Type 2 diabetes mellitus (T2DM), previously referred to as noninsulin-dependent diabetes or adult-onset diabetes and accounting for 90%–95% of all diabetes, is a disease caused by a gradual decrease in insulin secretion from β cells in the context of insulin resistance (1, 2). T2DM is a disease involving the interaction between genetic and environmental risk factors leading to the underlying pathophysiology of beta cell dysfunction as well as insulin resistance in liver and muscle (3, 4). Poorly controlled T2DM can lead to chronic diabetic complications such as microangiopathy (retinopathy and kidney disease), atherosclerotic cardiovascular disease, peripheral neuropathy (sensory dysfunction), and diabetic foot (5–10). In the above complications, the diabetic foot negatively affects the quality of both work and life of patients. The decline in daily living activities of patients with diabetic foot results in substantial physical and psychological burdens on the patients.

Diabetic foot is one of the most serious and costly chronic complications of diabetes. It refers to foot ulcer, infection, or deep tissue destruction related to peripheral neuropathy in the lower extremity and peripheral vascular disease (11, 12). Mild diabetic foot patients usually present with foot deformities, hypoesthesia, skin dryness, and loss of skin elasticity. Patients with severe diabetic foot may develop foot ulcers and gangrene. Diabetic foot is the main reason for non-traumatic amputations in orthopedics. At present, there are many studies on the individual risk factors of diabetic foot in patients with T2DM, but no consensus has been reached. Although there are relevant clinical guidelines as the reference basis for the formulation of clinical treatment plans, how to predict the probability of diabetic foot according to risk factors and determine the timing of interventional treatment is an urgent problem to be solved at present.

The nomogram is drawn by the individual risk factors determined by multivariate logistic regression analysis. The nomogram can graphically represent the numerical relationship between specific disease and risk factors and intuitively predict the incidence of adverse events through a scoring system without any complicated calculation formula (13). The nomogram can provide accurate and individualized risk predictions for each individual. It is convenient for clinicians to effectively screen out high-risk patients and timely take interventions. Therefore, this study aimed to develop a nomogram prediction model for the risk of diabetic foot in patients with T2DM. Early screening and identification of high-risk patients can provide the reliable reference basis for early clinical intervention.

Diabetic foot was one of the serious complications of diabetes mellitus. Diabetic foot imposed significant economic burdens on patients and society. T2DM accounted for a large proportion of diabetes mellitus. In patients with T2DM, the incidence of diabetic foot was as high as 12% to 15% (15–17). The overall incidence of diabetic foot in patients with T2DM in this study was approximately 14.9%. The incidence was similar to that reported in the above literature. Patients with T2DM who developed diabetic foot often had no obvious clinical symptoms and signs in the early stage, and sometimes only showed their decline of protective sensation. This made it easy to neglect the condition of the patient (18–20). Therefore, how to predict the risk of diabetic foot in patients with T2DM at an early stage and timely take intervention for high-risk patients is crucial.

Most of the current research on the relevant risk factors of diabetic foot has focused on intervening in the progression of diabetic foot to prevent severe ulcers and amputations (21–24). Although these studies are important, we believe that how to prevent diabetic patients from developing diabetic foot is more important and critical. On the one hand, there are few studies in this area. On the other hand, there is no consensus in this regard. In our study, univariate and multivariate logistic regression analysis found that the independent risk factors for diabetic foot in patients with T2DM were age, HbA1c, LDL, TC, smoke, and drink.

The pathogeneses of diabetic foot in patients with T2DM are as follows:

(1) Peripheral neuropathy

Patients with T2DM have various degree of neuropathy of distal lower extremities. Neuropathy of distal lower extremities mainly includes sensory, motor, and autonomic peripheral neuropathy (25). Sensory neuropathy is mainly characterized by the reduction or loss of vibration sensation (pallhypesthesia) and superficial sensation (pressure and touch) as well as subjective paresthesia. The sensation of pain in the foot is substantially declined as a consequence of chronic sensory neuropathy. As a result, the risk of foot trauma is significantly higher (26–29). Foot injuries and ulcers are neglected by patients and doctors due to the lack of pain symptoms (30, 31). Hence, foot injuries and ulcers often go undetected by doctors for weeks. Motor neuropathy of the foot is characterized by muscle atrophy, motor paralysis, and loss of muscle self-reflexes. The combination of sensory and motor peripheral neuropathy results in severe foot load accompanied by gait abnormality. As the disease progresses, the lesions of the foot will worsen due to neuropathy and increased plantar pressure load. Secretion of sweat is dysfunctional by motor paralysis due to autonomic neuropathy. Perspiration dysfunction could lead to dry skin on the foot and a reduced protective skin function, which increases the risk of injury and ulcers. Through multivariate logistic regression analysis, our study found that age, HbA1c, and drink were related to the occurrence of diabetic foot, and their OR values were 1.103, 2.509, and 1.407, respectively. According to our analysis of the patients, the sensory function of distal lower extremities of elderly patients with T2DM was worse than that of young patients, so the risk of foot injuries and ulcers was higher than that of young patients. HbA1c reflects the level of recent plasma glucose control of patients. The increase of HbA1c beyond the normal range usually indicates that the patient’s level of plasma glucose control is not ideal, leading to the occurrence of hyperglycemia. Previous studies showed that metabolic abnormalities due to hyperglycemia cause neuropathy (32). Hyperglycemia leads to nerve damage through four mechanisms, including increased levels of intracellular advanced glycation end products, activation of protein kinase C, hexosamine pathway, and polyol pathway (33). Alcohol has chronic neurotoxic effects (34), especially in patients with T2DM. This has a certain adverse effect on the sensory nerves of distal lower extremities of patients with T2DM, which leads to a decrease in the damage-sensing capacity of foot and increase the risk of diabetic foot in patients with T2DM.

