The main finding of the present study was that patients with SZ with abdominal obesity correlated with FA reductions in global regions and to cortical thinning in the temporal lobe and insula cortex, but not with reduction in subcortical volumes. Our results provide evidence of microstructural abnormalities of global WM integrity and focal cortical GM thickness in patients with SZ, which is seen to be in a more severe pattern in those with excessive abdominal obesity.
Our analysis of the whole brain found that the WM integrity loss and cortical GM thinning in patients with SZ was more pronounced in those with excessive abdominal obesity, while the patients without it demonstrated limited abnormalities when compared with healthy control subjects. To discuss the mechanism behind the influence of abdominal obesity on brain imaging of microstructural alternations, some previous neurobiological findings in the brain structural abnormality with obesity could bridge between microscopic and macroscopic levels. A genetic research project reported the particular genes and pathways which affect obesity and the roles of the central nervous system, such as synaptic plasticity and glutamate receptor activity 49. Meanwhile, in animal research in obese rats microglial alterations and decreased synaptic markers were found 2. Elsewhere, post-mortem research in SZ reported metabolic alterations in front-striatal-thalamic WM tracts 50. DTI research in two large independent cohorts of healthy subjects suggested that elevated obesity is associated with reduced global WM abnormalities 51. Various neurobiological research has advocated that obesity and SZ share several disrupted biological pathways, such as increased oxidative stress, hypothalamic-pituitary-adrenal disturbances, neurotransmitter imbalances, neurodegeneration, and dysregulated inflammatory pathways 52, and that chronic inflammation, particularly that associated with obesity, could influence brain plasticity 53. A systematic review with meta-analytic evidence on brain structural abnormalities reported that the complications of abdominal obesity and drug-naive first episode psychoses may increase the severity of brain WM or GM volumes 4, and the shared pathway in obesity and SZ could represent alternated WM integrity 5. Our results therefore corroborated with these previous reports, reinforcing the evidence of microstructural abnormalities of WM and cortical GM microstructures in patients with SZ, which is more prominent in those with excessive abdominal obesity.
The current study explored simple metabolic risk biomarkers other than WC, which partly mediated the relation of WC and the brain. As shown in Table 1, patients in the SZ-HWC group had significantly higher TG scores than patients in the SZ-LWC group. The combination of an increased WC and hypertriglyceridemia known as ‘hypertriglyceridemic waist’ has been proposed as a useful and inexpensive screening tool for identification of people at increased risk of coronary artery disease and type 2 diabetes 54,55. WC and TG levels have been reported to be significantly associated with the 10-year risk of coronary heart disease events 56. Such vascular risk factors may influence not only cerebrovascular pathology, but also neurodegenerative processes 9, and WM integrity loss in cortico-subcortical pathways and association fibers of fronto-temporal lobes may also relate to markers of microangiopathy 57. Negative correlation between brain microstructure and TG levels was shown in healthy subjects for the first time in 2017, with the indication that WM may be more sensitive to fat/lipid factors than GM 58. These findings might be explained by higher lipid content in myelinated WM and cerebrovascular architectural differences 59,60, differences in cerebrovascular architecture between cortical WM and GM where perfusion rate in cerebral WM is lower 61, and by sensitivity of WM to systemic inflammation and obesity-associated metabolic disorders 59,60,62,63. Our results corroborated with these previous reports, extending the evidence of global WM microstructural abnormalities and focal GM cortical thickness in patients with SZ, which is more prominent in those with excessive abdominal obesity and hypertriglyceridemia.
The progression in patients with SZ with excessive abdominal obesity might influence the brain microstructural abnormalities and memory decline more extremely than the patients without SZ, and it might accentuate their relationship. This interpretation may be partly supported by another interesting finding of our current study (Table 1), in which the STM and LTM performances in patients with SZ were worse in those with higher WC, who had more severe TG elevation, although this is a cross-sectional study. Additionally, the patients with SZ with abdominal obesity correlated to cortical thinning in the temporal lobe and insula cortex. These results are supported by several past findings. Animal research in obese rats found the presence of activated microglia and suggested neuroinflammation increases in the temporal association cortices, which regions impact memory, time perception, and auditory perception 64. Functional imaging research in late-middle-aged adults reported abdominal obesity and hypertriglyceridemia correlate with low cerebral blood flow across a large portion of the cortical surface, which is associated with lower memory function 65. In patients with SZ, the hypertriglyceridemia was negatively correlated with cognitive function 66. A review of metabolic studies of patients with SZ explored the relative effects of the metabolic syndrome components on cognition in SZ; the most commonly significant predictor was elevated TG level, followed by abdominal obesity 67.
The patients with SZ were taking a variety of pharmacological agents not only at the time of scanning, but prior to the scanning (i.e., lifetime use of medication). The influence of medication should be interpreted cautiously. Previous studies in SZ have shown associations between alterations in the FA values and antipsychotics 68 and a differential effect of first-generation antipsychotics (FGAs) and second-generation antipsychotics (SGAs) on WM 69. Further, FGAs and SGAs have divergent effects on cortical GM thickness during the first episode of psychosis that are independent from changes due to illness 70, and the difference between FGAs and SGAs could be attributed to smaller subcortical GM volumes in the clozapine users specifically 71. Furthermore, the low-potency FGAs and SGAs (clozapine, olanzapine, and quetiapine) are associated with a higher risk of hyperlipidemia, which increases noted primarily in TG levels 72–76. Metabolic disturbances appear to be greatest with clozapine and olanzapine, intermediate with quetiapine and low-potency FGAs, and lowest with aripiprazole, risperidone, ziprasidone, and high-potency FGAs 72. Clozapine is one of the most effective antipsychotic and is used for treatment-resistant schizophrenia 77. These previous reports might be partly supported by the finding of our current study (Table 1), in which the SZ-HWC group had a trend toward higher daily dosage of SGAs than the SZ-LWC group and several patients in the SZ-HWC group were taking clozapine while none of the patients in the SZ-LWC group were taking clozapine.
Some limitations of this study must be noted. First, the current study did not investigate the clinical parameters regarding lifestyle changes such as inadequate exercise or eating habits which could contribute to metabolic risk factors 78. Excessive supply of energy from food correlated with TG levels and high-density lipoprotein (HDL) cholesterol 79. And then, all subjects did not receive fasting blood tests, which may have affected the results such as especially blood glucose levels. Second, the fasting HDL was not measured in this research, and it is one of the key elements of diagnostic criteria for metabolic syndrome. Third, the patients with SZ were divided on high-risk criteria 24–27, but the HC group could not be divided into two subgroups because of small sample size. This was not possible because it was a retrospective subgroup analysis of our previous cohort 28,29. To confirm the results of our study, future studies are necessary using patients with limited medication exposure and matched lifetime use of medication, with subjects who matched for physical exercise levels or eating habits, with laboratory tests or other measures (e.g., HDL) of cardio metabolic health that were not included in the present study, and using a greater number of classifiable HCs.
Nonetheless, this is the first known study demonstrating the relationship between excessive abdominal obesity and microstructural alterations of WM-GM tissue in patients with SZ. Our findings contribute to the understanding of potential neurobiological overlaps between abdominal obesity and a disease-specific pathological process in SZ. Moreover, our findings might have potential implications for public health and managements of patients with SZ using simple and inexpensive measurements (WC and TG), and the evidence supports the findings of our previous studies of obesity-reducing aerobic exercise 28,29,80,81, which may improve brain network and cognitive function.