The clinical features and outcome of Kawasaki disease combined with G6PD deficiency Authors: Peisi Cheng, Tingyan He, Jun Yang Published: 2025 ABSTRACT AbstractPurposeThe characteristics and prognosis of Kawasaki disease (KD) patients with glucose-6-phosphate dehydrogenase deficiency (G6PDd) remain unknown. This study primarily aims to investigate whether G6PDd is an independent risk factor for persistent coronary artery lesions (CAL) in KD and secondarily to evaluate the prognosis of these patients without aspirin therapy.MethodsWe conducted a retrospective case-control study. All KD patients with G6PDd were included. The control group, consisting of KD patients with normal G6PD activity, was matched by age, sex and hospitalization period at a 1:3 ratio. The clinical features, proportion of high-risk KD, and incidence of CAL between the two groups were compared.ResultsThis study included 54 and 162 patients in the G6PDd and the control group, respectively. No differences were observed between the two groups in most clinical symptoms and laboratory indicators(all p > 0.05). The univariate analysis showed that G6PD deficiency was significantly associated with high-risk SANO classification (OR 3.51, 95% CI 1.60–7.72, p = 0.002). However, both groups had a similar proportion of non-SANO high-risk KD(16.67%, OR, 1.42; 95% CI, 0.60–3.35; p = 0.420). There were no significant differences in the incidence of persistent CAL beyond 8 weeks between the G6PDd group (without aspirin) and the control group (with aspirin) (12.77% vs. 6.21%; OR, 2.21; 95% CI, 0.76–6.44). Similarly, the prevalence of IVIG-resistant KD were comparable between two groups (8.51% vs. 8.70%; OR, 0.98; 95% CI, 0.31–3.12).ConclusionG6PD deficiency may not likely be a significantly risk factor for CAL in KD. SANO criteria may not be suitable for KD patients with G6PDd. Although aspirin is not routinely applied to KD patients with G6PDd, it seems unlikely have a major effect on the short-term prognosis of this subgroup.Keywords: Glucose-6-phosphate dehydrogenase deficiency, Kawasaki disease, Prognosis, Coronary artery lesions Background Kawasaki disease (KD) is an acute systemic vasculitis predominantly affecting children under the age of five, and can potentially lead to severe complications, including coronary artery dilation, aneurysms, shock, and macrophage activation syndrome. The etiology of KD remains unclear, although both genetic and environmental factors are believed to contribute to its pathogenesis [1–5]. Glucose-6-phosphate dehydrogenase deficiency (G6PDd) is an X-linked genetic disorder caused by defects in the erythrocyte enzyme G6PD, which plays a critical role in protecting erythrocyte from oxidative stress, through the generation of nicotinamide adenine dinucleotide phosphate (NADPH) [6, 7]. Emerging evidence suggests that G6PD deficiency is associated with elevated oxidative stress and dysregulated cytokine production, both of which can exacerbate inflammation in various pathological conditions [7–9]. New clinical observations, including our preliminary data, suggest that a small proportion of children with Kawasaki disease also have G6PD deficiency. While the direct involvement of G6PD deficiency in the development of coronary artery lesions (CAL) in KD has not been established, these inflammatory pathways overlap with those implicated in CAL formation [2–4], raising the possibility that G6PD deficiency may contribute to CAL progression in KD, warranting further investigation into this potential connection. Futhermore, G6PD deficiency has been reported to impair myocardial function, influence angiogenesis, and promote atherosclerosis [10–13], processes that are also relevant to KD-related cardiovascular complications [2–4]. These shared mechanisms suggest that G6PD deficiency could potentially exacerbate CAL in KD patients, a hypothesis that remains largely unexplored. Aspirin is a standard treatment for KD, aimed at controlling inflammation and reducing CAL progression [5]. However, its use is contraindicated in G6PD-deficient patients due to the risk of hemolysis [14], raising an important clinical question: does withholding aspirin