PDF
Citation
Print
INTRODUCTION
Diabetic foot is a chronic diabetic complication accompanied by peripheral neuropathy and vasculopathy due to complex metabolic pathways [1-3]. The incidence of diabetic foot was increased along with the incremental diabetes mellitus rate and the prolonged life expectancy of patients [2]. A lifetime incidence of foot ulceration in diabetic patients is 15% to 25% [4,5], and diabetic ulcer leads to poor quality of life [3,6,7] and high mortality [8,9]. Unfortunately, 28% of diabetic foot ulcers result in lower extremity amputation (LEA) [10]. Five-year mortality rates have been estimated to be 42% to 44% after the onset of primary diabetic ulceration and increase to 68% to 79% after amputation [11,12].
In the USA, the average annual cost of diabetic foot is more than $8,500 per patient [13], and the medical cost for treating diabetic foot ranges from $9 to $13 billion more that the cost of diabetes itself [14]. A total cost for an LEA ranges from $43,800 (minor amputation, foot level) to $66,215 (major amputation, above ankle), and 77% of the costs are incurred after the amputation [15]. Diabetic foot incurs considerable social and medical burdens as well as economic costs for patients [16].
Various risk factors for LEA of age, sex, peripheral neuropathy, peripheral artery disease, glycated hemoglobin, hypertension, nephropathy, duration of diabetes, and obesity have been revealed [17-19] to affect a diabetic ulcer, its exacerbation, and the survival rate of diabetic patients.
The present study analyzed cohort data from the National Health Insurance Service (NHIS) in Republic of Korea to reveal whether modifiable behavioral factors of current cigarette smoking, heavy alcohol consumption, and regular exercise are associated with risk of LEA in patients with diabetic foot.
Go to : 
METHODS
Data source and study population
This study used the database of the NHIS, which enrolls approximately 97% of the Korean population through a biennial medical examination sponsored by the Korean government [20]. NHIS data can be used by researchers with official review board approval of the study protocol, as was received here.
Among 2,746,078 type 2 diabetic patients assigned to the NHIS between the years 2009 and 2012, the medical records of 2,745,688 individuals aged 20 years and over were evaluated until December 31, 2018. The presence of type 2 diabetes mellitus was defined according to the presence of at least one claim per year for prescription of diabetic medication under International Classification of Diseases 10th Revision (ICD-10) codes E11–E14 or fasting blood glucose ≥126 mg/dL. The primary endpoint of this study was newly amputated lower extremities, which were defined with the ICD-10 codes (N0571-5, N0562, N0564, N0565, and N0566). Patients with missing data, previous LEA, and 1 year lag were excluded. Fig. 1 shows a flow of the selection of the study population.
Data collection
Standardized self-reporting questionnaires were used to collect data to establish a baseline for the following variables regarded as risk factors for LEA and were included as covariates in the multivariate analyses: age (years); sex; cigarette smoking (never, former, and current); alcohol consumption (non, light, and heavy); level of physical activity; yearly income (lower quintile vs. the remaining quintiles); body mass index (BMI); diagnosis of hypertension, dyslipidemia, or chronic kidney disease (CKD); medication with insulin or oral antidiabetic drugs; and duration of diabetes. Systolic blood pressure and diastolic blood pressure were also obtained (mm Hg). Values for serum glucose (mg/dL), white blood cell (WBC) count (100/µL), total cholesterol, high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C), and triglycerides (mg/dL) were also measured.
In this study, the criteria for classifying alcohol consumption were defined on the basis of the daily pure alcohol intake on 360 mL of soju (Korean distilled liquor) with an alcohol concentration of 20%. The study population was divided into three groups: non-drinkers, light drinkers (alcohol consumption less than 30 g per day), and heavy drinkers (30 g or more grams per day).
The diagnosis of hypertension was defined as the presence of at least 1 claim per year for antihypertensive prescription under ICD-10 codes I10–I13 and I15 or systolic/diastolic blood pressure ≥140/90 mm Hg. The diagnosis of dyslipidemia was defined according to the presence of at least one claim per year for antihyperlipidemic prescription under ICD-10 code E78 or a total cholesterol ≥6.21 mmol/L. The presence of CKD was defined as an estimated glomerular filtration rate <60 mL/min/1.73 m2 calculated using the CKD Epidemiology Collaboration equation [21,22].
