Background

Schizophrenia is a severe mental illness that has a lifetime risk of about 1% [1]. The prognosis for schizophrenia is typically poor with a low recovery rate [2]. Schizophrenia imposes burdens on not only the affected individuals but also their families as well as society. Accordingly, it is a serious public health problem and social issue [3].

The main clinical manifestations of schizophrenia include positive symptoms, negative symptoms, and cognitive impairment [4, 5]. Atypical antipsychotics have good efficacy for these three symptom domains and produce fewer extrapyramidal reactions than typical antipsychotics. Yet, the ability of these drugs to cause weight gain, hyperglycaemia, and hyperlipidaemia in patients has garnered significant attention [6,7,8]. These side effects impact patient medication adherence and limit the treatment options available to physicians for patient management. Even so, not all patients who are administered atypical antipsychotics develop a glucose/lipid metabolic disorder. The reasons for this have yet to be clarified. It is important to explore and understand the risk factors and protective factors for schizophrenia treatment so that clinicians can make more informed and accurate treatment choices.

Apolipoprotein E (APOE) is the most abundant apolipoprotein [9, 10]. It has six genotypes (ε2/ε2, ε2/ε3, ε3/ε3, ε/ε4, ε3/ε4, ε4/ε4) originating from three different alleles (ε2, ε3, ε4) [11]. Some researchers have found that the APOE genotype is related to the manifestation of glucose/lipid metabolic abnormalities in patients without schizophrenia [12,13,14]. However, the relationships among APOE gene polymorphisms, blood sugar, and blood lipid content are not fully understood. This may be due to the age-dependent influence of APOE genotype on glucose/lipid metabolism [15, 16], as most previous studies do not account for the potential effects of age [13]. Most studies of patients with schizophrenia have been conducted with subjects who were mainly under the age of 60 years. These studies have explored the relationship between the APOE genotype and susceptibility to schizophrenia, but not that between the APOE genotype and metabolic syndrome [17, 18]. Elderly individuals are more likely to develop glucose/lipid metabolic abnormalities than their younger counterparts [19,20,21]. Thus, we conducted a study on Han Chinese elderly patients (60 years or older) with schizophrenia and long-term history of antipsychotics use. We evaluated the APOE genotype, blood sugar, blood lipids, and other related indicators to explore the influence of APOE polymorphisms on lipid and glucose metabolism in hospitalized patients 60 years or older with schizophrenia.

Methods

Study design and participants

The study was conducted at three mental health centres in Shanghai, China (the Shanghai Mental Health Center, the Mental Health Center of Jiading District in Shanghai, and the Mental Health Center of Fengxian District in Shanghai) between July 1, 2015 and December 31, 2015 (Fig. 1). Information regarding sex, age, height, weight, body mass index (BMI), and current prescribed medicines was recorded for all patients. Obesity was classified according to the health of the People’s Republic of China industry standard WS/T 428–2013, which defines overweight as a BMI ≥ 24.0. Schizophrenia was diagnosed by a senior psychiatrist according to the International Classification of Diseases 10 diagnostic standard. Diabetes was previously diagnosed by an endocrinologist according to the World Health Organization 1999 criteria [22]. High blood pressure was defined as high average measured blood pressure (systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg) or previous diagnosis by a clinical specialist.

Fig. 1
figure 1

Flowchart depicting the selection of elderly patients with schizophrenia

Full size image

Measurements of blood sugar and blood lipids

Research subjects fasted from 20:00 in the evening prior to blood sampling and were prohibited from drinking water for 4 h prior to blood sampling. A total of 5 mL of venous blood was collected from each patient in the fasting state between 06:30 and 08:30. Samples were incubated at room temperature for 30 min and then centrifuged at 3000 rpm for 15 min to collect the serum. All samples were tested for serum glucose, triglyceride content, cholesterol content, high-density lipoprotein, and low-density lipoprotein.

APOE genotyping by polymerase chain reaction (PCR)-ligase detection reaction (LDR)

A total of 5 mL of peripheral blood was slowly injected into an EDTA-coated anticoagulant tube, shaken, and then centrifuged at 3000 rpm for 20 min at 4 °C. DNA in 1 mL of blood was extracted with a whole blood genome DNA isolation kit (spin column, Tiangen Biochemical Science and Technology Co., Ltd., Beijing, China).