therapy in this population affect CAL outcomes? The primary objective of this study is to investigate whether G6PD deficiency constitutes an independent risk factor for persistent coronary artery lesions in children with KD. Given the withholding aspirin therapy in G6PD-deficient individuals, a secondary objective is to evaluate whether the absence of aspirin therapy influences CAL outcomes in this specific subgroup. Patients and methods We conducted a retrospective case-control study. KD patients with G6PDd were consecutively enrolled from January 2016 to September 2021 at Shenzhen Children’s Hospital. All eligible KD patients with G6PD deficiency were included. The control group was matched to cases at a 1:3 ratio by age, sex and hospitalization peroid, comprising KD patients with normal G6PD activity. The primary endpoint was the persistence of CAL at 8 weeks. Analyses included comparisons of clinical features, the proportion of high-risk KD patients, and the incidence of IVIG-resistant KD in two groups. Study population and design The study was approved by the hospital’s ethics committee, and written informed consent was obtained from the legal guardians of all patients. G6PD deficiency was confirmed by laboratory testing demonstrating extremely low glucose-6-phosphate dehydrogenase activity. KD, incomplete KD, and IVIG resistant KD were diagnosed based on the American Heart Association criteria [5]. High-risk KD was defined by the presence of one or more of the following: scoring according to the KOBAYASHI [15], SANO [16], or EGAMI [17] criteria, or the occurrence of complications such as Macrophage Activation Syndrome(MAS) or Kawasaki disease shock syndrome(KDSS). MAS was diagnosed according to the 2016 EULAR/ACR/PRINTO classification criteria in KD patients [18], whereas KDSS was defined as sustained systolic hypotension or clinical signs of poor perfusion [5]. Exclusion criteria included patients with underlying cardiac diseases (e.g., congenital or acquired heart disease, arrhythmias and pericardial disease), metabolic diseases (other than G6PDd), confirmed inborn immune errors (confirmed by relevant genetic abnormalities), or cases missing G6PD activity data. Also patients with hemolytic anemia unrelated to G6PD deficiency were also excluded. Data collection Clinical data were collected from inpatient and outpatient medical records. The data included clinical manifestations, laboratory and imaging findings, treatment, and clinical outcomes. A list of clinical manifestations is provided in Table 1. The laboratory indicators included peripheral leukocyte counts, hemoglobin levels, platelet counts, urinary white blood cells, serum levels of C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), aminotransferases, albumin, ferritin, fibrinogen, triglycerides, total bilirubin, serum sodium, and serum potassium. Management included intravenous immunoglobulin (IVIG), glucocorticoids, aspirin, dipyridamole, and antibiotics. Table 1Clinical characteristics and of G6PDda with KD patientsWith G6PDd(n = 54)Without G6PDd(n = 162)Z/χ2p-valueFever daysa7(6,9)6(5,7)-2.6820.007Fever spike40.0(39.5,40.2)40.0(39.6,40.2)-0.4680.640Polymorphous rash45(83.33)139(85.80)0.1960.658Bilateral nonexudative conjunctivitis50(92.59)156(96.30)0.5590.455Oral mucous membrane changesb47(87.04)148(91.36)0.8620.353Peripheral extremity changesc42(77.78)126(77.78)0.0001.000Cervical lymphadenopathyd44(81.48)136(83.95)0.0180.673Redness at the BCG inoculation site8(14.81)22(13.58)0.0520.820Respiratory symptoms32(59.26)107(66.05)0.8140.367Gastrointestinal symptoms23(42.59)68(41.98)0.0060.937Limitation of neck movement5(9.26)16(9.88)0.0180.895Central nervous symptoms11(20.37)17(10.49)3.5020.061Data expressed as median (interquartile range) or n (%). (a) Variables such as fever days and spike were accessed by Wilcoxon Rank-Sum test. Others were analyzed by Chi-square test. (b) Oral mucous membrane changes: injected or fissured lips, injected pharynx, or strawberry tongue. (c) Peripheral extremity changes: erythema of palms or soles, edema of hands and periungual desquamation in convalescent phase. (d) Cervical lymphadenopathy: at least 1 lymph