Glucose metabolism in systemic juvenile idiopathic arthritis

We conducted a cross-sectional study in patients with SJIA who were followed up in Pediatric Rheumatology Clinic at Ramathibodi Hospital, Bangkok, Thailand between 2019 and 2021. The diagnosis of SJIA required the presence of arthritis of unknown etiology that began before the age of 16 years and persisted for at least 6 weeks preceded by quotidian fever at least 2 weeks, plus 1 or more of the following findings: typical evanescent, non-fixed erythematous rash, hepatomegaly or splenomegaly, generalized lymphadenopathy and serositis [1]. Exclusion criteria were patients with acute conditions (trauma, severe infection, exacerbation of chronic disease) and patients who were diagnosed with type 1 and type 2 diabetes prior to the study enrollment. None of the SJIA patients in our Rheumatology Clinic has had documented diabetes. Oral glucose tolerance test (OGTT) was performed in all SJIA patients. The control group is a cohort of children with simple obesity who underwent the routine OGTT in our Pediatric Endocrine Clinic according to their clinical criteria. Patients and their parents provided written informed consents before participation. This study was approved by the Ethics Committee of the Faculty of Medicine Ramathibodi Hospital, Mahidol University (MURA 2020/599). Demographic characteristics were recorded for each patient at diagnosis of SJIA: age, sex, pubertal status, disease duration, previous and concomitant therapies. Juvenile Arthritis Disease Activity Score (JADAS-27) and Systemic Manifestation Score (SMS) were used for assessing disease activity. [11, 12]. JADAS-27 was derived from 4 variables obtained on the same day of the OGTT testing. These 4 variables comprise the total number of active joints (27 joints were evaluated), erythrocyte sedimentation rate (ESR) normalized to a 0–10 scale ([ESR mm/h-20]/10), the physician's global assessment score (0–10) and the parent's or patient's global assessment of overall well-being score (0–10). Minimum and maximum JADAS-27 scores are 0 and 57, respectively [11]. In this study, the cumulative JADAS-27 scores across the course of the patients’ disease were estimated by summing the scores obtained serially since the diagnosis until the enrollment. The systemic manifestation score (SMS) was created by considering extraarticular symptoms as follows: fever = 1 point for temperature (T) of 37–38 °C, 2 points for T 38–39 °C, 3 points for T 39–40 °C, 4 points for T > 40 °C; rash = 1 point; generalized lymphadenopathy = 1 point; hepatomegaly and/or splenomegaly = 1 point; serositis = 1 point; anemia (hemoglobin < 9 g/dl) = 1 point; platelet count >  × 10⁹/L or ferritin > 500 ng/mL = 1 point [12]. Patients were divided into 3 groups according to disease status: 1) patient with systemic features with or without arthritis, 2) patient with arthritis but no systemic features and 3) patient with inactive disease [13]. Pubertal maturation was assessed using the Tanner stage [14, 15]. Body mass index (BMI) was calculated using the formula: weight (kg)/height² (m²). Owing to the wide age range of the enrolled patients, for comparison, BMI standard deviation score (BMI SDS) also presented. BMI SDS was calculated from WHO standards [16, 17]. Overweight and obesity were defined as BMI SDS of 1–1.99 and 2 or greater, respectively [16].

Patients and controls underwent a standard protocol of 2-h OGTT with 1.75 g glucose/kg (max. 75 g) at the Pediatric Endocrine Clinic. After an 8-h fasting, venous blood samplings were obtained for measuring plasma glucose, serum insulin and lipid profiles. Plasma glucose and insulin levels were measured at 30, 60, 90 and 120 min after drinking glucose solution. The diagnoses of diabetes and prediabetes were based on American Diabetes Association (ADA) criteria. Normal fasting glucose was defined as fasting plasma glucose (FPG) < 100 mg/dL. Normal glucose tolerance was defined as 2-h post load glucose of < 140 mg/dL. Impaired fasting glucose (IFG) was defined as fasting plasma glucose (FPG) 100–125 mg/dL. Impaired glucose tolerance (IGT) was defined as 2-h post load glucose between 140 and 199 mg/dL. Diabetes was defined as either FPG 126 mg/dL or greater or 2-h post load glucose 200 mg/dL or greater. Prediabetes was defined as either IFG or IGT or both [18].

Diagnosis of MS was based on International Diabetes Federation (IDF) criteria which included visceral fat obesity (waist circumference > 90th percentile) plus any two of the other four factors: triglyceride (TG) > 150 mg/dL, high density lipoprotein-cholesterol (HDL-C) < 40 mg/dL, blood pressure > 95th percentile, and either IFG or IGT [21‐23]. Due to lack of Thai National Standard for waist circumference, waist circumference percentile according to the previously published reference of Chinese children was used [24].

