Factors associated with lower‐extremity amputation in patients with diabetic foot ulcers in a Chinese tertiary care hospital

Abstract

Providing a better understanding of the risk factors for amputation in this particular region, Hunan province, in China might help patients with diabetic foot ulcers receive timely and appropriate medical care and help prevent amputation. Diabetic foot ulcer patients referred to the Third Xiangya Hospital during the period between December 2014 and September 2018 were enrolled. Participants who underwent amputations and received conservative treatments were compared using univariate and multivariate analyses to identify the independent predictors of amputation. Those who required amputation presented significantly higher levels of white blood cell counts, platelet counts, erythrocyte sedimentation rate, C‐reactive protein, and glycated haemoglobin (HbA1c) levels. However, levels of haemoglobin, postprandial plasma C‐peptide, triglyceride, high‐density lipoprotein cholesterol, albumin, and uric acid were decreased in patients with amputations. Patients with more advanced Wagner grades had much higher rates of amputation. Multivariable‐adjusted odds ratios in stepwise logistic regression model was 1.317 for HbA1c (95% CI: 1.015‐1.709), 0.255 for triglyceride (95% CI: 0.067‐0.975), and 20.947 for Wagner grades (95% CI: 4.216‐104.080). Independent risk factors for amputation in these Chinese diabetic foot ulcer patients included an elevated HbA1c level, lower triglyceride level, and higher Wagner grades.

1. INTRODUCTION

The number of diabetics is expected to rise to 592 million by 2035 worldwide.1 Diabetic patients are more susceptible to peripheral arterial disease (PAD) and foot ulceration compared with non‐diabetic individuals. About 15% to 25% of diabetic patients will develop foot ulceration during their lifetime, and among all causes of non‐traumatic limb amputation in both developed and developing countries, diabetes is the most common aetiology.2, 3, 4 The devastating consequences of diabetic foot diseases can become a major public health problem and impose a heavy burden on the Chinese health care system, where resources are variably limited in different regions, especially in rural areas.

Diabetic foot ulcers (DFU) might result from a variety of causes, including peripheral neuropathy, PAD, and immunosuppression. The prevalence of PAD is 10% to 26% in the general population, while diabetics with foot ulcers are usually at more than 50% risk of developing PAD, which leads to delayed wound healing and even amputation.5, 6, 7 Previous reports have shown that diabetes‐induced amputations have a high risk of mortality, with a 5‐year survival rate of 40% to 48%.8, 9, 10 According to a recent report of Cardoso et al, the bacterial genera Acinetobacter spp. and Klebsiella spp. identified in the infected ulcers of DFU patients were associated with a much higher risk of major amputation.11 Baumfeld et al found that risk factors for major amputations were ischaemia and previous amputations history, and antibiotic therapy before operations appeared to be a protective factor.12 Khalfallah et al reported that age was the only independent predictor for major lower‐extremity amputation (LEA) in DFU patients with a threshold age value of no less than 65 years in their study.13 In another study by Shatnawi et al, the presence of gangrenous tissues and uncontrolled blood glucose levels were significant predictive factors for type 2 diabetes‐related major LEA.14 The risk factors for LEA associated with diabetic foot diseases are extensive, including microvascular diseases, PAD, infections, long diabetes duration, poor glycaemic control, old age, and cardiovascular comorbidities.15, 16, 17, 18, 19, 20 Variations in these reports may be because of differences in study design and participant features, severities of DFU, standards and levels of medical care, and accesses to appropriate treatments.

In 2013, a report published in JAMA indicated that diabetes is a severe public health problem in China, with up to 113.9 million Chinese adults suffering from diabetes and 493.4 million remaining in pre‐diabetic status.21 Even though the population of patients with DFU is increasing dramatically in China, only limited studies were focused on the prevalence and predictors of LEA in Chinese diabetic patients. As China is a developing country with a vast territory, many aspects, including natural condition, lifestyle, and economic development levels, differ immensely for different regions. In 2011, Li et al showed that risk factors for amputation in Chinese DFU patients included presence of peripheral vascular diseases (PVD), elevated WBC, sCRP, and lower lipid level.22 Criteria for amputations might differ among different Chinese health care services. We retrospectively examined patients with DFU in the central and southern regions of China during the period between December 2014 and September 2018 to investigate the potential risk factors predicting amputations in these patients. A better understanding of the risk factors for amputation will help DFU patients receive early appropriate medical care and help prevent amputation.