Informed consent was not obtained for this study because the study included routinely collected data. The study was approved by the Institutional Review Board of The Catholic Medical Center Office of Human Research Protection Program (SC21ZIS-E0165).
Statistical analyses
Statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA) and R version 3.2.3 (The R Foundation for Statistical Computing, Vienna, Austria, http://www.Rproject.org). All data are expressed as mean±standard deviation for normally distributed continuous variables and as proportion for categorical variables. The t test and analysis of variance (ANOVA) were used to analyze continuous variables, and the differences between nominal variables were compared with the chi-square test. The incidence rate of LEA in diabetic feet was calculated by dividing the number of events by person-time at risk. To determine the independent association of current cigarette smoking, heavy alcohol consumption, and regular exercise with risk of LEA in diabetic feet, the Cox regression model was used after adjusting for age, sex, smoking, alcohol consumption, regular exercise, low income, hypertension, dyslipidemia, BMI, using insulin or oral antidiabetic drugs, and diabetes duration. The cumulative incidence of LEA was plotted on a Kaplan-Meier curve. A two-sided P value <0.05 indicated a statistically significant difference.
Go to :
RESULTS
Demographic characteristics
Table 1 shows the demographics of the study population according to LEA. Among a total of 2,644,440 diabetic patients, 0.33% (n=8,778) patients experienced LEA. The amputated patient group was generally older and included more men and more individuals who currently smoked; consumed alcohol heavily; did not exercise regularly; had limited income; had low BMI; were diagnosed with hypertension, dyslipidemia, and/or CKD; were treated with insulin or more than three oral antidiabetic drugs; and had higher serum glucose level/WBC count/systolic blood pressure/triglycerides and low diastolic blood pressure/high-density lipoprotein/low-density lipoprotein than patients without LEA.
Table 1.
Demographics of the Diabetic Foot Patients in the NHIS Cohort (2009 to 2012)
Values are expressed as mean±standard deviation, number (%), or median (interquartile range).
NHIS, National Health Insurance Service; BMI, body mass index; HP, hypertension; DYS, dyslipidemia; CKD, chronic kidney disease; WBC, white blood cell count; SBP, systolic blood pressure; DBP, diastolic blood pressure; HLD-C, high density lipoprotein cholesterol; LDL-C, low density lipoprotein cholesterol; TG, triglyceride; CI, confidence interval.
Risk of lower extremity amputation in diabetic foot according to current cigarette smoking, heavy alcohol consumption, and regular exercise
A multivariate analysis was performed with or without adjustment for age, sex, smoking, alcohol consumption, exercise, limited income, hypertension, dyslipidemia, BMI, use of insulin or oral antidiabetic drugs, and diabetes duration (Table 2). The risk of LEA increased with current cigarette smoking and heavy alcohol consumption (hazard ratio [HR], 1.436; 95% confidence interval [CI], 1.367 to 1.508 and HR, 1.082; 95% CI, 1.011 to 1.158), but it decreased with regular exercise (HR, 0.745; 95% CI, 0.706 to 0.786) after adjusting for the above variables.
Table 2.