node > 1.5 cm in diameter. G6PDd Glucose-6-phosphate dehydrogenase deficiency, BCG Bacillus Calmette-Guerin Clinical characteristics and of G6PDda with KD patients Data expressed as median (interquartile range) or n (%). (a) Variables such as fever days and spike were accessed by Wilcoxon Rank-Sum test. Others were analyzed by Chi-square test. (b) Oral mucous membrane changes: injected or fissured lips, injected pharynx, or strawberry tongue. (c) Peripheral extremity changes: erythema of palms or soles, edema of hands and periungual desquamation in convalescent phase. (d) Cervical lymphadenopathy: at least 1 lymph node > 1.5 cm in diameter. G6PDd Glucose-6-phosphate dehydrogenase deficiency, BCG Bacillus Calmette-Guerin Statistical analysis Normally distributed continuous variables are presented as mean ± standard deviation (SD) and compared using independent samples t-test. Non-normally distributed data are expressed as median (interquartile range, IQR) and compared using the Mann-Whitney U test. Categorical variables were analyzed using the Chi-square or Fisher’s exact test, with odds ratios and 95% confidence intervals (CIs) reported. Given the relatively small sample size, effect sizes with CIs were emphasized to provide more informative interpretation of the results. All statistical analyses were performed using SPSS version 26.0, and a a two-sided p-value < 0.05 was considered statistically significant. Result The clinical characteristics of G6PDd with KD patients We compared 54 KD patients with G6PDd to 162 KD patients with normal G6PD activity. The clinical characteristics of both groups are summarized and in Table 1. Except for the duration of fever, which was significantly longer in the G6PDd group (Z=-2.682, p = 0.007), most clinical features did not differ significantly between the two groups. These included classic KD manifestations such as fever spikes, polymorphous rash, bilateral nonexudative conjunctivitis, and oral mucous membrane changes (e.g. cracked, red lips and “strawberry tongue”) (all p > 0.05). Others features, including peripheral extremity changes (indurated edema on the dorsum of the hands and feet, diffuse erythema on the palms and soles, and periungual desquamation), cervical lymphadenopathy, redness at the BCG inoculation site, respiratory and gastrointestinal symptoms, limitation of neck movement, and central nervous system symptoms, were similarly distributed across both groups (all p > 0.05) (Table 1). The laboratory parameters between KD patients with G6PDd and the control KD patients Hemoglobin levels(t=-5.343, p < 0.001) were significantly lower in KD patients with G6PDd, whereas ESR (Z=-1.982, p = 0.048), ferritin (Z=-5.280, p < 0.001) and total bilirubin levels(Z=-3.577, p < 0.001) were significantly elevated. However, no significant differences were observed between the two groups regarding other laboratory parameters, including peripheral white blood cell count, neutrophil percentage, platelet count, urinary white blood cells, serum CRP, procalcitonin (PCT), alanine aminotransferase (ALT), aspartate transaminase (AST), albumin, and serum sodium (all p > 0.05) (Table 2). Table 2Laboratory findingsWith G6PDd(n = 54)Without G6PDd(n = 162)Z/tp-valueWBC count (×1012/µL)15.6(12.4,19.6)15.4(12.6,19.3)−0.1750.861Hemoglobin (g/dL)a96.1 ± 15.32107.8 ± 8.58−5.343 0.000 Neutrophil percentile (%)a67.9 ± 15.7367.06 ± 14.870.3910.696Platelet (×106 /µL)431.0(308.5,504.5)371.5(315.0,464.5)−1.0570.290CRP (mg/L)59.6(29.6,114.8)63.0(37.4,101.9)−0.6460.518ESR (mm/h)78.0(55.5,89.0)66.0(47.0,83.0)−1.982 0.048 PCT (ng/ml)0.4(0.2,2.0)0.5(0.3,1.7)−0.2190.827ALT (IU/L)39.5(20.0,77.5)42.0(16.0,96.3)−0.2980.766AST (IU/L)30.0(22.8,41.3)32.0(23.0,49.0)−0.5410.589Ferritin (ng/mL)254.0(162.4,393.9)136.1(100.0,203.1)−5.280 0.000 Triglyceride (mg/dL)1.4(1.1,1.8)1.3(1.1,1.7)−1.0620.288Fibrinogen (mg/dL)a5.23 ± 1.475.50 ± 1.51−1.1240.262Total bilirubin (umol/L)10.8(6.2,21.4)6.6(4.7,9.9)−3.577 0.000 Unconjugated bilirubin (umol/L)6.1(3.4,12.2)4(2.7,6.5)−3.280 0.001 Albumin (g/L)34.4(30.2,37.3)34.95(32.4,38.0)−1.8620.063Sodium (mmol/L)135.8(134.2,137.2)135.6(133.9,137.1)−0.1110.912Potassium (mmol/L)a4.14 ± 0.634.24 ± 0.60−1.0420.299Data expressed as mean, median (interquartile range) or n (%). a. Variables such as neutrophil percentile, hemoglobin, fibrinogen, and potassium were acessed by T-test. Others were analyzed by Wilcoxon rank sum test. WBC white blood cell, CRP C-reactive protein, ESR erythrocyte sedimentation rate, PCT procalcitonin, ALT alanine aminotransferase, AST aspartate aminotransferase Laboratory findings Data expressed as mean, median (interquartile range) or n (%). a. Variables such as neutrophil percentile, hemoglobin, fibrinogen, and potassium were acessed by T-test. Others were analyzed by Wilcoxon rank sum test. WBC white blood cell, CRP C-reactive protein, ESR erythrocyte sedimentation rate, PCT procalcitonin, ALT alanine aminotransferase, AST aspartate aminotransferase Proportion of high-risk KD The proportion of high-risk KD, as defined by the SANO criteria, was significantly higher in the G6PD deficiency group (27.78%, χ2=10.55, p = 0.001; OR, 3.52, 95% CI, 1.62–7.66). In contrast, no significant differences were observed between groups for high-risk KD based on the KOBAYASHI criteria(14.81%, OR 1.31, 95% CI 0.54–3.18, p = 0.553), Non-SANO high-risk KD (16.67%, OR 1.42, 95% CI 0.60–3.35, p = 0.420), incomplete KD (11.11%, OR 1.23, 95% CI 0.45–3.35, p = 0.691), MAS(1.85%, OR 1.00, 95% CI 0.10–9.95, p = 1.000), and KDSS (5.56%, OR 3.11, 95% CI 0.61–15.85, p = 0.167) (Table 3). Table 3High-Risk Kawasaki diseaseWith G6PDd(n = 54)Without G6PDd(n = 162)χ2p-valueOR(95% CI) Criteria variables KOBAYASHI criteria[12]8(14.81)9(11.73)0.3530.5531.31(0.54–3.18) SANO criteria[13]15(27.78)16(9.88)10.5580.0013.52(1.62–7.66) EGAMI criteria[11]0(0)4(2.47)-0.574- Criteria positivea15(27.78)25(15.43)4.0910.0432.11(1.04–4.28)Incomplete KD[5]6(11.11)15(9.26)0.1580.6911.23(0.45–3.35)MAS[14]1(1.85)3(1.85)-1.0001.00(0.10–9.95)KDSS[5]3(5.56)3(1.85)-0.1673.12(0.61–15.85)Non-SANO high-risk KDc9(16.67)290(12.35)0.6510.4201.42(0.60–3.35)High-risk KD[5]d16(29.63)26(16.05)4.7680.0292.21(1.09–4.50)Data was presented with n(%). Variables were analyzed by Chi-square test. (a) Criteria positive including KOBAYASHI, SANO and EGAMI criteria. (b) Non-SANO criteria including KOBAYASHI criteria and EGAMI criteria. (c) Non-SANO high-risk KD including KOBAYASHI criteria, EGAMI criteria, MAS and KDSS. (d) High-risk KD including all criteria, IVIG-resistant KD, MAS and KDSS. KD Kawasaki disease, MAS macrophage activation syndrome, KDSS Kawasaki disease shock syndrome High-Risk Kawasaki disease Data was presented with n(%). Variables were analyzed by Chi-square test. (a) Criteria positive including KOBAYASHI, SANO and EGAMI criteria. (b) Non-SANO criteria including KOBAYASHI criteria and EGAMI criteria. (c) Non-SANO high-risk KD including KOBAYASHI criteria, EGAMI criteria, MAS and KDSS. (d) High-risk KD including all criteria, IVIG-resistant KD, MAS and KDSS. KD Kawasaki disease, MAS macrophage activation syndrome, KDSS Kawasaki disease shock syndrome Multivariable analysis of SANO risk classification We conducted sequential logistic regression to identify predictive factors for high-risk classification according to the SANO criteria (Table 4). G6PD deficiency was significantly associated with SANO high-risk in univariate analysis (OR 3.51, 95% CI 1.60–7.72, p = 0.002), while this association was fully attenuated and lost statistical significance after adjusting for total bilirubin (OR 2.47, 95% CI 0.68–9.03, p = 0.172). In contrast, bilirubin remains a strong independent predictor (OR 1.089, 95% CI 1.055–1.125, p < 0.001). Table 4Multivariable logistic regression analysis of SANO High-Risk classificationModel 1OR(95% CI), pModel 2OR(95% CI), pModel 3OR(95% CI), pG6PD deficiency3.51(1.60–7.72), p = 0.0022.47(0.68–9.03), p = 0.1723.01 (0.68–13.30), p = 0.147Total bilirubin(µmol/L)-1.089 (1.055–1.125), p < 0.0011.084(1.049–1.121), p < 0.001Hemoglobin (g/dL)-0.993(0.946–1.043), p = 0.7721.005 (0.949–1.063), p = 0.874CRP(mg/L)--1.017 (1.005–1.029), p = 0.007Platelet count(×10⁹/L)--1.002 (0.998–1.006), p = 0.327Age (months)--1.015 (0.985–1.047), p = 0.332Sex (male)--1.922 (0.526–7.027), p = 0.323Model fit statisticsModel 1: Hosmer-Lemeshow test