2. PATIENTS AND METHODS

Approval for this retrospective study was obtained from the Third Xiangya Hospital Review Board. This study was based on the medical records of DFU patients admitted to the Third Xiangya Hospital during the period between December 2014 and September 2018. Patients were divided into an amputation group, including major and minor amputations, and non‐amputation group. Only LEAs that occurred during hospital admission were registered.

Clinical data of all the enrolled 475 participants were collected during admission. Age, diabetes duration, systolic blood pressure (SBP), diastolic blood pressure (DBP), body mass index (BMI), waist‐hip ratio (WHR), white blood cell (WBC) counts, haemoglobin (Hb), platelet (PLT) counts, erythrocyte sedimentation rate (ESR), glycated haemoglobin (HbA1c) levels, fasting/postprandial glucose levels, fasting/postprandial plasma C‐peptide levels, total cholesterol, low‐density lipoprotein cholesterol (LDL‐C), high‐density lipoprotein cholesterol (HDL‐C), triglyceride, albumin, alanine aminotransferase (ALT), blood urea nitrogen (BUN), creatinine, uric acid (UA), parathyroid hormone (PTH), 25(OH)VitD, and brain natriuretic peptide (BNP) were recorded as continuous variables. Gender; smoking/drinking status; C‐reactive protein (CRP); presence of diabetic retinopathy and peripheral neuropathy; presence of comorbidities such as hypertension, dyslipidaemia, coronary heart disease, cerebrovascular disease, and peripheral vascular disease; ankle brachial pressure index (ABPI); toe brachial pressure index (TBPI); and transcutaneous oxygen pressure (TcPO2) were recorded as categorical variables.

According to the International Working Group on the Diabetic Foot (IWGDF), foot ulcer referred to a full‐thickness skin lesion.23 Severities of DFU were graded according to Wagner's classification24 as follows: Grade 0, intact wound with a high risk of foot ulceration; Grade 1, superficial ulcer without clinical infection; Grade 2, deeper ulcer that reaches tendon, bone, or joint; Grade 3, usually involving bone tissues and abscess formation, osteomyelitis, or tendinitis; Grade 4, localised gangrene characterised by ischaemic ulcer; and Grade 5, extensive gangrene. The skin microvascular status can be well reflected by TcPO2, which measures tissue perfusion through a non‐invasive method. The IWGDF recommended that emergency angiography is necessary for individuals with TBP < 30 mm Hg and/or TcPO2 < 25 mmHg and to perform revascularisation as appropriate.23 Therefore, patients were stratified according to TcPO2 < and ≥ 25 mm Hg in this study. ABPI was calculated by dividing the highest systolic ankle pressure with the highest systolic arm pressure, and the reference range was 0.9 to 1.3.25 TBPI was calculated by dividing higher toe pressure with higher brachial systolic pressure, and the abnormal TBPI ratio was lower or equal to 0.7.26 When analysing amputation rates in this study, the values of ABPI, TBPI, and TcPO2 of the affected foot were used.