Risk Factors and Hazard Ratios of the Risk of Amputation in Patients with Diabetic Foot
| Variable | Number | Event | Duration | Incidence rate, /1,000 person-years |
Hazard ratio (95% CI) |
|
|---|---|---|---|---|---|---|
| Non-adjusted | Adjusteda | |||||
| Current cigarette smoking | ||||||
| No | 1,961,924 | 5,802 | 13,512,000 | 0.42940 | 1 (Reference) | 1 (Reference) |
| Yes | 682,516 | 2,976 | 4,630,204 | 0.64274 | 1.503 (1.438–1.571)b | 1.436 (1.367–1.508)b |
| Heavy alcohol consumption | ||||||
| No | 2,381,993 | 7,767 | 16,346,859 | 0.47514 | 1 (Reference) | 1 (Reference) |
| Yes | 262,447 | 1,011 | 1,795,345 | 0.56312 | 1.187 (1.112–1.268)b | 1.082 (1.011–1.158)b |
| Regular exercise | ||||||
| No | 2,096,400 | 7,126 | 14,317,538 | 0.49771 | 1 (Reference) | 1 (Reference) |
| Yes | 548,040 | 1,652 | 3,824,666 | 0.43193 | 0.865 (0.820–0.912)b | 0.745 (0.706–0.786)b |
| Scorec | ||||||
| 0 | 389,891 | 1,052 | 2,734,435 | 0.38472 | 1 (Reference) | 1 (Reference) |
| 1 | 1,583,446 | 4,859 | 10,856,128 | 0.44758 | 1.168 (1.092–1.248)b | 1.366 (1.277–1.462)b |
| 2 | 555,392 | 2,347 | 3,767,964 | 0.62288 | 1.631 (1.516–1.754)b | 1.752 (1.562–1.966)b |
| 3 | 115,711 | 520 | 783,677 | 0.66354 | 1.739 (1.566–1.932)b | 2.448 (1.998–3.001)b |
A Kaplan-Meier curve also showed that a cumulative incidence of LEA in diabetic foot was heightened by current cigarette smoking and heavy alcohol consumption but was lowered by regular exercise (Fig. 2).
We assigned zero points if all three lifestyle habits were healthy, such as smoking cessation, no alcohol consumption, and regular exercise, adding one point for each worsening lifestyle choice and assigning three points for current smoking, heavy drinking, and not exercising. After adjusting for the above-mentioned variables, the risk of LEA increased with every one-point increase (Score 1: HR, 1.366; 95% CI, 1.277 to 1.462; Score 2: HR, 1.752; 95% CI, 1.562 to 1.966; and Score 3: HR, 2.448; 95% CI, 1.998 to 3.001).
The effect of current cigarette smoking, heavy alcohol consumption, and regular exercise on amputation risk of diabetic foot according to amputation level
A multivariate analysis was performed according to the severity of amputation on diabetic foot. The severity of amputation was separated to minor amputation, foot level and major amputation, above ankle (Table 3). In the major amputation, the risk of LEA increased with current cigarette smoking (HR, 1.683; 95% CI, 1.487 to 1.905), but it decreased with regular exercise (HR, 0.638; 95% CI, 0.551 to 0.738) after adjusting for the above variables. However, correlation between heavy alcohol consumption and risk of LEA was not significant (HR, 1.012; 95% CI, 0.849 to 1.207; P=0.8946). Additionally, in the minor amputation, the risk of LEA increased with current cigarette smoking and heavy alcohol consumption (HR, 1.394; 95% CI, 1.321 to 1.471 and HR, 1.095; 95% CI, 1.018 to 1.179), but it decreased with regular exercise (HR, 0.765; 95% CI, 0.722 to 0.811) after adjusting for the above variables. The above-mentioned scoring calculation, the risk of LEA increased with every one-point increase in both major and minor amputation.
Table 3.
