Homocysteine level, body mass index and clinical correlates in Chinese Han patients with schizophrenia

Obesity is common comorbidity in patients with schizophrenia. Previous studies have reported that homocysteine (Hcy) is increased in schizophrenia. However, no study has reported the association between BMI and Hcy levels in schizophrenia. This cross-sectional naturalistic study aimed to evaluate the relationship between BMI, Hcy and clinical symptoms in Chinese Han patients with chronic schizophrenia. Clinical and anthropometric data as well as plasma Hcy level and glycolipid parameters were collected. Psychopathology was measured with the Positive and Negative Syndrome Scale (PANSS). Our results showed that compared with the low BMI group, the high BMI group had a higher PANSS general psychopathology subscore, higher levels of blood glucose, total cholesterol and high-density lipoprotein (HDL) cholesterol (all p < 0.05). Hcy levels were negatively associated with BMI in patients (p < 0.001). Hcy level, the PANSS general psychopathology subscale, total cholesterol and education (all p < 0.05) were the influencing factors of high BMI. Our study suggest that Hcy level may be associated with BMI in patients with schizophrenia. Moreover, patients with high BMI show more severe clinical symptoms and higher glucose and lipid levels.


Scientific RepoRtS
| (2020) 10:16119 | https://doi.org/10.1038/s41598-020-72934-3 www.nature.com/scientificreports/ Further research found that Hcy levels were positively correlated with the severity of negative symptoms 18 , duration of untreated psychosis 20 , and cognitive deficits 21 in patients with schizophrenia. Evidence for the association between different BMI and Hcy levels in the general population have been presented, with a particular focus on the course of physical disease 22,23 . However, results on the relationship between BMI and Hcy levels showed conflicting data 24 . A recent meta-analysis reported that Hcy was positively correlated with obesity 25 , while a recent study demonstrated opposite results in the general population 26 . A recent crosssectional study showed that Chinese bipolar disorder (BD) patients with hyperhomocystinemia had elevated BMI 16 . Taken together, it appears that Hcy may play a role in high BMI or obesity in both healthy population and patients with psychiatric disorders.
Based on the higher prevalence of obesity and increased Hcy levels in patients with schizophrenia and the close relationship between Hcy and obesity, it is of great interest to explore their possible association in schizophrenia. To our best knowledge, no study has reported the association between BMI and Hcy levels in patients with schizophrenia. Therefore, the main purposes of this study were to bridge this gap by examining: (1) whether there were differences in plasma Hcy, clinical data, glucose and lipid metabolism parameters between high and low BMI groups in patients with schizophrenia; (2) whether there were a correlation between BMI, Hcy and clinical symptoms in Chinese Han patients with schizophrenia.

Methods
Subjects. Using a cross-sectional naturalistic design, 120 patients with schizophrenia were recruited from the Affiliated Brain Hospital of Guangzhou Medical University, a psychiatric teaching hospital with more than 120 years of history in Guangzhou city. The study was conducted from June, 2018 to May, 2019. Inclusion criteria included the following: (1) aged 18-60 years, Han Chinese; (2) confirmed DSM-V diagnostic criteria of schizophrenia by two experienced psychiatrists according to the Structured Clinical Interviewed for DSM-V (SCID); (3) a stable dose of antipsychotic drugs for ≥ 4 weeks before the start of the study.
Each subject underwent a complete medical history, physical examination, and laboratory tests. Any subjects with major medical abnormalities were excluded, including acute or unstable medical illnesses or any organic brain diseases. Also, subjects suffering from drug or alcohol abuse/dependence were excluded. The trained research staff collected general information and social-demographic characteristics through questionnaires. Other information, including a history of mental illness and medications were collected from medical records. clinical measurements. The Positive and Negative Syndrome Scale (PANSS) was used to evaluate the psychiatric symptoms of all patients, which was assessed by two psychiatrists who had received special training to ensure consistency. The inter-rater correlation coefficient (ICC) of PANSS total score remained above 0.8 through repeated evaluation.
BMi measurements. BMI [weight (kg) / height squared (m 2 )] was calculated in a standardized fashion by assessing body weight and height. Height was measured when the subject was barefoot and upright. Weight was measured by a 0.1 kg electronic scale when the subjects wore light indoor clothing. Referring to the Chinese guidelines for adults 27 , it was considered as overweight with 24 ≤ BMI < 28 kg/m 2 , and BMI ≥ 28 kg/m 2 was classified as obese. In this study, all patients were divided into high BMI group (BMI ≥ 24 kg/m 2 ) and low BMI group (BMI < 24 kg/m 2 ).
Glycolipid metabolism and Hcy analysis. Venous blood was collected in the early morning from each patient after at least 8 h of fasting. The blood was collected with an EDTA tube and centrifuged (3000r/m) for 10 min within 30 min of collection. The plasma was isolated and stored at −80 °C for testing.
The Hcy levels were measured by Hcy assay kits (enzyme cycle assay), which were purchased from Beijing Leadman Biochemistry Co.Ltd. The assay method was in strict accordance with the instructions of the kit.
All biochemical analyses were performed by a technician in the clinical laboratory center of the hospital, who was blind to the status of subjects. Statistical analysis. Demographic and clinical characteristics between groups were compared using analysis of variance (ANOVA) for continuous variables and chi-squared tests for categorical variables. Since TG and Hcy levels did not follow a normal distribution (Kolmogorov-Smirnov one-sample test, p < 0.05), we first log transformed the data to a normal distribution and then used the transformed values in subsequent statistical tests. Further analysis of covariance (ANCOVA) was performed to control for the confounding factors including education, sex, age, age of onset and antipsychotic treatment (type, dose and duration of treatment). Associations between variables were examined by using Pearson correlation coefficients. We adopted Bonferroni correction to control for multiple tests. Multivariate regression analysis (enter model) was used to assess significant factors associated with BMI after controlling for variables. SPSS version 18.0 was used for all statistical analyses. All p values were defined with a 2-tailed significance level at 0.05.