Hypertension was diagnosed when the recorded value of SBP was at least 140 mm Hg and that of DBP was at least 90 mm Hg,27 or if patients were taking antihypertensive medicine. Diagnosis of diabetes was made according to the following criteria: a fasting blood glucose level ≥7.0 mmol/L (126 mg/dL); 2‐hour plasma glucose level ≥11.1 mmol/L (200 mg/dL) during an oral glucose tolerance test (OGTT); and a random plasma glucose level ≥11.1 mmol/L (200 mg/dL) for a patient with classic symptoms of hyperglycaemia or hyperglycaemic crisis.28 Dyslipidaemia was defined as increased total cholesterol (≥6.20 mmol/L [240 mg/dL]), LDL‐C (>4.13 mmol/L [160 mg/dL]), and triglyceride levels (>2.25 mmol/L [200 mg/dL]) or decreased HDL‐C (<1.03 mmol/L [40 mg/dL]).29 The self‐reported physician's diagnosis of diabetes or dyslipidaemia was also included. Baseline CRP values were only categorised as quartiles (<6.10 mg/L, 6.10‐26.99 mg/L, 27.00‐107.06 mg/L, ≥107.07 mg/L). BMI is defined as body weight divided by the squared body height (kg/m²). WHR is the ratio of waist circumference to hip circumference. Electrocardiography, echocardiography, or coronary angiography could provide evidence of the presence of coronary heart disease.30 Furthermore, cerebrovascular diseases were recorded based on the patient's case sheets.

Diabetic peripheral neuropathy presented neuropathic symptoms and impaired or absent sensation ability in a glove and stocking distribution, including vibration, light touch, pain, and temperature differences awareness.18 Loss of deep tendon reflex and absence of perception of the Semmes‐Weinstein monofilament (10‐g) at 2 of 10 standard plantar sites of either foot can help diagnose peripheral neuropathy for patients without any other significant diseases.18, 31 Once ophthalmologists observed non‐proliferative or proliferative retinopathy through fundus examination in diabetics, diabetic retinopathy was usually confirmed. PVD diagnosis was made by a history of intermittent claudication and/or lower limb vascular surgery performed before admission, absent/diminished pedal pulses, ABPI <0.9, or significant stenosis in low‐extremity arteries.32 LEA referred to bone removal from any part of the lower extremity. The decision of LEA was made in the conference of endocrinologist and surgeon; moreover, indications of operations included severe soft tissue infection, osteomyelitis, or gangrene.33 Amputation preserving the ankle joint with an intact healed wound was called minor amputation, whilst major amputation was defined as amputation above the ankle joint.23, 34

Statistical analysis was performed using IBM SPSS 22.0 software for Windows. Continuous variables were expressed as means with SD or medians with interquartile range (IQR), with the normality of the distribution verified by the Kolmogorov‐Smirnov test. If normality was confirmed, the Student's t test was used to evaluate differences. Otherwise, the non‐parametric Mann‐Whitney's U test was used. Categorical variables were represented as frequencies with percentage (%) and compared using the χ² test or Fisher's exact test as appropriate. Those factors with P‐values less than .05 in the univariable analyses were entered in a stepwise logistic regression model to identify the independent risk factors associated with LEAs. All statistics were two‐tailed, and a P‐value <.05 was considered statistically significant.

3. RESULTS

This study included a total of 475 diabetic patients (294 males and 181 females) with DFU. The age of patients ranged from 17 to 87 years, with a mean ± SD of 61.18 ± 12.01 years. The mean ± SD of diabetes duration was 9.44 ± 7.04 years. Wagner classification showed 20 cases (5.1%) were grade 1, 116 cases (29.6%) grade 2, 125 cases (31.9%) grade 3, 126 cases (32.1%) grade 4, and 5 cases (1.3%) grade 5. Of the 59 amputations, 4 were major amputations, and 55 were minor amputations. The overall rates of major and minor amputation rates were 0.8% and 11.6%, respectively.

There was no significant difference in the age, gender, duration of diabetes, smoking/drinking status, blood pressure, BMI, WHR, blood glucose level, TC, LDL‐C, ALT, BUN, creatinine, PTH, 25(OH)VitD, BNP, presence of comorbidities or complications mentioned above in the methods, ABPI, TBPI, and TcPO2 between the amputation and non‐amputation groups. When compared with DFU patients without amputation, patients with amputation had increased levels of WBC counts (P = .001), PLT counts (P = .012), ESR (P = .001), CRP (P = .001), and HbA1c (P = .012). However, levels of postprandial plasma C‐peptide (P = .002), Hb (P = .002), HDL‐C (P = .010), triglyceride (P = .030), albumin (P < .001), and UA (P = .015) were decreased in patients with amputations. Patients with advanced Wagner grades (P < .001) had much higher rates of amputation (Table 1).

