Introduction
Type 2 diabetes (T2D) currently represents a major global public health challenge, affecting more than 530 million adults in 2021, a figure that could reach 784 million by 2045 according to the International Diabetes Federation (1). In sub-Saharan Africa, the progression of T2D is particularly alarming, rising from a prevalence of 6.4% in 1990 to nearly 10.5% in 2022 (2). In Congo, this increase is driven by rapid urbanization, nutritional transition and sedentary lifestyles, exposing urban populations to high risk of insulin resistance and cardiovascular complications (3). These trends mirror patterns observed across Central Africa, where similar epidemiological shifts have been documented in neighboring countries (4).
T2D pathophysiology revolves around three major defects: tissues become resistant to insulin action, pancreatic β-cells progressively lose their function, and a state of persistent low-grade inflammation sets in (5, 6). Importantly, these pathological changes occur alongside excessive ROS generation and this oxidative imbalance severely affects how the body handles glucose and lipids (7). Among the major cellular sources of ROS, NADPH oxidases (NOX), particularly NOX2, occupy a central position (8). NOX2 stands out among cellular ROS sources as a key player in oxidative stress. This membrane-bound enzyme, abundant in immune and vascular endothelial cells, transfers electrons to oxygen molecules, producing superoxide, the primary trigger for oxidative stress (9). NOX2 activation leads to the formation of reactive derivatives such as hydrogen peroxide, which damage proteins, lipids, and DNA, contributing to endothelial dysfunction and apoptotic cell death (10, 11). Several studies have demonstrated that NOX2 is overexpressed in diabetes, promoting activation of the NF-κB pathway, production of pro-inflammatory cytokines (IL-6, TNF-α, COX-2), and increased circulating CRP, a classic marker of systemic inflammation (12–14).
Simultaneously, NOX2-dependent oxidative stress can activate the mitochondrial apoptosis pathway through caspase-3, leading to functional loss of pancreatic β-cells and T2D progression (15, 16). This dual interaction between oxidative inflammation and apoptosis contributes to worsening metabolic imbalance and establishment of a vicious cycle between hyperglycemia and oxidative stress (17, 18).
The hereditary component of T2D is well established, with studies showing that first-degree relatives of diabetic patients have a 2- to 6-fold higher risk of developing the disease (19). Polymorphisms in genes encoding NOX2 (CYBB) and its regulatory subunits (NCF1, NCF2) have been associated with increased susceptibility to oxidative stress and diabetic complications (20, 21). Moreover, the efficacy of conventional antidiabetic treatments (metformin, DPP-4 inhibitors, insulin) on modulating NOX2-dependent oxidative stress remains debated (22, 23).
Despite these advances, few studies have explored the correlation between NOX2 and inflammatory and apoptotic biomarkers in the African population, particularly in Congo. Yet, local genetic, nutritional, and environmental particularities could influence oxidative and inflammatory signaling pathways (24). In this context, the present study aims to evaluate the associations between NOX2 activity and the main inflammatory (IL-6, CRP, COX-2) and apoptotic (Caspase-3) biomarkers in Congolese patients with T2D; explore the influence of family history on NOX2 levels; and analyze the impact of antidiabetic treatments on the oxidative profile, in order to better understand the pathophysiological mechanisms involved and consider potential local therapeutic targets.
Discussion
The present study aimed to evaluate the involvement of NOX2, a key enzyme in oxidative stress, in regulating inflammatory and apoptotic pathways in Congolese patients with type 2 diabetes. Our findings show that diabetic patients exhibit significantly higher plasma NOX2 levels than non-diabetic controls, associated with marked metabolic, inflammatory, and apoptotic disruption.