Multiplex PCR reactions on the basis of the two SNP cores of the APOE gene were conducted using the rs429358 and rs7412 design primers as follows: P1:5′-GCCTACAAATCGGAACTGGA CAGCTCCTCGGTGCTCTG-3′ and P2:5′-TAAGCGGCTCCTCCGCGATGCCCCGGCCTGGTACACTG-3′. The total reaction volume was 20 μL and included the following: 1 μL (50 ng) genomic DNA, 2 μL 1e reaction buffer, 0.6 μL 3 mM Mg2+, 2 μL of 2 mM of each dNTP, 0.2 μL 1 U Taq enzyme, 12.2 μL double-distilled H2O (ddH2O), and 2 μL 0.5p Primer Mix. PCR products were examined using 3.0% agarose gel electrophoresis.

Details of the multiplex LDR reaction can be found in Table 1. The reaction volume was 10 μL and included the following: 1 μL reaction buffer, 1 μL 2 pmol/μL Probe Mix (each), 0.05 μL 2 U Taq DNA ligase, 4 μL ddH2O, and 4 μL PCR product. Finally, APOE genotypes were assigned according to product fragment sizes on 3.0% agarose gel electrophoresis (Table 2). All genotyping was performed in duplicate.

Table 1 Reaction probes used in the multiplex ligase detection reaction
Full size table
Table 2 APOE genotype determination method
Full size table

Data analysis

All data were inputted using the EpiData3.1 software and analysed with the SPSS 17.0 software package (SPSS Inc., Chicago, IL). APOE genotypes were calculated by alignment inspection according to the Hardy–Weinberg law. Continuous data (age, chlorpromazine equivalent dose, triglycerides, cholesterol, high-density lipoprotein, low-density lipoprotein, and fasting plasma glucose) are represented with descriptive statistics as means ± standard deviation. Categorical data (sex, overweight status, diabetes, high blood pressure, atypical antipsychotics use, lipid-lowering medication use, and glucose-lowering medication use) are represented as percentages of the total population. APOE genotypes were divided according to three categories: APOE ε2 (ε2/ε2 and ε2/ε3), APOE ε3 (ε3/ε3), and APOE ε4 (ε3/ε4 and ε4/ε4). Patients with the ε2/ε4 genotype were excluded from analysis because of known opposite effects of ε2 and ε4 on lipid levels. Inpatients who were ε2 carriers or non-ε2 carriers were labelled as APOE ε2+ or APOE ε2-, respectively. Continuous data were analysed between two groups using independent sample t tests and among three groups using a one-way analysis of variance. Categorical data were analysed using chi-squared (χ 2) tests; however, atypical antipsychotics use and lipid-lowering medication use were analysed using Fisher’s exact test. Low-density lipoprotein was examined with pairwise comparisons using Hochberg’s GT2(H) method, while cholesterol was examined with pairwise comparisons using the Games-Howell(A) method. Finally, a regression analysis was performed on cholesterol, low-density lipoprotein, and potential influencing factors. The significance level was set at P < 0.05.

Results

Demographic characteristics and APOE genotype distribution

All patients were of Han Chinese descent. There were 160 men (53.3%) and 140 women (46.7%). The mean age was 67.3 ± 6.66 years (range, 60–92 years). Two of the participants had the ε2/ε2 APOE genotype, 38 participants were ε2/ε3, 6 were ε2/ε4, 205 participants (were ε3/ε3, 43 were ε3/ε4, and 6 were ε4/ε4; Fig. 2). According to the Hardy-Weinberg law, there was χ 2 = 243.79, df = 3, and P > 0.05. This indicates that the distribution of the genotype was in accordance with Hardy-Weinberg equilibrium. The allele frequencies were as follows: ε2 = 48 (8.0%), ε3 = 491 (81.8%), and ε4 = 61 (10.2%).