Table 1.

Demographics and clinical characteristics of patients with and without amputations for diabetic foot

Characteristics	Amputation	Non‐amputation	P‐value
Patients	59 (12.4%)	416 (87.6%)
Age, years	60.00 [53.00, 68.00]	63.00 [52.00, 70.00]	.418
Gender			.882
Male	36 (12.2%)	258 (87.8%)
Female	23 (12.7%)	158 (87.3%)
Diabetes duration, years	10.00 [3.00, 13.00]	9.00 [3.00, 14.00]	.736
Smoking history			.065
Yes	23 (9.7%)	214 (90.3%)
No	36 (15.3%)	199 (84.7%)
Drinking history			.305
Yes	18 (10.5%)	154 (89.5%)
No	41 (13.7%)	258 (86.3%)
Systolic blood pressure, mm Hg	133.00 [124.00, 157.25]	136.00 [120.00, 152.00]	.621
Diastolic blood pressure, mm Hg	79.00 [72.00, 86.75]	80.00 [73.00, 89.00]	.340
Body mass index, kg/m2	22.63 ± 3.79	23.10 ± 3.86	.432
Meggitt‐Wagner grades for DFU			<.001*
Grade 1	0 (0.0%)	20 (100.0%)
Grade 2	1 (0.9%)	115 (99.1%)
Grade 3	10 (8.0%)	115 (92.0%)
Grade 4	46 (36.5%)	80 (63.5%)
Grade 5	1 (20.0%)	4 (80.0%)
Waist‐hip ratio	0.94 [0.90, 0.97]	0.94 [0.90, 0.98]	.878
WBC, 109/L	9.73 [7.80, 16.45]	8.40 [6.40, 11.34]	.001*
Hb, g/L	106.00 [90.75, 115.50]	115.00 [99.00, 131.00]	.002*
PLT, 109/L	270.00 [219.00, 340.75]	238.00 [180.00, 303.25]	.012*
ESR, mm/h	90.00 [72.75, 116.25]	58.00 [31.00, 91.00]	.001*
CRP, mg/L			.001*
<6.10	3 (4.2%)	69 (95.8%)
6.10‐26.99	6 (8.2%)	67 (91.8%)
27.00‐107.06	6 (8.2%)	67 (91.8%)
≥107.07	17 (23.3%)	56 (76.7%)
HbA1c, %	9.65 [7.90, 11.98]	9.10 [7.30, 11.00]	.012*
Fasting glucose level, mmol/L	8.05 [5.68, 11.75]	7.95 [6.20, 10.72]	.882
Postprandial glucose level, mmol/L	13.75 [7.85, 18.10]	12.90 [9.40, 16.90]	.761
Fasting plasma C‐peptide level, ng/mL	1.21 [0.70, 2.48]	1.70 [1.00, 2.91]	.055
Postprandial plasma C‐peptide level, ng/mL	1.90 [1.03, 2.75]	2.55 [1.29, 4.80]	.002*
Total cholesterol, mmol/L	3.63 [2.79, 4.57]	3.94 [3.07, 4.94]	.079
LDL‐C, mmol/L	1.92 [1.31, 2.60]	2.00 [1.43, 2.63]	.393
HDL‐C, mmol/L	0.92 [0.71, 1.11]	1.01 [0.84, 1.20]	.010*
Triglyceride, mmol/L	1.00[0.83, 1.34]	1.20 [0.83, 1.69]	.030*
Albumin, g/L	28.64 ± 4.75	32.60 ± 7.21	<.001*
ALT, U/L	15.00 [10.50, 28.50]	17.00 [11.00, 25.00]	.782
BUN, mmol/L	5.82 [4.39, 8.35]	6.13 [4.58, 9.23]	.337
Creatinine, umol/L	83.00 [61.50, 118.50]	78.00 [60.00, 123.25]	.650
UA, umol/L	255.00 [198.00, 329.00]	297.00 [230.00, 389.00]	.015*
PTH, pg/mL	31.10 [18.42, 47.07]	33.20 [22.28, 52.71]	.356
25(OH)VitD, ng/mL	10.47 [6.55, 15.43]	11.42 [5.78, 18.74]	.432
BNP, pg/mL	101.13 [26.54, 212.36]	103.45 [34.17, 370.41]	.526
Abnormal ABPI			.228
Yes	24 (15.7%)	129 (84.3%)
No	21 (11.2%)	166 (88.8%)
Abnormal TBPI			.142
Yes	29 (15.6%)	157 (84.4%)
No	11 (9.6%)	103 (90.4%)
TcPO2 < 25 mm Hg			.123
Yes	20 (18.0%)	91 (82.0%)
No	27 (11.8%)	201 (88.2%)
Hypertension status			.175
Yes	28 (10.7%)	233 (89.3%)
No	31 (14.9%)	177 (85.1%)
Dyslipidaemia			.387
Yes	7 (9.2%)	69 (90.8%)
No	41 (12.8%)	279 (87.2%)
Coronary heart disease			.225
Yes	8 (8.8%)	83 (91.2%)
No	51 (13.5%)	327 (86.5%)
Cerebrovascular disease			.052
Yes	3 (4.9%)	58 (95.1%)
No	56 (13.8%)	350 (86.2%)
Peripheral vascular disease			.372
Yes	6 (20.0%)	24 (80.0%)
No	52 (12.5%)	363 (87.5%)
Diabetic peripheral neuropathy			.076
Yes	52 (13.9%)	322 (86.1%)
No	6 (6.9%)	81 (93.1%)
Diabetic retinopathy			.082
Yes	31 (15.2%)	173 (84.8%)
No	22 (9.7%)	205 (90.3%)