A total of 103 children with SJIA in Rheumatology Clinic database were eligible. Fifteen were excluded due to insufficient data (n = 14) and death (n = 1), while 49 refused to participate in the study. Most of these 49 patients were followed by telemedicine-based care owing to the Covid-19 pandemic. It was not possible for many patients to come to the hospital for performing the OGTT. Therefore, a total of 39 patients were enrolled in this study. Patient characteristics are shown in Table1. Their mean age (SD) was 12.0 (4.6) years. Twenty-one (54%) were male. Twenty-nine (74%) patients have entered puberty, their median Tanner stage was III. The median age (IQR) at diagnosis was 5.8 (3.7–9.3) years. Thirty-one out of 39 patients (80%) had inactive disease (status 3) and 7 (18%) had only arthritis (status 2). The remaining 1 patient had systemic features and arthritis (status 1). Median disease duration (IQR) was 65 (45–117) months. Median BMI SDS (IQR) was 0.1 (-0.5 to 1.7). Four patients (10%) were overweight (OW) and seven (18%) were obese (OB). Median (IQR) cumulative dose of prednisolone was 272 mg/kg (113–661). We reviewed baseline characteristics of 49 SJIA patients who refused to participate in the study due to inconvenience. Comparing between these 49 patients and the 39 patients who were enrolled, there were no significant differences in age, sex, BMI, BMI SDS, percentage of OW/OB, prednisolone use and disease status.

Only 1 obese patient had prediabetes and none had diabetes. Three patients had hypertension and fulfilled criteria of metabolic syndrome. Twenty-two patients (56%) had dyslipidemia including decreased HDL-C (n = 7, 18%), elevated TG (n = 12, 31%), elevated low density lipoprotein-cholesterol (LDL-C) (n = 12, 31%) and elevated both TG and LDL-C (n = 5, 13%).

The patient with prediabetes was a 15-year old Thai male with severe obesity (BMI 39.2 kg/m², BMI SDS 3.6). He was diagnosed with SJIA at the age of 11 years with macrophage activation syndrome, obstructive sleep apnea and hypertriglyceridemia (TG = 259 mg/dL). His disease was in remission at age 14 years. He had discontinued all medications for one year prior to the study. He also had hypertriglyceridemia (TG = 129 mg/dL) and reduced HDL-C (34 mg/dL). He had no family history of diabetes.

Patients with SJIA were divided into 2 groups: lean (n = 28) and OW/OB (n = 11). Sex, age, Tanner stage, disease duration, systemic score, JADAS-27, disease status, cumulative dose of prednisolone, MTX, NSAIDs, biologic use and treatment duration were not significantly different between the 2 groups. In comparison with the lean group, current prednisolone use and cumulative prednisolone dose tended to be higher in the OW/OB group but were not significantly different. Three of 11 (27%) OW/OB patients had metabolic syndrome, and 1 of 11 (9%) had impaired glucose tolerance. In comparison with lean patients, OW/OB had a significantly higher fasting plasma insulin level and HOMA-IR but significantly lower WBISI, while no difference in fasting plasma glucose level (Table 1).

Due to the fact that obesity affects glucose metabolism, OW/OB patients (n = 11) were compared with sex- puberty- and age-matched OW/OB controls (n = 33). Of 11 OW/OB patients, 10 had disease status 3 (inactive) and only 1 had active disease (status 1). Six patients were currently treated with low dose prednisolone. The controls were recruited from healthy OW/OB children who had OGTT performed in Pediatric Endocrinology Clinic. In comparison with the OW/OB controls, OW/OB patients with SJIA (n = 11) had higher insulin resistance index, as shown by median (IQR) homeostasis model assessment for insulin resistance (HOMA-IR) [2.6 (2.1–3.3) vs 1.5 (0.8–2.0), p = 0.001], and lower insulin sensitivity index as shown by median (IQR) whole-body insulin sensitivity index (WBISI) [3.7 (2.7–5.9) vs 5.4 (4.5–8.7), p = 0.024]. The former had compensatory increased β-cell function indices as compared to the latter [median (IQR) insulinogenic index (IGI), 2.5 (2.0–3.5) vs 1.0 (0.8–1.9), p < 0.005; median (IQR) AUC-I/AUC-G, 0.7 (0.5–1.0) vs 0.4 (0.3–0.6), p = 0.019] (Table 2 and Fig. 1).

Since glucocorticoid treatment affects glucose metabolism, OW/OB patients currently treated with low dose prednisolone (2.5–10 mg/day) (n = 6) were compared with those without prednisolone treatment (n = 5). There were no significant differences in insulin sensitivity indices and β-cell function indices (data not shown). In addition, concerning cumulative prednisolone dose, comparing between patients who received the cumulative prednisolone doses of < 250 mg/kg (n = 5) and > 250 mg/kg (n = 6), there were no differences in all insulin sensitivity and β-cell function indices (data not shown).

Due to unavailable to recruit healthy lean controls to have OGTT, we compared our lean patients (n = 28) with the previously published normal weight children (n = 142) [25]. Fasting plasma glucose, insulin resistance index (HOMA-IR) and fasting plasma insulin in the patients seem to be comparable to those of lean controls (Table 3).

Since chronic inflammation causes insulin resistance, lean patients with SJIA (n = 28) in disease status 2 (arthritis only) (n = 7) were compared with those in status 3 (inactive disease) (n = 21). There were no significant differences in insulin sensitivity and insulin secretion indices (data not shown). Concerning prednisolone effect on these indices in the 28 lean SJIA patients, 7 (25%) were treated with low dose prednisolone at the time of the study enrollment. Comparing between these 7 patients and 21 patients who had no prednisolone treatment at the time of study enrollment, there were no differences in all insulin sensitivity and β-cell function indices.