Risk Factors and Hazard Ratios of the Risk of Lower Extremity Amputation in Diabetic Foot according to Amputation Level
| Variable |
Number |
Event |
Hazard ratio (95% CI) |
|||||
|---|---|---|---|---|---|---|---|---|
| Major | Minor | Major | Minor |
Non-adjusted |
Adjusteda |
|||
| Major | Minor | Major | Minor | |||||
| Current cigarette smoking | ||||||||
| No | 1,961,924 | 1,961,924 | 845 | 4,957 | 1 (Reference) | 1 (Reference) | 1 (Reference) | 1 (Reference) |
| Yes | 682,516 | 682,516 | 495 | 2,481 | 1.717 (1.537–1.918)b | 1.467 (1.398–1.539)b | 1.683 (1.487–1.905)b | 1.394 (1.321–1.471)b |
| Heavy alcohol consumption | ||||||||
| No | 2,381,993 | 2,381,993 | 1,191 | 6,576 | 1 (Reference) | 1 (Reference) | 1 (Reference) | 1 (Reference) |
| Yes | 262,447 | 262,447 | 149 | 862 | 1.141 (0.962–1.353) | 1.196 (1.114–1.284)b | 1.012 (0.849–1.207) | 1.095 (1.018–1.179)b |
| Regular exercise | ||||||||
| No | 2,096,400 | 2,096,400 | 1,121 | 6,005 | 1 (Reference) | 1 (Reference) | 1 (Reference) | 1 (Reference) |
| Yes | 548,040 | 548,040 | 219 | 1,433 | 0.729 (0.631–0.843)b | 0.89 (0.840–0.943)b | 0.638 (0.551–0.738)b | 0.765 (0.722–0.811)b |
| Scorec | ||||||||
| 0 | 389,891 | 389,891 | 138 | 914 | 1 (Reference) | 1 (Reference) | 1 (Reference) | 1 (Reference) |
| 1 | 1,583,446 | 1,583,446 | 724 | 4,135 | 1.325 (1.105–1.59)b | 1.144 (1.065–1.229)b | 1.55 (1.289–1.864)b | 1.338 (1.244–1.439)b |
| 2 | 555,392 | 555,392 | 393 | 1,954 | 2.08 (1.713–2.525)b | 1.563 (1.445–1.69)b | 2.428 (1.785–3.304)b | 1.655 (1.462–1.874)b |
| 3 | 115,711 | 115,711 | 85 | 435 | 2.165 (1.653–2.837)b | 1.675 (1.494–1.877)b | 4.287 (2.481–7.408)b | 2.226 (1.788–2.772)b |
The impacts of current cigarette smoking, heavy alcohol consumption, and regular exercise on risk of amputation in diabetic foot according to age, diabetic duration, and use of more than three oral diabetic drugs
The HRs of LEA by current cigarette smoking, heavy alcohol consumption, and regular exercise for age, diabetic duration, and use of more than three oral diabetic drugs are presented in Table 4. The risks of LEA increased more in older diabetic patients (65 years and older) than in younger diabetic patients (under 65 years) in those who were smoking and drinking heavily. On the other hand, the risk of LEA was reduced more in older diabetics (65 years and older) than in younger diabetics (under 65 years) in those who exercised regularly.
Table 4.
Hazard Ratio for Risk of Amputation in Diabetic Foot according to Age, Diabetes Duration, and Oral Antidiabetic Drugs Stratified by Cigarette Smoking, Heavy Alcohol Consumption, and Regular Exercise
| Subgroup |
Current cigarette smoking |
Heavy alcohol consumption |
Regular exercise |
|||
|---|---|---|---|---|---|---|
| Group | Hazard ratio (95% CI)a | Group | Hazard ratio (95% CI)a | Group | Hazard ratio (95% CI)a | |
| Age, yr | ||||||
| <65 | No | 1 (Reference) | No | 1 (Reference) | No | 1 (Reference) |
| Yes | 1.407 (1.323–1.496) | Yes | 1.053 (0.971–1.141) | Yes | 0.768 (0.716–0.824) | |
| ≥65 | No | 1 (Reference) | No | 1 (Reference) | No | 1 (Reference) |
| Yes | 1.522 (1.402–1.652) | Yes | 1.185 (1.047–1.341) | Yes | 0.702 (0.645–0.764) | |
| Duration over 5 years | ||||||
| No | No | 1 (Reference) | No | 1 (Reference) | No | 1 (Reference) |
| Yes | 1.722 (1.587–1.869) | Yes | 1.131 (1.020–1.255) | Yes | 0.792 (0.720–0.871) | |
| Yes | No | 1 (Reference) | No | 1 (Reference) | No | 1 (Reference) |
| Yes | 1.29 (1.212–1.372) | Yes | 1.038 (0.948–1.135) | Yes | 0.719 (0.674–0.768) | |
| More than three oral antidiabetic drugs | ||||||
| No | No | 1 (Reference) | No | 1 (Reference) | No | 1 (Reference) |
| Yes | 1.446 (1.370–1.526) | Yes | 1.091 (1.014–1.174) | Yes | 0.747 (0.704–0.792) | |
| Yes | No | 1 (Reference) | No | 1 (Reference) | No | 1 (Reference) |
| Yes | 1.384 (1.225–1.564) | Yes | 1.050 (0.880–1.254) | Yes | 0.728 (0.634–0.835) | |
In terms of diabetic duration, current cigarette smoking and heavy alcohol consumption increased the risk of LEA, and the increase was greater in patients with a diabetes duration of 5 or fewer years compared to patients with duration longer than 5 years. However, the risk was reduced by a larger amount in patients with a duration longer than 5 years when they exercised regularly compared to such patients with a duration of 5 or fewer years.