Notes: Values are presented as number (%), mean ± SD, or median [IQR].

Abbreviations: ABPI, ankle brachial pressure index; ALT, alanine aminotransferase; BNP, brain natriuretic peptide; BUN, blood urea nitrogen; CRP, C‐reactive protein; DFU, diabetic foot ulcer; ESR, erythrocyte sedimentation rate; Hb, haemoglobin; HbA1c, glycated haemoglobin; LDL‐C, low‐density lipoprotein cholesterol; HDL‐C, high‐density lipoprotein cholesterol; PTH, parathyroid hormone; IQR, interquartile range; PLT, platelet; TBPI, toe brachial pressure index; TcPO2, transcutaneous oxygen pressure; UA, uric acid; WBC, white blood cell.

^* Statistical significance at P < .05 through univariate analysis.

Then, binary logistic regression model was used to examine the independent risk factors predicting amputation in a stepwise manner. Multivariate analysis showed that elevated HbA1c (odds ratio [OR] 1.317), decreased triglyceride (OR 0.255), and higher Wagner grades (OR 20.947) at treatment were associated with increased risk of amputations (Table 2). Other variables included in the logistic regression model were not found to be statistically significant in determining the amputation risk.

Table 2.

Multivariate logistic regression analysis of risk factors to predict amputation in the diabetic foot ulcer patients

Factors	B	SE	OR	LCL	UCL	P‐value
HbA1c	0.275	0.133	1.317	1.015	1.709	.039
Triglyceride	−1.365	0.684	0.255	0.067	0.975	.046
Meggitt‐Wagner grades	3.042	0.818	20.947	4.216	104.080	<.001

Abbreviations: B, regression coefficient; LCL, lower control limit; HbA1c, glycated haemoglobin; OR, odds ratio; SE, standard error; UCL, upper control limit.

4. DISCUSSION

Our study, through a univariate analysis, has identified the following factors as being predictive for LEA in patients with DFU: WBC and PLT counts, Hb, ESR, CRP, HbA1c, postprandial plasma C‐peptide, HDL‐C, triglyceride, albumin, UA, and Wagner classification. In the multivariate analysis, higher HbA1c levels, decreased triglyceride levels, and higher Wagner grades appeared to be independent predictive factors for LEA in DFU patients.

