Predicting hypertension in type 2 diabetes mellitus: Insights from a nomogram model

Abstract

The prevalence of type 2 diabetes mellitus (T2DM) is rising, with hypertension as a common comorbidity that significantly increases cardiovascular and microvascular risks. Accurate prediction of hypertension in T2DM is essential for early intervention and personalized management. In this editorial, we comment on a recent retrospective study by Zhao et al, which developed a nomogram model using a large cohort of 26850 patients to predict hypertension risk in patients with T2DM. The model incorporated key independent risk factors, including age, body mass index, duration of diabetes, low-density lipoprotein cholesterol and urine protein levels, demonstrating promising discriminative power and predictive accuracy in internal validation. However, its external applicability requires further confirmation. This editorial discusses the clinical value and limitations of the predictive model, highlighting the unfavorable impact of hypertension on T2DM patients. Future research should evaluate the potential contribution of other risk factors to enhance risk prediction and improve the management of T2DM comorbidities.

Key Words: Hypertension; Type 2 diabetes mellitus; Diabetes; Risk prediction; Nomogram model; Insulin resistance; Inflammatory markers; Blood pressure variability; Serum uric acid

Core Tip: In this editorial, we comment on the retrospective study by Zhao et al, recently published in the World Journal of Diabetes, which developed a nomogram model to predict the risk of hypertension in patients with type 2 diabetes mellitus (T2DM). We highlight its limitations, including the need for external validation and broader hypertension diagnosis. Future studies should consider this and include additional factors, such as insulin resistance, inflammatory markers, blood pressure variability and serum uric acid, to further refine risk prediction and improve T2DM management.

INTRODUCTION

Type 2 diabetes mellitus (T2DM) is a growing global health concern, with its prevalence projected to rise from 537 million in 2021 to 783 million by 2045[1]. Hypertension is the most prevalent comorbidity among T2DM patients, and together these two conditions are leading causes of premature mortality and disability worldwide[2]. In China, 8.7% of the general population has both hypertension and T2DM[3], whereas 59.9% of individuals with T2DM are affected by hypertension[4].

As the most common modifiable cardiovascular risk factor, hypertension further accelerates the onset and progression of cardiovascular and microvascular complications in T2DM, significantly contributing to global mortality[5]. Current clinical guidelines recommend initiating antihypertensive therapy when blood pressure (BP) exceeds 130/80 mmHg in T2DM patients, along with routine screening[6-8]. However, existing prediction models have shown limited accuracy and lack robust validation across diverse populations, highlighting the need for improved risk stratification tools. Therefore, the accurate prediction of hypertension in T2DM is crucial for timely intervention and precision management.

HYPERTENSION BURDEN IN T2DM

Hypertension significantly increases all-cause mortality and vascular morbidity in T2DM patients. In the Framingham cohort, its coexistence was associated with a 72% higher risk of all-cause mortality and a 57% higher risk of cardiovascular events in T2DM patients, substantially higher than the 7% and 9% adjusted risks observed in those without hypertension, underscoring its predominant contribution to adverse outcomes[9]. Notably, hypertension-related damage can occur early in the course of BP elevation. For example, one study reported that over 25% of T2DM patients with hypertension or high-normal BP showed an increased risk of cardiac injury, as reflected by elevated N-terminal proBNP levels[10]. Moreover, the coexistence of T2DM and hypertension significantly increased the risk of major adverse cardiovascular events and microvascular complications, including retinopathy, neuropathy, and nephropathy[11].

The pathophysiological mechanisms underlying heightened susceptibility to hypertension in T2DM are multifactorial (Figure 1). In early stages, hyperglycemia and hyperinsulinemia increase circulating plasma volume, while in later stages, it is driven by vascular remodeling[12]. Specifically, elevated oxidative stress and fibrosis impair endothelial integrity and initiate vascular remodeling. These processes are further exacerbated by activation of the renin-angiotensin-aldosterone system and sympathetic nervous system, which increase peripheral resistance and arterial stiffness, ultimately disrupting BP homeostasis and accelerating microvascular injury[12,13]. Additionally, common modifiable risk factors in T2DM, such as obesity, smoking, and physical inactivity further exacerbate hypertension progression[11].

Figure 1 Mechanisms contributing to increased susceptibility to hypertension in patients with type 2 diabetes mellitus. Shared risk factors contribute to both type 2 diabetes mellitus (T2DM) and hypertension. In T2DM, hyperglycemia and hyperinsulinemia increase circulating plasma volume during early stages and promote vascular remodeling in later stages. The late stages are further characterized by increased oxidative stress, fibrosis, and activation of the renin-angiotensin-aldosterone system and sympathetic nervous system, which together impair blood pressure regulation and elevate hypertension risk. RAAS: Renin-angiotensin-aldosterone system; SNS: Sympathetic nervous system.