p = NA; Overall accuracy = 85.6%; Sensitivity = 0%Model 2: Hosmer-Lemeshow test p = 0.010; Overall accuracy = 92.6%; Sensitivity = 58.1%Model 3: Hosmer-Lemeshow test p = 0.658; Overall accuracy = 94.0%; Sensitivity = 64.5%Abbreviations: OR, odds ratio; CI, confidence interval; CRP, C-reactive proteinNote: Model 1: Unadjusted; Model 2: Adjusted for bilirubin and hemoglobin; Model 3: Fully adjusted model Multivariable logistic regression analysis of SANO High-Risk classification Model fit statistics Model 1: Hosmer-Lemeshow test p = NA; Overall accuracy = 85.6%; Sensitivity = 0% Model 2: Hosmer-Lemeshow test p = 0.010; Overall accuracy = 92.6%; Sensitivity = 58.1% Model 3: Hosmer-Lemeshow test p = 0.658; Overall accuracy = 94.0%; Sensitivity = 64.5% Abbreviations: OR, odds ratio; CI, confidence interval; CRP, C-reactive protein Note: Model 1: Unadjusted; Model 2: Adjusted for bilirubin and hemoglobin; Model 3: Fully adjusted model This pattern persisted in the fully adjusted model, which included hemoglobin, CRP, platelet count, age, and sex. In this model, G6PD deficiency remained non-significant (OR 3.01, 95% CI 0.68–13.30, p = 0.147), while both bilirubin (OR 1.084, 95% CI 1.049–1.121, p < 0.001) and CRP (OR 1.017, 95% CI 1.005–1.029, p = 0.007) retained independent predictive value. The final model demonstrated good calibration (Hosmer-Lemeshow p = 0.658) and high predictive accuracy (94%). Management and prognosis A total of 7 KD patients with G6PDd received low-dose aspirin treatment, whereas the remaining 47 patients in the G6PDd group did not receive aspirin (Table 5). Except for aspirin, dipyridamole and glucocorticoids, no other antiplatelet or anti-inflammatory agents were administered. In the control group, 161 patients were treated with aspirin and one was not. The length of hospital stays was comparable between the G6PDd group without aspirin and the control group with aspirin (p = 0.142) (Table 6). There were no significant differences in the proportion of CAL that persisted beyond 8 weeks between the G6PDd group without aspirin and the control group with aspirin (161 patients) (12.77% vs. 6.21%; OR 2.2, 95% CI 0.758–6.438, p = 0.241) or in the prevalence of IVIG resistant KD (8.51% vs. 8.70%; OR 0.98, 95% CI 0.306–3.122, p = 1.000) (Table 6). Table 5ManagementWith G6PDd (n = 54)Without G6PDd (n = 162)Z/χ2p-valueIVIG (day)a6(5,8)6(5,7)−1.0520.293Aspirin7161--Glucocorticoids16(29.63)27(21.60)1.4460.229Dipyridamole51(94.44)151(93.21)01.000Antibiotics50(92.59)154(95.06)0.1180.732Vasoactive drugs0(0)2(1.23)-1.000Data expressed as median (interquartile range) or n (%). a. IVIG means the day of administration of intravenous immunoglobulin(day). And IVIG was accessed by Wilcoxon rank sum test. Other variables were analyzed by Chi-square test Management Data expressed as median (interquartile range) or n (%). a. IVIG means the day of administration of intravenous immunoglobulin(day). And IVIG was accessed by Wilcoxon rank sum test. Other variables were analyzed by Chi-square test Table 6Prognosis between G6PDd group and normal G6PD groupG6PDd group (n = 47)Normal G6PD group (n = 161)Z/χ2P valueOR (95% CI)Hospital stays (ds)6(5,10)6(5,8)-1.4670.142CAL > 8wks6(12.77)10(6.21)1.3750.2412.2(0.758–6.438)IVIG-R KD[5]4(8.51)14(8.70)0.0001.0000.98(0.306–3.122)Data expressed as median (interquartile range) or n (%). (a) Hospital days was accessed by Wilcoxon Rank-Sum test. Others were analyzed by Chi-square test. (b) CAL > 8wks coronary artery lesions that persisted beyond 8 weeks. (c) IVIG-R KD intravenous immunoglobulin resistant Kawasaki disease. CAL coronary artery lesions Prognosis between G6PDd group and normal G6PD group Data expressed as median (interquartile range) or n (%). (a) Hospital days was accessed by Wilcoxon Rank-Sum test. Others were analyzed by Chi-square test. (b) CAL > 8wks coronary artery lesions that persisted beyond 8 weeks. (c) IVIG-R KD intravenous immunoglobulin resistant Kawasaki disease. CAL coronary artery lesions Discussion This study provides a preliminary exploration of the potential interactions between G6PD deficiency and KD, two conditions with overlapping pathophysiological features