(2) Peripheral vascular disease

Peripheral vascular disease is often present in the course of T2DM. Peripheral vascular disease is an atherosclerotic occlusive disease of the lower extremity. Patients with T2DM have a higher risk of peripheral vascular disease (35). In patients with T2DM, peripheral vascular disease is an important cause of the occurrence and development of diabetic foot (36). Patients with T2DM have a higher incidence of thickened basement membranes of the capillaries, atherosclerosis, endothelial cell hyperplasia, and arteriolosclerosis (37). As a result, patients with T2DM suffer from a lack of blood supply to their arteries. Poor peripheral blood supply can lead to poor wound healing of foot and worsen the condition. Patients with T2DM have reduced blood perfusion in the foot. As a result, the patients are at risk of ulcers and infections and eventually developing diabetic foot. Through multivariate logistic regression analysis, our study found that LDL, TC, and smoke were related to the occurrence of diabetic foot, and their OR values were 1.796, 1.690, and 1.539, respectively. After analyzing the patients, we believed that patients with T2DM would have a great risk of diabetic foot if their LDL and TC were higher than the normal range.LDL, a cholesterol-rich lipoprotein, is one of the risk factors for atherosclerosis (38). After chemical modification, LDL is ingested by phagocytes, forming foamy cells, and remaining in the vascular wall, resulting in a large amount of cholesterol deposition, which contributes to the formation of atheromatous plaque in the arterial wall (39–41). Thus, elevated LDL increases the risk of peripheral vascular disease in patients with T2DM, resulting in poor blood supply to the foot. In this context, diabetic foot ensues. TC is one of the risk factors for atherosclerosis in clinic (42). Therefore, increased TC will increase the risk of peripheral vascular atherosclerosis in patients. Increased TC has adverse effects on the blood supply to the foot of patients with T2DM, leading to the development of diabetic foot. Smoking can reduce the release of prostacyclin in patients, and then platelets tend to adhere to the arterial wall (43, 44). Smoking can also reduce high density lipoprotein cholesterol and increase TC in blood, resulting in atherosclerosis (45–47). Hence, smoking predisposes the patients with T2DM to atherosclerosis, which reduces peripheral blood supply to the foot. This increases the risk of diabetic foot in patients with T2DM.

At present, the preventive measures for diabetic foot mainly include the following. On the premise of glycemic control, the foot of the patients with T2DM should be checked regularly and protected preventatively (such as wearing loose-fitting shoes and socks). In clinical work, how to identify which patients need early clinical intervention is worth pondering. Meanwhile, there is a lack of related research on the nomogram prediction model for the risk of diabetic foot in patients with T2DM. Hence, we developed and validated a nomogram prediction model for the risk of diabetic foot in patients with T2DM and evaluated its clinical application value. In our study, when the cutoff value (11.6%) was taken as the threshold of DCA curve, we observed that the net clinical benefit of patients was higher than that of the other two extreme curves. This suggests that when the risk of diabetic foot is higher than 11.6% predicted by the nomogram prediction model, immediate intervention will benefit the patients clinically. When the risk of diabetic foot is lower than 11.6% predicted by the nomogram prediction model, doctors and patients can temporarily not take intervention and continue to pay attention to the dynamic changes of the disease. This facilitates clinical decision making in patients with T2DM.

There are some limitations to this study. First of all, it is a retrospective study, which requires further prospective studies in the later stage. Secondly, this study is a single-center study at the present stage. If the data of the patients from multiple centers can be included in the later stage to increase the sample size and the range of observed variables, we will further strengthen the nomogram prediction model.

In conclusion, our study found that the independent risk factors for diabetic foot among patients with T2DM were age, HbA1c, LDL, TC, smoke, and drink. In addition, our study developed an individualized nomogram prediction model, which made the prediction model visualized and easy for clinical application. The nomogram prediction model has good discrimination, calibration, and clinical usefulness in both training cohort and validation cohort. This facilitates early prediction and identification of patients at high risk of developing diabetic foot.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by The Medical Ethics Committee of the Second Affiliated Hospital of Xi’an Jiaotong University. The patients/participants provided their written informed consent to participate in this study.

Study conception and design: JW and HL; data collection and data analysis: JW and TX; manuscript drafting: JW, TX, and SG. All authors were involved in the revision of the manuscript and approved the final version of the paper.

This study was supported by the Fundamental Research Funds for the Central Universities (No. ZRZD2017008).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.