Baseline characteristics
In the present case-control study, 56 NMOSD patients with seropositive AQP4-IgG and 100 controls without family history of MS or NMOSD were involved. Participants were recruited between December 2019 and Feb 2020.
The baseline and clinical characteristics of study groups are summarized in Table 1. NMOSD patients did not use corticosteroid at least 3 months before sampling. Among NMOSD patients, 14.28% (n = 8) had at least a history of a chronic disorder; vitiligo (1 out of 56), SLE (1 out of 56), asthma (3 out of 56), type 1 diabetes (1 out of 56), arthritis (1 out of 56) and Crohn's disease (1 out of 56).
In our population study, there were no significant differences in age, gender, and body mass index (BMI) between the two groups (P > 0.05).
Serum levels of insulin were higher in NMOSD patients while serum levels of FBS were lower in NMOSD patients than the control group even after adjusting with age, sex and BMI (p = 0.0001, p = 0.002; respectively).
The data analysis showed no significant differences in lipid profile levels between the two groups (p > 0.05).
Table 1
Clinical and biochemical characteristics of the NMOSD and control groups under study
Baseline characteristics
|
NMOSD
N = 56
|
Control
N = 100
|
p-value
|
Age (years)
|
35.89 ± 9.39
|
37.40 ± 6.68
|
0.24
|
Sex (Female)
|
89.3% (50)
|
93.0% (93)
|
0.42
|
BMI (kg/m2)
|
26.61 ± 5.15
|
26.82 ± 5.45
|
0.81
|
History of alcohol consumption (whisky and/or vodka)
|
3.57% (2)
|
1% (1)
|
0.26
|
History of smoking
|
5.35% (3)
|
6% (6)
|
0.86
|
Physical activity (Ln MET-min/week)
|
7.36 + 2.09
|
7.60 + 1.07
|
0.38
|
Duration of the disease (years)
|
4.16 ± 2.91
|
-
|
-
|
Age of onset (years)
|
32.28 ± 9.11
|
-
|
-
|
Medications
Azathioprine
Rituximab
|
23.2% (13)
76.8% (43)
|
-
|
|
Comorbidity
|
14.28% (8)
|
-
|
|
EDSS score*
|
3 (3.5)
|
|
|
Biochemical and laboratory analysis
|
|
|
|
FBS (mg/dL)
|
85.47 ± 6.93
|
90.42 ± 9.03
|
0.003
|
Ln.HDL (mg/dL)
|
3.89 ± 0.23
|
3.93 ± 0.20
|
0.29
|
LDL (mg/dL)
|
100.71 ± 71
|
105.97 ± 97
|
0.27
|
Ln.TG (mg/dL)
|
4.71 ± 0.46
|
4.57 ± 0.48
|
0.07
|
Chol (mg/dL)
|
171.75 ± 32.62
|
177.77 ± 34.67
|
0.29
|
Ln.Insulin (µU/L)
|
2.78 ± 0.88
|
2.12 ± 0.48
|
0.0001
|
Ln. SAT (U/L)
|
3.30 ± 0.28
|
3.24 ± 0.31
|
0.15
|
Ln.ALT (U/L)
|
2.94 ± 0.31
|
2.81 ± 0.45
|
0.038
|
Ln.IL-17 (pg/ml)
|
2.00 ± 1.18
|
1.11 ± 0.81
|
0.0001
|
Ln.IL-6 (pg/ml)
|
3.76 ± 0.74
|
3.34 ± 0.62
|
0.008
|
*median (IQR)
Numerical variables were expressed as the mean ± SD except for EDSS (median (IQR)). Categorical variables were presented as percentages. Values in bold indicate statistical significance (P < 0.05).
ALT, alanine aminotransferase; SAT, serum agglutination test, BMI, body mass index; EDSS, expanded disability status scale; FBS, fast blood sugar; Chol, total cholesterol; HDL, high-density lipoprotein cholesterol; IL-6, interlukine-6; IL-17, interlukine-17; LDL, low-density lipoprotein; Ln, natural log; MET, metabolic equivalent of task; TG, triglyceride.
Metabolic Syndrome Components:
The prevalence rates of metabolic syndrome components are presented in Table 2. Hyperinsulinemia was approximately 1.5 times higher in NMOSD patients than the control group (in women participants, 50% in NMOSD group and 26.9% in the healthy group).
Among other metabolic syndrome components, the data analysis showed no significant differences between the two groups.
In women participants, the prevalence rates of hyperglycemia, dyslipidemia, and hypertension were statistically similar to the control group (p > 0.05). There was no separate data analysis for the male subgroup because of its small size.
In the control group, the data analysis showed there was direct correlation between serum levels of insulin and FBS (p = 0.02, r = 0.23). However, in NMOSD group, there was not any significant correlation between serum levels of insulin and FBS (p = 0.1).
Table 2
The prevalence of metabolic syndrome components based on modified WHO criteria in NMOSD and control groups
Metabolic syndrome component
|
NMOSD
N = 56
|
Control
N = 100
|
p-value
|
Hypertension (≥ 140/90 mmHg)
|
5.4% (3)
|
2%(2/100)
|
0.35*
|
Dyslipidemia (triglycerides ≥ 150 mg/dl or HDL cholesterol < 40 mg/dl)
|
37.7% (21)
|
26.0% (26)
|
0.13
|
Hyperglycemia (Fasting glucose ≥ 110 mg/dl)
|
0.0% (0)
|
4%(4)
|
0.29*
|
Hyperinsulinemia (forth Quartile)
|
48.2% (27)
|
26.0% (26)
|
0.005
|
waist girth ≥ 94 cm
|
35.0%(20)
|
16.0% (16)
|
0.063
|
ISQUICKI
|
0.32 ± 0.04
|
0.35 ± 0.02
|
0.0001
|
*Fisher exact test
Insulin Sensitivity In Nmosd Patients:
Insulin sensitivity (Quicki index) was calculated based on fasting serum levels of insulin and FBS. There was significant lower IS in patients with NMOSD compared with the control group (0.32 ± 0.04 vs. 0.35 ± 0.02, p = 0.0001). In logistic regression model, after adjusting for age, sex and BMI, there was significant association between lower IS and NMOSD risk (95% CI: Beta = 0.73, 0.62 to 0.86, p = 0.0001).
