Family history of diabetes is associated with diabetic foot complications in type 2 diabetes

To investigate the relationship between diabetic foot complications (DFCs) and clinical characteristics, especially the number and types of first-degree family members with diabetes. A total of 8909 type 2 diabetes patients were enrolled. The clinical characteristics of these patients, including DFCs and family history of diabetes (FHD), were collected from medical records. Multiple regression was used to investigate the association between FHD and DFCs after adjusting for confounding factors. The patients with one and more than one first-degree family member with diabetes accounted for 18.7% and 12.8%, respectively. The proportions of the participants with a father with diabetes, a mother with diabetes, both parents with diabetes, siblings with diabetes, father and siblings with diabetes, mother and siblings with diabetes, and both parents and siblings with diabetes were 3.5%, 6.2%, 1.1%, 14.4%, 1.5%, 4%, and 0.7%, respectively. The multiple regression analysis showed that the number of family members with diabetes was positively associated with DFCs. However, among the different types of FHD, only the patients with a mother with diabetes showed a statistical association with DFCs. In addition to FHD, other factors, including gender, body mass index, platelet count, hemoglobin levels, albumin levels, high-density cholesterol levels, diabetic peripheral neuropathy, and the use of lipid-lowering agents, oral hypoglycemic agents, and insulin, were also associated with DFCs. DFCs were associated with different numbers of family members with diabetes and types of FHD. This association reveals the importance of genetic and environmental factors in DFCs and highlights the importance of adding FHD to public health strategies targeting detecting and preventing the disease.

Definitions. The definition of FHD was the presence or absence of diabetes among the first-degree family members, including father, mother, and siblings. According to the different types of first-degree family members with diabetes, all patients were subdivided into seven groups, including fathers with diabetes, mothers with diabetes, parents with diabetes, siblings with diabetes, fathers and siblings with diabetes, mothers and siblings with diabetes, or both parents and siblings with diabetes. The formula of the body mass index (BMI) and waisthip ratio (WHR) were as follows: weight (kg)/height (m 2 ) and waist/hip, respectively. FBG and fasting c-peptide levels were used to calculate homoeostatic model assessment 2 of β-cell function (HOMA2-B) and insulin resistance (HOMA2-IR). The estimated glomerular filtration rate (eGFR) was calculated based on the Modification of Diet in Renal Disease (MDRD) formula. The definition of drinkers or smokers was a drinking or smoking time ≥ 1 year. The treatment of angiotensin-converting enzyme inhibitor (ACEI), angiotensin receptor blockers (ARB), oral hypoglycemic agents (OHA), or insulin were referred to regular medications taking ≥ 3 months.
The diagnosis of DFCs was based on the following three criteria: (1) diagnosed with diabetes; (2) foot tissue dystrophy (ulcer or gangrene); and (3) accompanied by neuropathy or/and vasculopathy 18 . Neurologic examination was used to assess the foot dorsum and position in the toes sense and to evaluate the patellar reflexes and deep tendon. DR is an important manifestation of diabetic microvascular disease, which is classified as no proliferative diabetic retinopathy (NPDR) or proliferative diabetic retinopathy (PDR) based on new retinal blood vessel formation. Diabetic peripheral neuropathy (DPN) was diagnosed by clinical symptoms of sensory nerve and autonomic nerve (such as persistent pain and/or sensory disturbance in the extremities or lower extremities; decreased sensation), the examination of neurology and electrophysiologic investigations. Diabetic nephropathy (DN) was defined by a urinary albumin creatinine ratio (UACR) ≥ 30 mg/24 h. Microalbuminuria was defined as UACR ≥ 30 mg/24 h but ≤ 300 mg/24 h, and macroalbuminuria was defined as UACR ≥ 300 mg/24 h. The definition of coronary heart disease (CHD) was impaired myocardium attributable to imbalanced coronary blood flow and myocardial demands, including acute coronary syndrome and chronic coronary syndrome, as described in a previous study 19 . Cerebrovascular disease (CVD) is mainly caused by intracranial blood circulation disorders caused by brain vessels, such as stoke and cerebral infarction, as previously described 20 . Hypertension (HTN) was defined as the SBP greater than 130 mmHg or the DBP greater than 80 mmHg. Statistical analysis. Continuous variables are presented as the mean ± standard deviation, and discrete variables are described as percentages. For continuous variables, one-way ANOVA or the Kruskal-Wallis H test were used to compare the differences among groups. For discrete variables, we used the Chi-square test to compare groups. Multivariate regression analysis was used to estimate the association between DFCs and clinical variables. The variables selected for the multivariate regression analysis were based on the statistical significance of univariate regression analysis, including sex, DBP, durations, smoking, BMI, FHD, Hb levels, PLT levels, LDL-C levels, HDL-C levels, TC levels, Alb levels, HOMA-IR, eGFR, lipid-lowering agents, OHA and insulin (see supporting information Table 3), whereas disease onset age and HbA 1c and DPN were forcibly selected for the multivariate regression analysis. All variables in the regression analysis were described as positive or negative based on a cut-off number, which is described in detail in the supplementary materials (see supporting www.nature.com/scientificreports/ information Table 2). The analysis was performed by SPSS 25.0, and a two-tailed P value < 0.05 was considered statistically significant.