such as oxidative stress and inflammatory dysregulation. Previous research has suggested that G6PD deficiency may enhance inflammatory responses through cytokine modulation [10, 11, 19] and endothelial injury mechanism [12, 13], theoretically increasing the risk of CAL in KD. However, in our study, the incidence of CAL persistent beyond 8 weeks show no significant differences (OR 2.21, 95% CI 0.758–6.438) between KD patients with low or normal G6PD activity. These findings suggest that G6PD deficiency may not substantially increase CAL risk in this population. Similarly, no significant differences were observed between the two groups in clinical manifestations, laboratory parameters, the proportion of high-risk KD (excluding SANO criteria), or the incidence of IVIG-resistant KD. Although, the study was sufficiently powered to detect large effects (e.g., risk ratio >2.0), the wide confidence intervals for these results limit the precision of these estimates, reflecting the challenges inherent to studying this rare patient population. The neutral CAL outcome observe in G6PD-deficient patients may reflect a hypothesis of complex immunometabolic balance. Although G6PD deficiency increases oxidative stress, it is also associated with anti-inflammatory effects, including impaired T cell proliferation, reduced cytokine production (such as IFN-γ, IL-17), and weakened cytotoxic function, thereby suppressing adaptive immunity. This T-cell suppression inhibits M1 polarization and IFN-γ-dependent macrophage activation, thereby attenuating a key inflammatory pathway. Meanwhile, NADPH deficiency may inhibit NLRP3 inflammasome activation and promote the differentiation of anti-inflammatory M2 macrophages, supported by the activation of the Nrf2 pathway [20, 21]. In KD, vascular injury is mainly driven by innate immunity. Although G6PD deficiency may suppress adaptive immunity, this effect appears insufficient to alleviate the overall inflammatory response. Conversely, G6PD deficiency may also induce pro-inflammatory response under certain conditions. These opposing effects likely counterbalance one another, resulting in no net increase in CAL risk [11, 20, 21]. Further mechanistic studies are warranted to elucidate these complex interactions. Interestingly, KD patients with G6PDd presented with lower hemoglobin levels, higher ESR, elevated ferritin, and increased total bilirubin, consistent with stress-induced hemolysis. Furthermore, a higher proportion of these patients were classified as high-risk KD according to the SANO criteria, whereas the proportion of high-risk KD defined by non-SANO criteria was similar between groups. We hypothesize that this difference arises from the the inclusion of bilirubin (a marker that typically increases in hemolytic state [6]) as a key component of the SANO criteria, potentially leading to risk overestimation in G6PD-deficient patients. Multivariate logistic regression analysis supported this hypothesis. After adjusting for bilirubin, the statistical significance of G6PD status disappeared, while bilirubin remained strongly and was independently associated with SANO criteria. These findings suggest that the apparent increased risk in G6PD-deficient patients is primarily attributed to hemolysis-related bilirubin elevation, rather than the severity of KD, highlighting potential limitations of applying the SANO criteria in this population. This hypothesis warrants further validation in future studies. Regarding aspirin therapy is controversial in G6PDd patients, our exploratory and observation analysis found no significant difference in CAL progression between patients who received aspirin and those who did not. This aligns with documented evidence from two cases [22, 23], who reported successful management and complete coronary recovery in G6PDd KD patients treated with IVIG alone. This observation, though limited by the sample size, this finding challenges concerns about withholding aspirin in this subgroup and suggests that the absence of aspirin therapy may not have a major impact on short-term outcomes. This align with prior studies demonstrating that high-dose aspirin does not improve