In NMOSD patients, there were not any significant correlation between IS and duration of disease (p = 0.08) and EDSS score (p = 0.27). Also, there was not any significant difference in serum levels of insulin or insulin sensitivity in NMOSD patients with different type of medication; rituximab and azathioprine (p = 0.91, p = 0.40; respectively).
Inflammatory Markers:
The data showed that serum levels of IL-17 were significantly higher in NMOSD patients than the control group (mean Ln-IL17 ± SD: 2.00 ± 1.18 pg/ml vs. 1.11 ± 0.81 pg/ml, p = 0.0001, respectively) even after adjusting age, sex and BMI (95% CI, beta: 2.35, 1.63 to 3.37, p = 0.0001) (Fig. 1a).
The serum levels of IL-6was significantly higher in NMOSD patients (mean Ln-IL6 ± SD: 3.76 ± 0.74 pg/ml vs. 3.34 ± 0.62 pg/ml, p = 0.008, respectively) (Fig. 1b) even after adjusting with age, sex and BMI (95% CI, beta = 2.39, 1.23 to 4.67, p = 0.01).
In NMOSD group, there was only significant direct correlation between IL-6 and TG (P = 0.025; r: 0.341). In the control group, there was only significant correlation between IL-6 and IL-17 with TG (P = 0.04; r: 0.322, P = 0.001; r: 0.338, respectively).
There were not any correlations between IL-6 and IL-17 with liver enzymes in each group (Table 3).
There was not any significant correlation between inflammatory markers and duration of diseases in NMOSD patients (P > 0.12) and EDSS score (p > 0.34).
Regarding to different type of medications, rituximab and azathioprine, there was not any significant differences in serum levels of IL-6, IL-17 and insulin in NMOSD patients (p = 0.81, p = 0.91, p = 0.44); respectively).
Also, there were not any significant correlation between IL-6, IL-17 with disease statues (duration of disease and EDSS) (p > 0.12).
Table 3
Correlation between biochemical and inflammatory markers in NMOSD patients and control group
|
NMOSD
|
control
|
Ln. IL-6 (pg/ml)
|
Ln.IL-17(pg/ml)
|
Ln. IL-6 (pg/ml)
|
Ln.IL-17(pg/ml)
|
r
|
P-value
|
R
|
P-value
|
r
|
P-value
|
r
|
P-value
|
Ln.insulin (µU/L)
|
0.32
|
0.03*
|
0.19
|
0.15
|
-0.05
|
0.75
|
-0.02
|
0.81
|
FBS (mg/dL)
|
0.11
|
0.54
|
0.003
|
0.98
|
0.07
|
0.66
|
0.16
|
0.10
|
Ln.TG (mg/dL)
|
0.34
|
0.02*
|
0.16
|
0.23
|
0.32
|
0.04*
|
0.34
|
0.001*
|
Ln.HDL (mg/dL)
|
-0.27
|
0.08
|
-0.02
|
0.88
|
-0.03
|
0.83
|
-0.06
|
0.56
|
LDL (mg/dL)
|
0.05
|
0.74
|
0.02
|
0.86
|
0.11
|
0.51
|
-0.008
|
0.93
|
Chol. (mg/dL)
|
0.003
|
0.98
|
0.03
|
0.84
|
0.16
|
0.34
|
0.10
|
0.32
|
Ln.AST (U/L)
|
0.12
|
0.44
|
0.19
|
0.18
|
-0.02
|
0.89
|
0.11
|
0.26
|
Ln.ALT (U/L)
|
-0.08
|
0.62
|
-0.03
|
0.81
|
0.06
|
0.73
|
-0.11
|
0.28
|
* After adjusting with BMI, age and sex, there was constant significant correlation.
ALT, alanine aminotransferase; SAT, serum agglutination test, Chol, total cholesterol; HDL, high-density lipoprotein cholesterol; IL-6, interlukine-6; IL-17, interlukine-17; LDL, low-density lipoprotein; Ln, natural log; TG, triglyceride.
Correlation between circulating levels of inflammatory markers and insulin sensitivity
In unadjusted data analysis, there were significant correlations between insulin levels and IL-6 (p = 0.037, r = 0.31) but not with IL17 (p = 0.15) in NMOSD patients.
In the control group, there was not any significant correlation between insulin levels and inflammatory markers, IL-6 (p = 0.75) and IL17 (p = 0.81).
In logistic regression model, after adjusting with age, sex, BMI, taking medicines and IL-6, there was only an independent direct association between IL-6 and NMOSD risk (95% CI: beta = 2.29, 1.06 to 4.96, p = 0.035). There was not any constant association between IS and NMOSD risk (p = 0.08).
To examine the effect modifier role of IL-17, logistic regression model was used. After adjusting with age, sex, BMI, taking medicines and IL-7, there was independent direct association between IL-17 (95% CI: Beta = 2.15, 1.42–3.26, p = 0.0001) and lower IS (p = 95% CI: beta = 0.8, 0.68 to 0.93, p = 0.004) with NMOSD risk.