Results
Demographics characteristics. A total of 120 patients including 76 males and 44 females were recruited.
Their average age was 46.92 ± 4.41 years and the mean education level was 11.05 ± 3.59 years. The average illness duration was 256.85 ± 155.42 months and the age of onset was 25.94 ± 9.16 years. Socio-demographic data and clinical profiles of patients with high BMI (n = 51) and low BMI (n = 69) groups are displayed in Table 1. There was no significant difference in demographic and clinical characteristics between the two groups (all p > 0.05), except for the education level and BMI (both p < 0.001; Bonferroni corrected both p < 0.01). clinical symptoms in high versus low BMi groups. The high BMI patients had higher PANSS general psychopathology subscore than the low BMI patients (p < 0.05; Bonferroni corrected p > 0.05). When education, sex, and age were added as covariates, there was still a significant difference in PANSS general psychopathology subscore (F 1,118 = 4.013, p = 0.049, r 2 = 0.008). There was no significant difference in PANSS total, positive symptom or negative symptom scores between the two groups (all p > 0.05).
In addition, since the number of patients treated with clozapine/olanzapine was imbalanced between the two groups (50.98%% vs. 60.87%), we performed ANCOVA with antipsychotic type as a covariate. The result still showed significant difference in PANSS general psychopathology subscore (F 1,118 = 4.331, p = 0.040, r 2 = 0.012) ( Table 1).
Homocysteine and glycolipid metabolism in high versus low BMi groups. As shown in Table 2, compared to the low BMI group, the high BMI group had significantly higher levels of glucose (F 1,118 = 5.812, Table 1. Socio-demographic data and clinical profiles of schizophrenia patients with high or low BMI. BMI body mass index, PANSS positive and negative symptoms scale, P positive symptom, N negative symptom, G general psychopathology syndrome.  Multiple regression analysis showed that Hcy levels (β = -3.789, t = -2.542, p = 0.012), the PANSS general psychopathology subscale (β = 0.103, t = 2.092, p = 0.039), total cholesterol (β = 0.979, t = 2.518, p = 0.013) and education (β = 0.293, t = 2.817, p = 0.006) were independent contributors to the BMI after controlling for age, gender, glucose, and HDL cholesterol and antipsychotic drugs (Table 3).