Although the factors associated with diabetic patients developing foot ulcers have already been well defined in previous reports,35 risk factors for amputation performed in Chinese patients are less clear, especially in the Hunan province. Many studies have identified various independent risk factors for LEA in diabetics, including a history of foot ulcer,36 limb ischaemia, presence of gangrene, deep wounds, advanced age, elevated inflammatory markers,37 poor glycaemic control,38 specific ethnicities and geographical regions, and nephropathy.39 Markowitz et al40 observed that the DFU amputation rate was significantly increased with male gender and the presence of renal disease and peripheral vascular disease in a retrospective case‐control study. Carlson et al41 demonstrated that digital deformity, diabetic neuropathy, and ischaemia were obvious risk factors for toe amputations, while gender, foot infection, foot abscess, osteomyelitis, diabetic retinopathy, and diabetic nephropathy were less obvious risk factors. Chaturvedi et al42 suggested that elevated levels of glucose and triglyceride and the presence of retinopathy were critical risk factors for amputation.42 Miyajima et al, in a review of 210 diabetic patients over the past 9 years, showed that atherosclerosis obliterans with multiple stenosis, haemodialysis history, and elevated HbA1c levels were independent risk factors for major amputation.43 Thus, various risk factors have been identified by different studies. This variability may be because of variations in the study designs, as well as differences in the genetic profiles and cultural or ethnical features of the populations studied.

Although there were inconsistencies among studies regarding risk factors, HbA1c elevation was found to be closely associated with increased risk of amputation in different reports.17, 19, 43, 44, 45, 46, 47 In our analysis, elevated HbA1c levels also led to a significantly higher rate of amputations (OR 1.317), which was compatible with previous studies. According to Pemayun et al,18 significant risk factors of LEA in their study included hypertension status (OR 2.85, 95% CI: 1.23‐6.60), presence of PAD (OR 6.80, 95% CI: 2.67‐17.32), presence of foot necrosis or gangrene (OR 25.88, 95% CI: 6.97‐96.13), FPG ≥ 126 mg/dL (OR 9.43, 95% CI: 1.13‐78.78), HbA1c ≥8% (OR 9.54, 95% CI: 2.03‐44.89), and triglycerides ≥150 mg/dL (OR 4.16, 95% CI: 1.75‐9.86); thus, good control of blood glucose and lipid levels and management of comorbidities such as PAD and hypertension are considered important factors in reducing amputation risk. According to a recent retrospective study by Shatnawi et al,14 higher HbA1c was also significantly associated with increased risk of major amputation (OR 4.0; 95% CI: 1.3‐12.4) in multivariate analysis, and diabetes duration longer than 15 years (OR 6.0), renal dysfunction (OR 3.5), and gangrene (OR 4.2) were also closely associated with increased risk of major amputation. A meta‐analysis by Zhou et al45 showed a statistically significant association between HbA1c and LEA risk (χ (2) = 65.51, P < .001), and the OR for LEA incidence was 1.229 (95% CI: 1.169‐1.292) for every 1% HbA1c increase in linear model; moreover, for HbA1c between 5% and 9%, the OR of LEA risk increased with HbA1c levels in the spline model. In summary, the strong association of HbA1c with amputations caused by DFU could indicate that long‐term hyperglycaemia plays a key role in impaired wound healing, PAD, and increased susceptibility to infection.48, 49, 50