Fig. 2
figure 2

APOE genotype distribution in elderly patients with schizophrenia

Full size image

Comparison of clinical characteristics among the APOE genotype groups

Statistically significant differences were observed for lipid-lowering medication use and low-density lipoprotein according to the APOE genotype (P < 0.05). Near statistically significant differences were observed for cholesterol according to the APOE genotype (P = 0.052). There were no statistically significant differences in sex, age, high blood pressure, diabetes, glucose-lowering medication use, atypical antipsychotics use, plasma glucose, triglycerides, or high-density lipoprotein according to genotype (P > 0.05) (Table 3).

Table 3 Comparison of clinical characteristics according to APOE genotype in elderly patients with schizophrenia
Full size table

Comparisons of blood sugar, blood lipids, and other related indicators according to APOE genotype, ε2 allele carrier status, sex, and drug use

With lipid-lowering medication used as a covariate, cholesterol was significantly lower in elderly patients with schizophrenia with the APOE ε2 genotype than in those with the APOE ε3 and APOE ε4 genotypes (P < 0.05). No statistically significant differences were identified between the APOE ε3 and APOE ε4 genotypes (P > 0.05) (Fig. 3a). Low-density lipoprotein was also lower in patients with the APOE ε2 genotype than in those with the APOE ε3 and APOE ε4 genotypes (P < 0.01). No statistically significant differences were identified between the APOE ε3 and APOE ε4 genotypes (P > 0.05) (Fig. 3b). Cholesterol and low-density lipoprotein were lower in ε2 allele carriers with APOE ε2+ genotypes than in those with APOE ε2- genotypes (independent sample t tests, P < 0.05). No statistically significant differences were identified for triglycerides, plasma glucose, or high-density lipoprotein in ε2 allele carriers (P > 0.05) (Figs. 3c–g).

Fig. 3
figure 3

Comparisons of blood sugar, blood lipids, and other related indicators according to APOE genotype, (a-b), ε2 allele carrier status (c-g) in elderly patients with schizophrenia. Note: APOE ε2+ indicates carrier status for the ε2 allele and APOE ε2- indicates non-carrier status for the ε2 allele. * indicates P < 0.05, ** indicates P < 0.01, and NS indicates non-significant, P > 0.05

Full size image

Women had higher cholesterol levels (5.02 ± 1.10 vs. 4.47 ± 0.90 nmol/L, P < 0.01) and higher low-density lipoprotein levels (2.88 ± 0.81 vs. 2.66 ± 0.77 nmol/L, P < 0.05) than men. Compared to patients administered typical antipsychotics, those administered atypical antipsychotics had higher triglyceride levels (1.43 ± 0.86 vs. 1.09 ± 0.43 nmol/L, P < 0.01), higher cholesterol levels (4.78 ± 1.05 vs. 4.33 ± 0.85 nmol/L, P < 0.05), and higher low-density lipoprotein levels (2.80 ± 0.80 vs. 2.46 ± 0.72 nmol/L, P < 0.05).

Regression analysis of abnormal cholesterol and low-density lipoprotein in elderly patients with schizophrenia

Stepwise linear regressions were conducted with cholesterol, low density lipoprotein as the dependent variable and age, sex (0 = women, 1 = men), overweight status (0 = no, 1 = yes), APOE ε2 genotype (0 = no, 1 = yes), APOE ε3 genotype (0 = no, 1 = yes), APOE ε4 genotype (0 = no, 1 = yes), atypical antipsychotic drug use (0 = no, 1 = yes), diabetes (0 = no, 1 = yes), high blood pressure (0 = no, 1 = yes), and lipid-lowering medication use (0 = no, 1 = yes) as independent variables. The results showed that cholesterol and low-density lipoprotein were influenced by sex, APOE ε2 genotype, and atypical antipsychotics use (Table 4).