Our study included 143 type 2 diabetic patients and 136 healthy controls. Both groups were comparable for sex ratio (p = 0.892) and height (p = 0.687), but differed significantly for age (p = 0.003). Diabetic patients presented an anthropometric profile characterized by major weight excess, with mean weight 12.3 kg higher (p < 0.001), waist circumference increased by 18.6 cm (p < 0.001), and mean BMI of 31.8 kg/m² versus 24.3 kg/m² in controls (p < 0.001). Obesity (BMI ≥ 30 kg/m²) affected 48.95% of diabetic patients versus no controls, confirming the central role of weight excess, particularly abdominal obesity, in type 2 diabetes pathophysiology.
Socioeconomically, significant disparities were observed: 52.44% of diabetics were unemployed versus 29.41% of controls and higher education level was more frequent among controls (48.52% vs 27.27%). Mean diabetes duration was 10.2 ± 5.8 years. These socioeconomic differences may have important implications for disease management, treatment adherence, and chronic stress levels, all of which can influence oxidative stress pathways.
In our study, 64.3% of patients received metformin, reflecting its status as first-line treatment (5). Despite its well-documented antioxidant properties (mitochondrial complex I inhibition, AMPK activation, AGE reduction) (22, 25), mean NOX2 levels in these patients remained very high, approximately 4-fold higher than controls. This observation suggests that metformin, while effective for glycemic control, is insufficient to normalize NOX2-dependent oxidative stress in established T2D. Residual hyperglycemia (mean HbA1c of 9.56%) maintains NOX2 activation through multiple pathways (7), and metformin primarily acts on mitochondrial oxidative stress with limited effect on direct NOX2 activation.
Regarding DPP-4 inhibitors, prescribed in 36.4% of our patients, they have demonstrated pleiotropic effects beyond glycemic control, notably increased GLP-1 which modulates cytokine production and improves endothelial function, as well as reduced oxidative stress through NOX inhibition via a PKA-dependent pathway (26). In our study, patients on DPP-4 inhibitors alone presented numerically lower adjusted NOX2 levels (18.9 ± 1.2 ng/mL) compared to those on metformin alone (19.8 ± 0.9 ng/mL) or sulfonylureas (20.6 ± 1.1 ng/mL), although these differences did not reach statistical significance in post-hoc comparisons. This trend could reflect a specific antioxidant effect, but larger comparative studies adjusted for diabetes duration and HbA1c would be necessary to confirm this potential protective effect.
Insulin therapy concerned only 9.1% of patients despite an overall HbA1c of 9.56%, suggesting delayed therapeutic intensification well documented in sub-Saharan Africa (3). After adjusting for HbA1c, BMI, and diabetes duration through ANCOVA, patients on insulin therapy presented significantly higher NOX2 levels (22.8 ± 1.5 ng/mL) compared to those on metformin or DPP-4 inhibitors alone (p = 0.041 and p = 0.028, respectively). This likely reflects several factors: more advanced and severe diabetes requiring therapeutic intensification, prolonged exposure to chronic hyperglycemia before insulin initiation, and potentially more marked treatment resistance. Importantly, this cross-sectional analysis cannot determine whether insulin therapy itself affects NOX2 levels. While some studies suggest insulin may have antioxidant properties through PI3K/Akt pathway activation, others report pro-oxidant effects at supraphysiological doses. Our data argue for earlier insulin initiation in patients with inadequate glycemic control, consistent with international recommendations (27).
Our results show that 47.6% of T2D patients had first-degree family history of diabetes versus 39.7% of controls (OR = 1.38 [95% CI: 1.08-2.94], p = 0.023), confirming the important role of hereditary component (19). The most interesting observation is that patients with family history presented significantly higher NOX2 levels (22.4 vs 19.3 ng/mL, p = 0.002), suggesting genetic predisposition to increased oxidative stress.
The CYBB gene, located on the X chromosome, encodes the gp91phox catalytic subunit of NOX2. Several variants of this gene have been associated with increased enzymatic activity and excessive ROS production (20, 21). Similarly, polymorphisms in NCF1 (p47phox) and NCF2 (p67phox) genes, which encode cytosolic subunits necessary for NOX2 complex assembly, can modulate ROS production (21). The frequency of these variants could differ in African populations compared to Caucasian or Asian populations, which would explain specific risk profiles and justify dedicated pharmacogenomic studies.
