L.,
Meltzer, H.Y., Hsiao, J., Scott, Stroup T., Lieberman, J.A., 2005. Prevalence
of the metabolic syndrome in patients with schizophrenia: baseline
results from the Clinical Antipsychotic Trials of Intervention Effective-
ness (CATIE) schizophrenia trial and comparison with national estimates
from NHANES III. Schizophr. Res. 80, 19–32.
Mokdad, A.H., Ford, E.S., Bowman, B.A., Dietz, W.H., Vinicor, F., Bales, V.S.,
Marks, J.S., 2003. Prevalence of obesity, diabetes, and obesity-related
health risk factors, 2001. JAMA 289, 76–79.
Mukherjee, S., Schnur, D.B., Reddy, R., 1989. Family history of type 2 diabetes
in schizophrenic patients. Lancet 1, 495.
Mukherjee, S., Decina, P., Bocola, V., Saraceni, F., Scapicchio, P.L., 1996.
Diabetes mellitus in schizophrenic patients. Compr. Psychiatry 37,
68–73.
Nasrallah, H.A., Meyer, J.M., Goff, D.C., McEvoy, J.P., Davis, S.M., Stroup, T.S.,
Lieberman, J.A., 2006. Low rates of treatment for hypertension,
dyslipidemia and diabetes in schizophrenia: data from the CATIE
schizophrenia trial sample at baseline. Schizophr. Res. 86, 15–22.
Newcomer, J.W., 2007. Antipsychotic medications: metabolic and cardiovas-
cular risk. J. Clin. Psychiatry 68 (Suppl 4), 8–13.
Newman, S.C., Bland, R.C., 1991. Mortality in a cohort of patients with
schizophrenia: a record linkage study. Can. J. Psychiatry 36, 239–245.
Osborn, D.P., Wright, C.A., Levy, G., King, M.B., Deo, R., Nazareth, I., 2008.
Relative risk of diabetes, dyslipidaemia, hypertension and the metabolic
syndrome in people with severe mental illnesses: systematic review and
meta-analysis. BMC Psychiatry 8, 84.
Pilote, L., Dasgupta, K., Guru, V., Humphries, K.H., McGrath, J., Norris, C., Rabi,
D., Tremblay, J., Alamian, A., Barnett, T., Cox, J., Ghali, W.A., Grace, S.,
Hamet, P., Ho, T., Kirkland, S., Lambert, M., Libersan, D., O'Loughlin, J.,
Paradis, G., Petrovich, M., Tagalakis, V., 2007. A comprehensive view of
sex-specific issues related to cardiovascular disease. Can. Med. Assoc. J.
176, S1–S44.
Saari, K.M., Lindeman, S.M., Viilo, K.M., Isohanni, M.K., Järvelin, M.R., Laurén,
L.H., Savolainen, M.J., Koponen, H.J., 2005. A 4-fold risk of metabolic
syndrome in patients with schizophrenia: the Northern Finland 1966
Birth Cohort Study. J. Clin. Psychiatry 66, 559–563.
Scheen, A., De Hert, M., 2007. Abnormal glucose metabolism in patients
treated with antipsychotics. Diab. Metab. 33, 169–175.
Sikich, L., Frazier, J.A., McClellan, J., Findling, R.L., Vitiello, B., Ritz, L., Ambler,
D., Puglia, M., Maloney, A.E., Michael, E., De Jong, S., Slifka, K., Noyes, N.,
Hlastala, S., Pierson, L., McNamara, N.K., Delporto-Bedoya, D., Anderson,
R., Hamer, R.M., Lieberman, J.A., 2008. Double-blind comparison of first-
and second-generation antipsychotics in early-onset schizophrenia and
schizo-affective disorder: findings from the treatment of early-onset
schizophrenia spectrum disorders (TEOSS) study. Am. J. Psychiatry 165,
1420–1431.
Smith, M., Hopkins, D., Peveler, R.C., Holt, R.I., Woodward, M., Ismail, K., 2008.
First- v. second-generation antipsychotics and risk for diabetes in
schizophrenia: systematic review and meta-analysis. Br. J. Psychiatry
192, 406–411.
Susce, M.T., Villanueva, N., Diaz, F.J., de Leon, J., 2005. Obesity and associated
complications in patients with severe mental illnesses: a cross-sectional
survey. J. Clin. Psychiatry 66, 167–173.
Suvisaari, J.M., Saarni, S.I., Perälä, J., Suvisaari, J.V., Härkänen, T., Lönnqvist, J.,
Reunanen, A., 2007. Metabolic syndrome among persons with schizo-
phrenia and other psychotic disorders in a general population survey. J.
Clin. Psychiatry 68, 1045–1055.
Thakore, J.H., Mann, J.N., Vlahos, I., Martin, A., Reznek, R., 2002. Increased
visceral fat distribution in drug-naïve and drug free patients with
schizophrenia. Int. J. Obes. Relat. Metabol. Disord. 26, 137–141.
Tschoner, A., Engl, J., Laimer, M., Kaser, S., Rettenbacher, M., Fleischhacker,
W., Patsch, J., Ebenbichler, C., 2007. Metabolic side effects of antipsy-
chotic medication. Int. J. Clin. Pract. 61 (8), 1356–1370.
Warner, R., de Girolamo, G., 1995. Metabolic side effects of antipsychotic
medication. World Health Organization, Geneva, Switzerland.