Table 4 Regression analysis of abnormal cholesterol and low-density lipoprotein in elderly patients with schizophrenia
Full size table

Discussion

Davignon et al. [23] found that APOE polymorphisms are important genetic factors influencing blood lipids, such as cholesterol. One study found that the presence of the APOE ε4 allele was significantly correlated with high serum cholesterol and low-density lipoprotein. ε2 and ε4 have opposite effects on lipid levels. ε4 allele carriers had the highest cholesterol and low-density lipoprotein values, followed by carriers of the ε3 allele, with ε2 allele carriers having the lowest values [24]. It has also been suggested that the presence of the ε2 allele reduces low-density lipoprotein [25] and that the ε4 allele increases low-density lipoprotein [26]. These findings are in agreement with our results in elderly patients with schizophrenia; serum cholesterol and low-density lipoprotein values were significantly lower in patients with the ε2 allele, while they were significantly higher in patients with the ε4 allele. Taken together, these data show that the effects of the APOE genotype on serum blood lipids are consistent between patients with schizophrenia and healthy subjects.

It is worth noting that the elderly patients with schizophrenia in this study had specifically been administered atypical antipsychotic drugs for years. In this context, ε2 carriers had more healthy cholesterol and low-density lipoprotein values than ε3 or ε4 carriers. These data suggest that ε2 carrier status is a protective factor against lipid metabolism abnormalities. This hypothesis was corroborated by a regression analysis. Therefore, we deduced that even after being administered atypical antipsychotics for a long time, elderly APOE ε2 carriers with schizophrenia may have a lower occurrence of lipid metabolism abnormalities than ε2 non-carriers. This is reflected in the fact that the ε2 carriers had the lowest rate of lipid-lowering drug consumption. Further investigations are required to explore whether the ε2 genotype can protect patients from metabolic problems during the initial years of antipsychotics treatment.

Studies have reported sex differences in the clinical features of schizophrenia [27]. Moreover, it has been reported that antipsychotic drugs exert sex-specific influences on blood lipids (i.e., larger effects on women). Consistent with this research, we found that female elderly patients with schizophrenia had higher serum lipid levels and were more likely to have lipid metabolism abnormalities than men.

At present, antipsychotic drugs are the first-line therapy for the treatment of schizophrenia. Among them, atypical antipsychotics are common but can cause weight gain as well as greatly increase a patient’s risk of developing diabetes and other metabolic syndromes [28, 29]. Moreover, the wide application of atypical antipsychotics has revealed that these drugs have a larger influence on blood lipids than typical antipsychotics [30,31,32]. In the present study, we found similar results and confirmed this effect of atypical antipsychotics use.

The correlation between the APOE genotype and triglycerides has been inconsistently reported in previous studies [23, 33]. A study with elderly Brazilian women found that triglyceride values were higher in ε2 carriers than in ε4 carriers [34]. In contrast, no obvious relationship between APOE status and serum triglycerides was reported in a study with patients with metabolic syndrome [35], which is in line with our findings. Moreover, the lack of correlation between high density lipoprotein and the APOE genotype that we noted in this study is consistent with other results [36].

In a previous study with a community sample of elderly Han Chinese individuals, we showed a potential association between the APOE ε3/ε3 genotype and increased susceptibility to diabetes [13]. However, this study showed that there were no statistical differences among the three genotype groups. We consider that these results require confirmation in a larger community sample, as the use of a hospitalized sample in the present study may have influenced our results.

A major factor limiting the generalizability of this study was our selection of hospitalized elderly patients with schizophrenia; specifically, the inclusion of long-term hospitalized patients with a relatively fixed diet and regular blood glucose monitoring. The influence of hospital environmental factors on glucose/lipid metabolism cannot be completely discounted, and accordingly our results should be validated in a community sample of elderly patients with schizophrenia. Moreover, we did not include patients who were abstinent from antipsychotic medication or non-elderly patients as control subjects. Therefore, this study does not inform us regarding the effects of specific antipsychotic drugs or those of age on glucose/lipid metabolism. In addition, the study results may be subject to survival bias. We thus could not assess the effects of the APOE genotype on the lifespan of elderly patients. Future studies should address these limitations.

Conclusion

Our study showed that the APOE ε2 allele and male sex are possible protective factors against blood lipid abnormalities in the context of atypical antipsychotic use in patients with schizophrenia 60 years or older. This finding informs the relationship between APOE genotype status and glucose/lipid metabolism in patients with schizophrenia of Han ethnicity (and possibly other Asian ethnicities), and can assist the development of a more personalized approach to schizophrenia treatment selection.