Beyond genetic factors, families share lifestyle habits and similar socioeconomic environment. In our study, 52.4% of T2D patients were unemployed, a factor associated with chronic psychosocial stress activating the hypothalamic-pituitary-adrenal axis and stimulating NOX2 via cortisol (3). The interaction between genetic predisposition and unfavorable environmental factors could therefore amplify oxidative stress in individuals with family history of diabetes.
These observations reinforce the need for early T2D screening in first-degree relatives, particularly in the presence of additional risk factors. Integration of oxidative biomarkers like NOX2 into risk stratification algorithms could allow identification of high-risk subjects requiring intensive preventive intervention (27).
The NOX2 increase observed in this study confirms the existence of chronic oxidative stress in T2D. These results are consistent with previous work showing that prolonged hyperglycemia induces overproduction of reactive oxygen species (ROS) through NADPH oxidase 2 activation (20, 23). This oxidative hyperactivation contributes to several deleterious processes including non-enzymatic protein glycation, endothelial dysfunction, and worsening insulin resistance, which are central mechanisms in diabetes pathophysiology (7, 10, 11).
The positive correlations we observed between NOX2 and glycemic control parameters (HbA1c, glucose, insulin) reinforce the hypothesis that NOX2-dependent oxidative stress constitutes a marker of T2D metabolic severity. Our multiple regression analysis identified HbA1c as the most powerful independent predictor of NOX2 levels (β = 0.31, p < 0.001), suggesting that a 1% increase in HbA1c is associated with a 1.6 ng/mL elevation in NOX2. These observations are comparable to those reported in European and Asian populations (12, 14), but constitute the first data of this magnitude in the Congolese population, where genetic particularities (CYBB gene polymorphisms) and environmental factors (nutritional transition, psychosocial stress) could differently modulate NOX2 activation (24, 28).
Partial correlation analyses controlling for HbA1c revealed that the relationships between NOX2 and inflammatory markers are partly mediated by glycemic control. Specifically, the correlation between NOX2 and IL-6 became non-significant after adjusting for HbA1c (r_partial = -0.08, p = 0.31), while the NOX2-COX-2 correlation remained significant but attenuated (r_partial = -0.18, p = 0.03). This suggests that hyperglycemia may partially mediate the observed relationships between oxidative stress and inflammation.
Furthermore, mediation analysis demonstrated that HbA1c partially mediates the relationship between BMI and NOX2 (indirect effect: β = 0.14, p = 0.04; proportion mediated: 25%). This indicates that while adiposity has direct effects on oxidative stress likely through adipokine secretion and chronic inflammation a substantial portion of its impact is channeled through worsening glycemic control.
The negative correlations observed between NOX2 and inflammatory biomarkers (IL-6, COX-2, CRP) deserve particular attention as they contrast with the classical paradigm whereby oxidative stress stimulates inflammation via NF-κB activation (12–14). Several mechanisms can explain these inverse correlations: prolonged NOX2 activation could induce depletion of inflammatory mediators through lipid peroxidation damaging immune cells, direct cytokine inactivation by oxidation, or desensitization of cytokine receptors (22, 23, 29).
In our cohort, mean diabetes duration was 10.2 years, suggesting that these negative correlations possibly reflect a negative feedback phenomenon, where sustained oxidative stress progressively reduces cellular inflammatory capacity. This hypothesis is supported by recent studies documenting “cytokine exhaustion” in chronic metabolic inflammation (22). However, it must be acknowledged that this interpretation remains speculative and requires additional mechanistic studies, particularly at cellular and molecular levels, as the relationship between NOX2 and inflammation strongly depends on diabetes evolutionary stage and tissue context (14).