Regulation of BP plays a crucial role in T2DM management. The United Kingdom Prospective Diabetes Study was the first to demonstrate that maintaining BP below 150/85 mmHg reduced diabetes-related mortality by 32% and microvascular complications by 37%[14]. Subsequent studies confirmed these findings, showing that intensive control of BP lowers composite cardiovascular risk, although the optimal targets remain debated[15,16]. A meta-analysis found that a target of 130/80 mmHg provides optimal cardiovascular protection in older patients[17]. Additionally, intensive BP control significantly reduces adverse renal outcomes, slows the progression of retinopathy and albuminuria, and offers protection against dementia and cognitive decline[18-21]. A recent large cohort study involving over 12000 patients further demonstrated that reducing systolic BP (SBP) from 140 to 120 mmHg decreased the relative risk of major cardiovascular events by 21% in T2DM patients with hypertension[21].

ADVANCES IN PREDICTING T2DM HYPERTENSION

Previous efforts to predict hypertension in T2DM patients have faced significant limitations, hindering their clinical applicability. Traditional tools such as the Framingham Cardiovascular Risk Score were originally designed for the general population to estimate 10-year cardiovascular event risk[22]. However, they are not specifically designed for hypertension prediction and fail to capture the distinct pathophysiological interactions between diabetes and hypertension, thereby limiting their predictive accuracy in this population. To enhance predictive performance, researchers have explored nomogram-based approaches. Deng et al[23] developed a nomogram to estimate hypertension risk in the general population, achieving an area under the curve (AUC) of 0.750 in receiver operating characteristic (ROC) analysis. Yang et al[24] constructed a nomogram based on data from 706 T2DM patients, reporting an AUC of 0.762 in the training set. However, its applicability remains limited due to its small sample size and the omission of critical predictors, such as body mass index (BMI) and urinary albumin, both of which are highly relevant to hypertension risk in T2DM[23].

To overcome these limitations, in this issue of the World Journal of Diabetes, Zhao et al[25] conducted a retrospective study involving 26850 T2DM patients from Anhui Province, China. Utilizing routinely collected clinical data, they developed a hypertension risk prediction nomogram based on multivariable logistic regression. Key predictors were selected via LASSO regression, and model performance was evaluated through AUC, calibration curve, and decision curve analysis. The final model incorporated key predictors including age, low-density lipoprotein cholesterol, BMI, diabetes duration and urine protein levels. In contrast to earlier tools, this model integrated both conventional risk factors and T2DM-specific indicators, such as urinary proteins. Internal validation demonstrated robust discriminative ability, with AUCs of 0.823 and 0.812 in the training and validation sets, respectively. However, these values fall slightly below the ideal threshold of 0.85.

This nomogram model effectively identifies high-risk T2DM patients, supporting timely preventive strategies. Its large sample size and reliance on readily available clinical parameters enhance its feasibility and generalizability in real-world practice. Overall, this model enables early identification of at-risk patients and reduces cardiovascular and renal complications, thereby alleviating the burden of this common comorbidity.

EXPANDING PREDICTION: CURRENT LIMITATIONS AND ADDITIONAL RISK FACTORS

The model proposed by Zhao et al[25] represents a valuable advancement, yet several limitations constrain its broader clinical utility. Firstly, it was developed using data from a single Chinese center without external validation, which may limit its generalizability. To address this, future studies should pursue external validation through multicenter and multinational collaborations. Public international cohorts, such as the United Kingdom Biobank or NHANES, may serve as valuable resources for cross-population validation. Moreover, given the known ethnic and regional variations in BP patterns and their association with T2DM[26], incorporating heterogeneous populations will be essential to improve the model’s robustness and clinical relevance. Secondly, hypertension was defined by office measurements, yet elevated ambulatory BP is independently associated with increased cardiovascular risk, even when office readings are normal[27]. This supports current USPSTF guidelines to utilize 24-hour ambulatory BP monitoring (ABPM) to detect masked or nocturnal hypertension[28,29]. Additionally, cumulative SBP load (proportion of SBP ≥ 130 mmHg) has demonstrated superior predictive value for cardiovascular events and mortality compared to single-threshold definitions[24], highlighting the need for a broader hypertension definition. Thirdly, this model excluded adolescents with T2DM, among whom approximately 25% exhibit both hypertension and albuminuria[30], as well as individuals with prediabetes who have been shown to benefit from intensive SBP control in post-hoc analyses of the SPRINT trial[19]. Finally, restricted range of included variables may overlook other risk factors, as discussed later (Figure 2).