Results
The clinical characteristics of the participants. A total of 8909 participants diagnosed with T2DM were recruited for the study. The mean age of all patients was 57.89 ± 12.2 years, and the mean disease onset age was 49.33 ± 11.37 years. Men accounted for 55.4% of the sample. A total of 2804 (31.5%) patients had firstdegree family members with diabetes. Among them, the patients with one and more than one first-degree family member with diabetes accounted for 18.7% and 12.8%, respectively. The proportions of the participants with a father with diabetes, a mother with diabetes, both parents with diabetes, siblings with diabetes, a father and siblings with diabetes, a mother and siblings with diabetes, and both parents and siblings with diabetes were 3.5%, 6.2%, 1.1%, 14.4%, 1.5%, 4%, and 0.7%, respectively. The mean concentrations of FBG and HbA 1c were 8.60 ± 4.02 mmol/l and 9.06% ± 2.39%, respectively. The patients with HOMA-IR ≥ 1 accounted for 48.4%. Moreover, among the diabetes-related complications, nearly half of the subjects had DPN (50.7%), and only a small proportion of patients had DFCs (7.7%). Additionally, regarding the treatments, a large proportion of people took lipid-lowering agents (71.3%), OHA (70.9%) and insulin (70.2%). Only 17.5% and 23.1% of participants took ACEIs and ARBs, respectively (see Table 1).
Comparison of the clinical characteristics among the groups based on different numbers of first-degree relatives with diabetes. Based on the number of first-degree relatives with diabetes, the study was divided into 3 groups as follows: "No FHD" group referred to patients with no first-degree relatives with diabetes; the "One member with FHD" group referred to patients with only one first-degree relative with diabetes; the " ≥ 2 members with FHD" group referred to patients with more than 2 first-degree relatives with diabetes. The following clinical characteristics were significantly different across the 3 groups: age, the proportion of men, DBP, the durations of diabetes, disease onset age, smoking and drinking habit, Hb levels, FBG, HbA 1c levels, Alb levels, TC levels, proteinuria, eGFR, HOMA2-IR and HOMA2-B, the prevalence of DR, DPN, DF, diabetic ketosis (DK), CHD, CVD, HTN, the use of lipid-lowering agents and OHA (P < 0.05). Other parameters, including BMI, WHR, SBP, PLT levels, LDL-C levels, HDL-C levels, uric acid levels, the prevalence of DN, and the use of ACEI, ARB and insulin, were not significantly different across groups (P ≥ 0.05) (see Table 2).
Comparison of the clinical characteristics among the groups based on different types of first-degree relatives with diabetes. When comparing the clinical characteristics of the 8 groups, we found that the following parameters showed significant differences: age, the proportion of men, SBP, durations of the disease, disease onset age, smoking and drinking, Hb levels, FBG levels, HbA 1C levels, Alb levels, TC levels, LDL-C levels, proteinuria, uric acid levels, eGFR, HOMA2-B, DR, DPN, DN, DK, CHD, CVD, HTN, ARB, lipid-lowering agents, OHA, and insulin (P < 0.05). Other variables, including WHR, PLT levels, HDL-C levels, HOMA2-IR, DFCs and the use of ACEI, did not differ significantly across groups (P ≥ 0.05) (see Table 3). When the different types of first-degree relatives with diabetes was entered into the analysis instead of the number of first-degree relatives with diabetes, the results revealed that only having a mother with a history diabetes was positively associated with diabetes (OR = 1.484, 95%CI 1.022-2.154); other types, including a father, siblings, siblings and a mother, siblings and a father, both parents and siblings, both parents, displayed no significance (P > 0.05). Regarding other variables, such as BMI, Hb levels, PLT levels, DPN, and OHA, similar results were observed with the abovementioned variables in Table 4 (see Table 5).