The same pattern as in the duration of diabetes was seen in the use of more than three oral diabetic drugs. Current cigarette smoking and heavy alcohol consumption increased the risk of LEA more in patients using three or fewer oral diabetic drugs than in individuals treated with more than three oral diabetic drugs. However, regular exercise showed a greater reduction in the risk in patients using three or fewer oral diabetic drugs compared to patients treated with more than three oral diabetic drugs.
Go to :
DISCUSSION
Diabetic foot is a challenging and chronic complication that affects nearly 6% of patients with diabetes mellitus accompanied by peripheral neuropathy and peripheral arterial disease in the lower extremity due to complex metabolic pathways [1-3]. Its incidence has increased along with the growing prevalence of diabetes mellitus and the extended patient’s lifespan [2].
An incidence of foot ulcer is up to 15% to 25% in diabetic patient’s whole life [4,5]. The worldwide prevalence of diabetic ulcer is 6.3% [2], is highest in North America at 15% [23,24], and relatively low in Asia at 5.4% [25,26]. Diabetic ulcer impairs patients’ quality of life [6,7], disturbs their social activity and existence [3], and contributes to high mortality [8,9]. The risk of death at 10 years after onset of diabetic ulcer is twice as high as that of those without diabetic ulcer [27]. Furthermore, 5-year mortality rates have been estimated to be 42% to 44% since primary diabetic ulceration and increase to 68% to 79% after amputation [11,12].
Diabetic ulcer usually develops in body areas subjected to repeated vertical or shear stress in a patient with peripheral neuropathy or peripheral artery disease [17,28]. Of the two, peripheral arterial disease is more likely to show diabetic ulcers progressing to nonhealing, infection, and even amputation [29].
About 40% of diabetic ulcers recur within 1 year after healing, 60% within 3 years, and 65% within 5 years [9]. One-third of diabetic foot ulcers result in LEA [10], and the rest heal conservatively [30]. Diabetes-related LEA rates are 0.04% to 1.5% [3,31]. In this study, the LEA rate of diabetic patients in Korea was 0.33%, but the proportion of diabetic feet among diabetic patients was unknown. In addition, although LEA, which is unrelated to diabetes, might be present in this study population, the effect on the statistical significance of the results would be negligible because the risk of an LEA with diabetes has been estimated at 23 times that without diabetes [32].
Various risk factors for LEA, including age, sex, peripheral neuropathy, peripheral artery disease, glycated hemoglobin, hypertension, nephropathy, longer diabetes duration, and obesity, have been shown to affect [17-19] early diagnosis, healing of a diabetic ulcer, exacerbation including infection and LEA, and survival rate of diabetic patients. As Asian data about diabetic foot diseases are scarce [16], the present study analyzed whether modifiable behavioral factors of current cigarette smoking, heavy alcohol consumption, and regular exercise are associated with risk of LEA in Korean patients with diabetic foot.