Discussion
To our best knowledge, this is the first study to examine the association between Hcy levels and BMI in patients with schizophrenia. The main findings of this study included: (1) The level of Hcy was significantly lower in the high BMI group than that in the low BMI group in patients with schizophrenia. (2) The high BMI patients had higher glucose and lipid levels than low BMI patients. (3) There were significant associations between BMI, Hcy, clinical symptoms and glycolipid metabolism in patients with chronic schizophrenia.
Our study showed that Hcy levels were significantly lower in the schizophrenia patients with high BMI, and that Hcy levels were negatively associated with BMI, which were consistent with some previous studies reporting a negative correlation between Hcy and BMI in the general population 28,29 . Also, a recent large sample of study including 5300 overweight/obese individuals and 5807 normal weight individuals showed that Hcy levels were significantly lower in overweight/obese persons than normal weight controls 26 . However, some previous studies have reported inconsistent results about the relationship between Hcy levels and obesity. For example, some studies showed that compared to non-obese subjects, the Hcy levels were significantly greater in obese subjects suffering from physical disease, such as coronary artery disease and polycystic ovaries [30][31][32] . The inconsistent results may be due to the differences in subjects that were from different diseases, regions, races, or genetic backgrounds. Other demographic and clinical factors such as diet habits, physical activity, drug treatment, chronic inflammation or the methods of measuring Hcy may also be implicated in the these inconsistent results. Interestingly, we found that Hcy was negatively associated with BMI in schizophrenia, which was inconsistent with most of previous studies showing that high Hcy may be related to obesity and increase the risk of coronary heart  www.nature.com/scientificreports/ disease 33,34 . This may be attributed to the effects of different diseases, different stages of disease development 18 , or medication treatment regiments 35 . Petronijević et al. 18 proposed that schizophrenia patients had increased plasma Hcy levels in the acute phase, but Hcy levels were significantly decreased after one month of antipsychotic treatment in the remission phase. Several drugs such as antiepileptic drugs 36 , fibrates 37 and antipsychotic drugs 35 may interfere with the one-carbon metabolic pathway, leading to changes in Hcy levels 36 . We speculated that the changes in Hcy levels observed in this study may be due to antipsychotic treatment and altered symptoms in patients with schizophrenia. Moreover, some researchers found that an increase in Hcy concentration was associated with an increase in the percentage of adipose tissue 38 , which regulated body fat through epigenetic inheritance of gene expression, leading to the occurrence and development of obesity 31,39 . Conversely, a previous study reported that Hcy had a negative association with both fat mass and lean mass during weight loss 40 . BMI, as a proxy for general obesity, cannot distinguish lean mass and adipose tissue, which may also confuse the relationship between obesity and HCY. Although the mechanism was still unclear, these results suggest that the Hcy metabolic pathway may play an important role in obesity 25 . In addition, we found that Hcy level was negatively correlated with blood glucose and HDL, which is consistent with previous reports 41 that hyperhomocysteinemia (HHcy) was associated with low levels of HDL. Also, Hcy was associated with glucose metabolism, lipid metabolism, hyperglycemia 42 , and hyperlipidemia 43 .
Another important finding of our study was that obese patients had more severe clinical symptoms, showing that the patients with high BMI had higher PANSS general psychopathology subscore than those with low BMI, and BMI was positively associated with PANSS general psychopathology subscore, which is in line with the finding of previous studies 2, 44 . Patients with schizophrenia who had greater severity of psychopathological symptoms may directly affect their enthusiasm for activity, leading to more significant obesity 45 . However, some previous studies have reported opposite results, showing that weight gain was associated with less psychopathological symptoms 46,47 , or that there was no correlation between PANSS and BMI in patients with schizophrenia treated with multiple antipsychotic drugs for one year 44 . Kemp et al. also found that some complex factors, such as medical adherence and treatment time may confuse the results 48 . In addition, several studies also showed that weight gain may be due to treatment of antipsychotic medications, including clozapine, olanzapine, and haloperidol [48][49][50] .
In order to rule out the effects of certain drugs, we included the antipsychotic treatment as covariates and found that there was still a significant correlation, suggesting that the relationship between high BMI and clinical psychopathological symptoms may be more than just a matter of antipsychotic treatment. Although the results were inconsistent in these studies, they all indicated that high BMI showed a certain correlation with clinical symptoms; however, more in-depth studies are needed to explore their causal relationship.
Several related factors for obesity were also identified in patients with schizophrenia in our study. We found that BMI was associated with glycolipid metabolism, showing that compared with low BMI patients, high BMI patients had higher blood glucose and lipid levels, especially total cholesterol, which is consistent with most of previous studies [51][52][53] . Moreover, a recent study also reported that BMI was positively associated with TGs in Chinese patients with schizophrenia 54 . Also, we found that high BMI was inversely correlated to education levels, which is consistent with previous studies 2, 55,56 . In addition, most of the previous studies demonstrated that atypical antipsychotic treatment may lead to increased obesity in schizophrenia 57,58 . In this study, however, we did not find that BMI was significantly associated with clozapine/olanzapine treatment. Clinically, psychiatrists were cautious in the use clozapine or olanzapine to high BMI patients, often switching to other antipsychotic drugs when a patient was becoming obese. Perhaps this is the main reason that we did not find significant association between obesity and clozapine/olanzapine treatment.
There are several methodological limitations of this study. First, a cross-sectional study design cannot directly show the causal relationship between obesity and Hcy in schizophrenia patients. Second, most of the subjects in this study were inpatients and some of them had been in the hospital for a long time. In general, they had more severe negative symptoms, a longer courses of disease, and a long history of antipsychotic treatment, which may have influenced their Hcy levels 35 . Therefore, our findings in this study should be confirmed in a large sample of patients with first-episode and drug-naïve schizophrenia. Third, there are some other confounding factors that may affect Hcy levels and obesity, such as physical exercise, diet, nutrition and psychotropic drugs. However, we did not collect these data in this study, which should be remedied in future studies. Fourth, previous studies found that there was sex difference in obesity in patients with schizophrenia 59 . However, in this study, we included much fewer female patients than male patients, which may have affected the results due to an imbalance in the sex ratio. Finally, our study did not have a healthy control group. Therefore, the results in this study should be considered preliminary.

conclusions
In summary, we demonstrated that high BMI was associated with lower Hcy levels, higher total cholesterol levels, and more severe psychopathological symptoms in patients with chronic schizophrenia. Moreover, Hcy was negatively associated with high BMI in these schizophrenia patients. Also, our results showed that high BMI patients had higher education levels but was not correlated with clozapine/olanzapine use. Future study in a large sample size with balanced male and female patients using a prospective and longitudinal design will be needed to explore the relationship between Hcy and obesity in first-episode and drug naïve patients with schizophrenia. compliance with ethical standards. All procedures were based on the Declaration of Helsinki issued by the National Institutes of Health. The research scheme has been approved by the Ethics Committee of the Affiliated Brain Hospital of Guangzhou Medical University before the study began. Written informed consents were obtained from each subject before participating in this study. www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.