Figure 2 Comparison of risk factors from Zhao et al’s nomogram[25] (left) with additional potential risk factors (right) for predicting hypertension in type 2 diabetes mellitus. The nomogram incorporates age, low-density lipoprotein cholesterol, body mass index, diabetes duration, and urine protein. Additional factors such as insulin resistance (e.g. homeostatic model assessment of insulin resistance, metabolic score for insulin resistance, triglyceride-glucose index), inflammatory markers [e.g. C-reactive protein, interleukin (IL)-6, IL-1β, tumor necrosis factor alpha, interferon-γ, neutrophil-to-lymphocyte ratio, systemic immune-inflammation index], blood pressure (BP) variability (e.g. 24-hour ambulatory BP monitoring, pulse wave velocity) and serum uric acid may further improve risk prediction beyond the original model. This figure was created by BioRender.com (Supplementary material). BMI: Body mass index; LDL-C: Low-density lipoprotein cholesterol; HOMA-IR: Homeostatic model assessment of insulin resistance; METS-IR: Metabolic score for insulin resistance; TyG: Triglyceride-glucose; CRP: C-reactive protein; IL: Interleukin; ABPM: Ambulatory blood pressure monitoring; TNF-α: Tumor necrosis factor alpha; IFN-γ: Interferon-γ; NLR: Neutrophil-to-lymphocyte ratio; SII: Systemic immune-inflammation.

Beyond Zhao’s model[25], recent studies have underscored other potential predictors of hypertension risk in T2DM. Insulin resistance, a hallmark of T2DM, contributes to vascular dysfunction and can be assessed using the homeostatic model assessment of insulin resistance (HOMA-IR), which is calculated based on fasting glucose and insulin levels[31]. In the HCHS/SOL cohort, HOMA-IR was positively associated with SBP and DBP independent of BMI[31]. Similarly, the metabolic score for insulin resistance (METS-IR), which better accounts for triglycerides, BMI and HDL-C than HOMA-IR, was independently associated with hypertension risk in the NHANES database, with a 3.89-fold higher risk observed in the highest vs lowest quartile[32]. The triglyceride-glucose (TyG) index, another practical surrogate derived from fasting triglyceride and glucose levels, has also been proposed as a predictor of hypertension, showing a positive correlation when values are below 8.1 and an inverse relationship above this threshold[33]. Among the three mentioned markers, TyG and METS-IR have been incorporated into several hypertension risk models in the general population[32,33]. Although not yet applied in T2DM-specific models, their calculation based on routine parameters and closer links to metabolic dysfunction suggest their potential to enhance predictive accuracy and warrant further investigation.

Inflammation is another important pathophysiological pathway linking T2DM and hypertension. Classical markers such as C-reactive protein and interleukin (IL)-6 are well-established in this context[34], while cytokines like IL-1β, tumor necrosis factor alpha, and interferon-γ have been linked to incident hypertension independent of traditional risk factors[35]. In addition, emerging composite indices, such as systemic immune-inflammation index and neutrophil-to-lymphocyte ratio, have emerged as independent predictors[36,37]. These inflammatory responses contribute to hypertension via mechanisms like endothelial dysfunction, oxidative stress, and mitochondrial dysfunction[11,38]. Although antihypertensive treatments can affect cytokine levels[38], the persistent low-grade inflammation in T2DM highlights the potential of these markers for risk stratification in this population, warranting further investigation.

BP variability (BPV), defined as temporal fluctuations in BP and typically assessed through 24-hour ABPM, has been implicated in the development of arterial stiffness and adverse cardiovascular outcomes[28,39]. Evidence also supports a bidirectional causal relationship between arterial stiffness and hypertension[40]. Arterial stiffness, commonly measured by pulse wave velocity (PWV), has been shown to improve the predictive accuracy for hypertension risk combined with traditional risk factors[41]. Although these findings highlight the value of BPV and PWV in hypertension risk assessment, their predictive utility in individuals with T2DM remains underexplored and warrants further investigation.

Serum uric acid (SUA) has also been identified as a promising marker. In adolescents with T2DM, elevated baseline SUA is associated with a higher risk of hypertension and urinary proteins excretion[42]. Moreover, longitudinal data showed that the relationship between elevated SUA and cardiovascular disease risk is weakened after adjusting for BP, underscoring the close link between SUA and hypertension[43]. However, most evidence is derived from general or pediatric populations, and whether SUA serves as an effective predictor of hypertension in adults with T2DM needs further validation.

Additionally, previous nomogram-based models for predicting hypertension risk in the general population have identified hematologic parameters, including mean corpuscular hemoglobin concentration, mean platelet volume, total bilirubin, along with family history of hypertension as independent risk factors[23]. Additionally, lifestyle factors such as smoking, psychological stress and dietary potassium or sodium intake, are also associated with hypertension onset and progression[44,45].

Collectively, although factors such as insulin resistance, inflammatory markers, BPV and SUA are good predictors of hypertension in the general population, their roles in T2DM-specific prediction models remain insufficiently explored. These variables should be considered as promising rather than definitive risk predictors. Given that some of these markers may be influenced by confounders such as BMI, future studies are necessary to assess their independent predictive value and determine whether incorporating selected indicators could improve the accuracy of hypertension risk stratification in T2DM populations.

CONCLUSION

In conclusion, hypertension is a major comorbidity in T2DM, significantly impacting cardiovascular and microvascular outcomes. The nomogram model developed by Zhao et al offers promising predictive value for hypertension in T2DM patients[25]. However, its external applicability needs further validation. Future research should consider additional factors such as insulin resistance, inflammatory markers, BPV, and SUA to refine prediction models and improve T2DM management.