Meta-analysis including five cohort studies and three case control studies demonstrated that cigarette smoking significantly increased the risk of LEA (odd ratio, 1.65; P<0.0001) without publication bias [33]. Smoking cessation also reduces the risk of LEA in diabetic foot, consistent with the results of the present study that current smoking increases the risk of LEA. Current smoking is the most potent but avoidable risk factor associated with cardiovascular, cerebrovascular, and peripheral vascular disease [34]. Its pathogenesis is reduced capacity to transport oxygen in the blood due to harmful by-products of cigarette smoking and results in tissue hypoxia and arteriospasm [35]. This damage leads to compensatory erythrocytosis that increases blood viscosity and decreases tissue perfusion [36]. Decreased tissue perfusion and oxygenation inhibit healing of diabetic ulcer, which can increase the risk of LEA.
Chronic and even regular moderate consumption of alcohol with diabetes results in hyperglycemia and peripheral neuropathy, contributing to development of diabetic ulcer, which can increase the risk of LEA [37]. This supports the results of the present study that heavy alcohol consumption enhanced the risk of LEA. Chronic alcohol consumption in diabetes tends to accompany poor compliance with diet and medication, which hampers glycemic control [38].
Six controlled clinical trials suggested that physical activity and exercise improved diabetic foot outcomes and prevented complications including diabetic ulcer, infection, and LEA [39]. The present study also indicated that regular exercise decreased the risk of LEA because physical activity and exercise in diabetic patients has improved glycemic control, nerve velocity conduction, and gait function [40,41]. Regular exercise in type 2 diabetic patients has resulted in positive glycemic control, which is effective in preventing diabetic peripheral neuropathy, a key risk factor of diabetic ulcers [42]. Exercise in diabetes patients increases both motor and sensory nerve velocity conduction of the lower limbs, with reduced distal latency [43]. It delays diabetic peripheral neuropathy and improves skin sensitivity and the vibration perception threshold [43,44]. In addition, diabetic patients who exercise regularly verified improved balance in one-legged posture; better foot rollover with redistribution of dynamic plantar loading; and improvement in quality of life [40,44]. Also, weight-bearing from physical activity is not associated with risk of diabetic foot re-ulceration [45].
The combination of smoking and chronic intake of alcohol in diabetic patients has demonstrated a higher rate of LEA and longer hospital stays [38]. The present study also indicated that addition of unhealthy behaviors increased the risk of LEA, with the highest risk for combined lack of exercise, current smoking, and heavy alcohol intake. A synergistically increased risk of LEA was observed with addition of each unhealthy behavior.
Diabetic foot occurs more frequently in male patients than in female patients [2]. The gender difference might be due to greater physical exertion in males [46]. This study also found a higher risk of LEA in males. However, vulnerability associated with current cigarette smoking and the beneficial effects of regular exercise showed a great impact on the risk of LEA in females.
The association between BMI and risk of diabetic foot ulceration is controversial [47,48] Recent studies revealed that higher risk of diabetic ulcer is related to a J-shaped BMI curve (BMI <25 kg/m2 and BMI ≥45 kg/m2) [49] or a specific range (BMI ranging from 25 to 30 kg/m2) [2]. In the present study, low BMI (BMI <23 kg/m2) and higher risk of LEA were associated in the Korean population. In addition, factors that enhanced the risk of LEA included older age; longer duration of diabetes; nephropathy; hypertension; and high levels of WBC, triglycerides, LDL-C, HDL-C, and serum C-reactive protein [50].
In the present study, older diabetic patients (65 years and older) showed higher risk of LEA with cigarette smoking and heavy alcohol intake than did younger patients, but the difference decreased with regular exercise. Patients with shorter diabetes duration (under 5 years) and a combination of three or fewer oral diabetic drugs were more susceptible to increased risk of LEA with smoking and heavy alcohol intake. In patients with diabetes duration longer than 5 years and use of more than three oral diabetic drugs, regular exercise was associated with a lower risk of LEA.
In conclusion, this nationwide population-based Korean study revealed that the risk of LEA in diabetic patients was increased by current cigarette smoking and heavy alcohol consumption but was reduced by regular exercise. A synergistically increased risk of LEA was observed with addition of each unhealthy behavior. Modification of behaviors including current smoking, heavy alcohol intake, and frequency of exercise is likely to decrease the incidence of LEA and improve the physical, emotional, and social quality of life in diabetic patients.
Go to :
XML Download