The absence of significant correlation between NOX2 and CRP (r = +0.04, p = 0.62) in diabetic patients, despite both being markedly elevated compared to controls, provides additional insight into these complex relationships. This finding suggests that NOX2 and CRP may reflect different aspects of the inflammatory process: CRP is primarily produced by hepatocytes in response to IL-6 and reflects systemic, chronic inflammation and NOX2 is activated locally in immune cells, endothelial cells, and adipocytes and responds rapidly to metabolic perturbations. Within the diabetic population, the temporal dynamics of these markers likely differ: NOX2 responds rapidly to acute hyperglycemic spikes (hours), whereas CRP reflects chronic inflammatory burden accumulating over days to weeks. Additionally, CRP levels are influenced by factors independent of oxidative stress, including obesity-related adipokine secretion, subclinical infections, and genetic variants (CRP gene polymorphisms).
The weak within-group correlation suggests that while both NOX2 and CRP are elevated in T2D compared to controls (p < 0.001 for both), their variability within diabetic patients is driven by different mechanisms. This finding highlights the multifactorial nature of diabetic inflammation and reinforces that NOX2 captures oxidative stress pathways not fully reflected by classical inflammatory markers.
Similarly, the very weak negative correlation between NOX2 and Caspase-3 (r = -0.004, p < 0.05) suggests complex dynamics. Although NOX2 activation is classically associated with apoptosis via cytochrome c release from mitochondria (15, 16), excessive and prolonged activation can paradoxically lead to mitochondrial dysfunction that alters classical apoptotic pathways (30). Direct oxidation of caspase-3 can reduce its enzymatic activity (31), and cells exposed to intense oxidative stress can develop resistance to apoptosis (32).
This phenomenon of apoptosis resistance could have important consequences in diabetes pathophysiology. Indeed, prolonged survival of dysfunctional cells, particularly at the endothelial and immune levels, could contribute to maintaining a chronic low-grade inflammatory state and favor progression of microvascular and macrovascular complications. However, this hypothesis needs to be validated through longitudinal studies and specific tissue analyses.
Partial correlation analyses controlling for HbA1c revealed that the relationships between NOX2 and inflammatory markers are partly mediated by glycemic control. Specifically, the correlation between NOX2 and IL-6 became non-significant after adjusting for HbA1c (r_partial = -0.08, p = 0.31), while the NOX2-COX-2 correlation remained significant but attenuated (r_partial = -0.18, p = 0.03). This suggests that hyperglycemia may partially mediate the observed relationships between oxidative stress and inflammation, consistent with established frameworks for understanding mediating variables in complex biological systems (33, 34).
Our multivariate analysis identified four independent predictors of NOX2 levels: HbA1c (β = 0.31, p < 0.001), BMI (β = 0.24, p = 0.003), diabetes duration (β = 0.19, p = 0.012), and family history (β = 0.16, p = 0.028). This model explains 38% of NOX2 variance (R² = 0.38, p < 0.001). These results position NOX2 as an integrator of metabolic (hyperglycemia, adiposity), temporal (disease duration), and genetic (family history) signals, suggesting that oxidative stress constitutes a central mechanism linking different risk factors to diabetic complications. This integrative vision has important implications for risk stratification and therapeutic personalization. NOX2 measurement could complement traditional markers (HbA1c, glucose, lipid profile) to better stratify individual risk and guide therapeutic intensification, particularly in patients with family history or poor glycemic control. A NOX2 threshold > 20 ng/mL could identify patients requiring enhanced surveillance and more aggressive therapeutic optimization.
Several factors specific to the Congolese population may modulate NOX2 activity and its pathophysiological impact, making our findings particularly relevant for understanding diabetes in African contexts: The urban population of Brazzaville is undergoing rapid dietary transition characterized by increased consumption of ultra-processed foods rich in trans fatty acids and refined sugars, decreased traditional diet rich in fiber (cassava, plantain, leafy vegetables), and higher dietary AGE (advanced glycation end-product) load from fried and grilled foods. This nutritional shift could exacerbate NOX2-dependent oxidative stress. Dietary AGEs directly activate NOX2 through RAGE (receptor for AGEs) signaling, creating a vicious cycle between hyperglycemia-induced AGE formation and oxidative stress amplification (35).