the KD prognosis compared to median- or low-dose aspirin [24–28], or even to withholding aspirin, which was supported by a randomized clinical trial [29]. We emphasize that confounding by indication remains a potential limitation, and well-designed randomized controlled trials are needed to more definitively assess the role of aspirin in this patient population. Several limitations should be acknowledged. The small sample size, due to the rare coexistence of G6PD deficiency and KD, limits detection of smaller associations. This single-center cohort lacked longitudinal follow-up, restricting generalizability and evaluation of long-term cardiovascular outcomes. G6PD deficiency is genetically heterogeneous, and varying detection methods precluded enzymatic subgrouping, highlighting the need for future studies integrating molecular genotyping with WHO-based activity classification [6]. Aspirin was largely withheld in G6PD-deficient patients, with a few receiving low-dose aspirin without safety concerns, reflecting real-world clinical practice. Finally, while Ravelli’s criteria [30] for MAS perform reasonably in KD, overlapping features may cause diagnostic uncertainty, emphasizing the need to refine MAS criteria for KD. Multi-center prospective studies with larger cohorts and longer follow-up are warranted to validate these findings. Conclusion Despite these limitations, this study provides valuable preliminary evidence on the interaction between G6PD deficiency and KD. Our findings suggest that G6PD deficiency is unlikely to substantially worsen coronary outcomes in KD patients receiving standard treatment protocols. Moreover, the SANO criteria may not be appropriate for accurately assessing the risk in this specific subgroup. Although aspirin is not routinely administered to KD patients with G6PDd, our results suggest that its omission may be unlikely have a major effect on their short-term prognosis. However, these conclusions should be interpreted with caution due to the limited sample size and study design, they offer an important foundation for future research and warrant validation in larger, well-characterized cohorts with extended follow-up. Acknowledgements We thank all the patients and their families for their participation. Abbreviations Glucose-6-phosphate dehydrogenase deficiency Kawasaki disease Coronary artery lesions Intravenous immunoglobulin resistant Kawasaki disease Macrophage activation syndrome Kawasaki disease shock syndrome Aspartate aminotransferase Erythrocyte sedimentation rate C-reactive protein Procalcitonin Author contributions Peisi Cheng was involved in the conception of the paper, the acquisition, analysis and interpretation of the data, and drafted the initial manuscript, and incorporated edits from other authors. Jun Yang reviewed the manuscript. Tingyan He was involved in the conception of the paper, supervised the work, interpretation of the data and reviewed the manuscript. All authors reviewed and approved the final manuscript. Funding Shenzhen Clinical Research Center (20220819113341005); Shenzhen Clinical Research Center for Child Health and Disease (szcrc2024_04); National Natural Science Foundation of China (82302056); National Key Research and Development Program of China (2021YFC2702001); Shenzhen Fund for Guangdong Provincial High level Clinical Key Specialties (SZGSP012). Data availability The datasets were collected from the medical records of participating patients at Shenzhen Children’s Hospital. Data will be available on request by E-mailing the first author or corresponding author. Declarations Ethics approval and consent to participate All participating members were enrolled with the approval of the ethics committee of Shenzhen Children’s Hospital and provided written consent from legal guardians. Consent for publication Written consent for publication of this anonymous information was obtained from patients’ legal guardians. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets were collected from the medical records of participating patients at Shenzhen Children’s Hospital. Data will be available on request by E-mailing the first author or corresponding author.