Discussion
In this study, we found that DFCs among T2DM patients were associated with FHD, including both the number and type of first-degree family members with diabetes. However, among the different types of first-degree family members with diabetes, only the patients with a mother with a history of diabetes showed a statistical association with DFCs. In addition to FHD, other factors, including gender, BMI, PLT levels, Hb levels, Alb levels, HDL-C levels, DPN, and the use of lipid-lowering agents, OHA, and insulin, were also associated with DFCs.
The management of diabetes is an important part of public health policy. However, it is not going well, especially the management of diabetic complications, such as DFCs 21,22 . A previous study reported that the overall  [23][24][25] . In our study, we observed a strong relationship between FHD and DFCs, especially of the different numbers of relatives with diabetes, in which the number of patients with one and more than 1 relative with diabetes led to 1.377-and 1.402-fold increases in the risk of DFCs, respectively, compared with those without FHD. FHD has been considered a reflection of both genetic effects and environmental effects 26 . Regarding the genetic effects, with the rapid advancement in currently emerging genetic methods, such as genome-wide association studies (GWAS), a wide variety of candidate genes associated with T2DM have been identified 27 . The genetic risk score (GRS) has been reported to be associated with FHD 28,29 . A study investigated whether the GRS value was associated with parental numbers with FHD, indicating an important role of genes and FHD 28 . Similarly, genetic factors may be vital in DFCs, and diabetes-related complications have been reported to have a phenomenon of family clustering in type 1 diabetes 30 and type 2 diabetes 31 . Genetic and ethnic factors play an important role in not only DPN but also DFCs 32 . Several studies have identified that some genetic variants may be potential risk factors for DFCs, such as the polymorphisms of MAPK14 33 , TNFRSF11B 34 , MCP1 35 , VEGF 36,37 , TNF-α 38 , and osteoprotegerin gene 39 . In addition, one genome-wide association study conducted by Meng et al. 33 , which included 699 diabetic foot ulcers and 2695 controls, indicated that the single-nucleotide polymorphism rs80028505 of MAPK14 was a risk factor for diabetic foot ulcers. Another study aimed to investigate the relationship between VEGF gene polymorphisms and DFCs in a Chinese population showed that the VEGF gene could be a susceptible gene of DFCs 36 . Another study found that polymorphisms in cytokine/chemokine genes had a close association with amputation and severe infection in individuals with DFCs 38 . Therefore, we conjectured from the abovementioned studies that the higher prevalence of DFCs in patients with FHD may be due to genetic impact.
A study to investigate the association between different relatives with diabetes and the prevalence of T1DM, LADA and T2DM found that any first-degree member with diabetes, such as a father with diabetes, a mother with diabetes, or siblings with diabetes, would increase the prevalence of LADA and T2DM 14 . When we explored the relationship between different types of FHD and DFCs, the results suggested that only having a mother with a history of diabetes was associated with DFCs after adjusting for some usual confounders of DFCs; other types of diabetes showed no statistical association. The mechanism remains unknown. Some studies observed maternal transmission in T2DM 40,41 . The hypotheses attributed the maternal transmission to the distinctly maternal genetic and environmental impact 42 . Thus, we speculated that maternal transmission likely existed in DFC individuals and that the high prevalence of DFCs in patients with a mother history of diabetes may be caused by maternal transmission. Another possible explanation was that environmental factors, such as smoking and drinking 23,43,44 and BMI 45 , also contributed to DFCs. In our study, participants in mother history seemingly had unhealthy lifestyles with worse weight management and glycemic control and a higher percentage of smokers and drinkers (Table 3). Additionally, family lifestyles, including eating habits, are mainly derived from mothers, and mothers' history of diabetes may be easily inclined to live with unhealthy lifestyles, leading to a higher prevalence of DFCs.
Our study found that male gender was a risk factor for DFCs. A study aimed to investigate 88 potential risk factors for amputation in DFC patients implied that male gender increased the likelihood of amputation by eightfold compared to female gender 46 . On the one hand, males tend to have higher foot pressure and a higher prevalence of peripheral insensate neuropathy, which may contribute to DFCs 47,48 . On the other hand, females may be inclined to be more self-managed and self-caring and show an active attitude towards foot care; however, males are more likely to express fear, inactive attitudes and an uncooperative manner in behavior 49,50 .

Scientific Reports
| (2020) 10:17056 | https://doi.org/10.1038/s41598-020-74071-3 www.nature.com/scientificreports/ www.nature.com/scientificreports/   www.nature.com/scientificreports/ prothrombotic state may result in atherogenesis attributable to the impaired endothelium or ruptured plaque by abnormal adhesion and aggregation of PLT levels 52 . Furthermore, inflammation has an association with endothelial injury and regulates platelet action-associated proteins 53 . Atherogenesis and inflammation may contribute to the occurrence of PAD, which has been demonstrated to be a risk factor for DFCs 7,54 . The relationship between BMI and DFCs remains controversial 23,46,55 . On the one hand, a study investigated whether BMI had no association with diabetic foot ulcers 23 . On the other hand, another study showed a positive association between BMI and DFCs 8 , contrary to our results. This may be caused by racial differences. Dyslipidemia, such as the decreased level of HDL and increased level of triglycerides, has been proven to be related to DFCs 44,56 , consistent with our study. Due to the negative association between DFCs and HDL, it was easy to comprehend the higher percentage of lipid-lowering agent use in DFC patients. Moreover, one multicentric cross-sectional study revealed that insulin was positively associated with DFCs 57 , consistent with our findings that DFC patients had a higher proportion of insulin.
Several limitations were observed in the study. First, it was not clear whether the first-degree family members with diabetes had type 1 diabetes or type 2 diabetes. Second, this was a cross-sectional and single-center observation, and more prospective and multicenter studies should be conducted.
In conclusion, DFCs were associated with different numbers and types of family members with diabetes. This association reveals the importance of genetic and environmental factors in DFCs and highlights the importance of adding FHD to public health strategies targeting detecting and preventing the disease.  www.nature.com/scientificreports/