Studies in Central Africa have documented significant deficiencies in key antioxidants: vitamin E (35-45% prevalence of deficiency), selenium (28% below optimal levels), and zinc (40% subclinical deficiency in urban populations) (36). These deficiencies impair endogenous antioxidant defenses glutathione peroxidase requires selenium, superoxide dismutase requires zinc potentially amplifying the oxidative damage caused by NOX2 activation, creating a double burden of excessive ROS production coupled with inadequate antioxidant defenses. Indeed, earlier work in African populations has demonstrated strong associations between oxidative stress markers and metabolic syndrome features in diabetic patients (37), underscoring the clinical relevance of this oxidative-nutritional interplay.
The high unemployment rate in our cohort (52.4% vs 29.4% in controls) is accompanied by chronic psychosocial stress. Stress activates the hypothalamic-pituitary-adrenal axis, and cortisol has been shown to stimulate NOX2 expression and activity in immune and vascular cells. Chronic stress may therefore represent an underappreciated driver of oxidative stress in urban African populations undergoing rapid socioeconomic changes.
These airborne pollutants represent an often-overlooked source of oxidative stress, as they activate NOX2 in pulmonary macrophages and endothelial cells (38).
African populations exhibit distinct allelic frequencies for genes involved in oxidative metabolism. Polymorphisms in CYBB (NOX2), NCF1 (p47phox), and NCF2 (p67phox) may differ from Caucasian or Asian populations, potentially modulating basal NOX2 activity and response to hyperglycemia. This genetic diversity remains largely unexplored and justifies dedicated pharmacogenomic studies in African cohorts.
These contextual factors reinforce the importance of conducting research in diverse populations rather than extrapolating findings from Western studies. The interplay between genetic susceptibility, nutritional deficiencies, and environmental stressors may create a unique oxidative stress phenotype in sub-Saharan African patients with diabetes.
Pharmacological targeting of NOX2 represents a promising therapeutic strategy. Several selective NOX inhibitors have been developed, notably GKT137831 and VAS2870, which have shown efficacy in reducing oxidative stress and inflammation in murine diabetes models (31, 39, 40). While NOX2 predominates in immune cells and has been our focus here, it's worth noting that other members of the NOX family, particularly NOX5, also play important roles in vascular cells (41), and the broader family of NADPH oxidases contributes to various aspects of vascular pathology in diabetes (42). These molecules act by blocking NOX2 complex assembly or directly inhibiting catalytic subunit enzymatic activity. Although promising in preclinical research, none of these molecules has yet been approved for clinical use in humans, primarily due to persistent questions regarding their selectivity, pharmacokinetics, and potential adverse effects.
Our data also highlight the need to integrate NOX2-dependent oxidative stress into early biomarker panels for diabetes and its cardiovascular complications. NOX2 measurement could complement traditional markers (HbA1c, glucose, lipid profile) to better stratify individual risk and guide therapeutic intensification, particularly in patients with family history or poor glycemic control. A NOX2 threshold > 20 ng/mL could identify patients requiring enhanced surveillance and more aggressive therapeutic optimization. This approach may be especially valuable in sub-Saharan African healthcare settings, where resources for comprehensive diabetes care are often limited and tools for predicting complications are urgently needed (43). Furthermore, given the sensitivity analysis showing that statin use did not significantly affect NOX2 levels (20.5 ± 5.0 vs 21.0 ± 5.3 ng/mL in non-users, p = 0.58), targeting oxidative stress through NOX-specific pathways may provide benefits beyond those achieved with conventional cardiovascular medications.
This study nevertheless has certain limitations, notably its cross-sectional nature does not allow establishment of strict causal relationships between NOX2 and observed metabolic alterations; analyses do not account for diurnal variations or potential nutritional influences; and finally, the sample, although representative, could be extended to other regions of Congo to strengthen result generalization. Despite these limitations, our observations provide novel data in sub-Saharan African populations, contributing to better understanding of oxidative stress’s role in diabetes.