Type 2 diabetics typically presented elevated triglycerides and decreased HDL‐C levels. These lipid abnormalities were testified as independent risk factors of cardiovascular disease and commonly associated with poor prognosis.51, 52 Only limited studies have been reported about the relationship between lipid or lipoprotein abnormalities and increased risk of DFU amputations. In our study, decreased levels of triglyceride and HDL‐C were found to be significantly associated with higher amputation rates. Furthermore, decreased triglyceride level was identified as an independent predictive factor of lower limb amputation (OR 0.255). Another retrospective study about Chinese diabetic patients also showed a protective role of triglyceride in amputation caused by DFU, with an OR value of −0.488.22 However, some reports indicated that DFU patients with abnormally elevated triglyceride levels were at higher amputation risk.42, 53 A study by Zubair et al53 reported that the levels of fasting triglyceride (>150 mg/dL), cholesterol (>150 mg/dL), LDL‐C (>100 mg/dL), and HDL‐C (<40 mg/dL) were associated with amputation risk. Chaturvedi et al also found that increased serum triglyceride and cholesterol levels were related to higher incidences of lower‐extremity gangrene and/or amputation.42 Decreased levels of lipid profiles in DFU patients usually indicated malnutrition, which caused rapid protein loss and wound‐healing delay. They may also suffer from malnutrition because of limited dietary protein intake and excessive urinary protein excretion. However, another study by Lacle et al47 from Costa Rica and a multinational study by Chaturvedi et al42 demonstrated that neither LDL‐C nor HDL‐C was a risk factor for LEA.

The Wagner classification system is commonly used in clinical practice, which evaluates the wound depth and the presence of osteomyelitis and necrosis or gangrene.54 However, the prognostic accuracy of the Wagner grades and the risk evaluation for LEA were not clearly clarified because of insufficient data. Treece et al reported that the depth of the ulcer was an important predictor of the prognosis.55 Deep ulcerations were significantly associated with threefold increased risk of amputation.56 Similarly, Pickwell et al57 found that diabetic patients with infected deep foot ulcers were at higher risk of amputation in 10 European countries. Armstrong et al58 also demonstrated that major amputation risk was much higher when the wound penetrated to the bone. Overall, Wagner grades were significantly correlated with the increased risk of amputation.59 A recent meta‐analysis demonstrated that Wagner Grades 4 (OR 4.3) and 5 (OR 6.4) were identified as predictors of amputation in DFU.60 It is rational that higher Wagner scores, especially grades 4 and 5, were strongly associated with amputation in the present study (OR 20.947) as Wagner grades 4 and 5 indicate the presence of localised or diffused gangrene because of ischaemia and infection.

In the present study, decreased albumin levels were found to be significantly associated with amputation through univariate analysis. A low albumin level usually indicates poor nutritional status and, possibly, renal dysfunction. Inflammation could lower albumin levels under different nutritional circumstances.61 Aetiologies of decreased albumin levels in acute and chronic inflammatory states include increased albumin degradation resulting from high catabolic rate and extravascular albumin leakage because of increased capillary permeability.62 Low albumin levels were also found to be associated with poor wound healing after Syme amputations, which caused major LEA in diabetics.63 According to a cross‐sectional study about T2DM patients by Yusof et al, low albumin levels were also found to be a significant predictive factor of major LEA among these participants, along with another two risk factors, namely, T2DM duration ≥10 years and positive bacterial culture.64 In our study, although we failed to define lower albumin as an independent risk factor for LEA, the facts that patients who underwent amputations had lower levels of UA (P = .015) and albumin (P < .001), combined with the triglyceride (OR 0.225, 95% CI: 0.067‐0.975) being an independent risk factor for LEA, indicated that a lack of nutrition could largely increase the amputation risks in these diabetic populations.

WBC counts, CRP, and ESR were found to be significantly associated with amputation in our univariate analysis. Weigelt et al65 demonstrated that CRP was associated with the severity of the ulceration and ulcer sizes according to the University of Texas classification. Eneroth et al57 showed that WBC counts > 12 × 10⁹/L were related to increased amputation risk. The severities of limb ischaemia and infection and the presence of osteomyelitis could largely reflect the acute‐phase responses in DFU patients. In a pilot study, Kaleta et al66 suggested that an ESR level of 70 mm/h might be the optimal cut‐off point to predict osteomyelitis with a sensitivity of 89.5% and specificity of 100%. ESR levels higher than 70 mm/h were reported to be highly efficient in correctly diagnosing diabetic osteomyelitis from cellulitis among the haematological parameters.67 A large Turkish cohort conducted by Yesil et al indicated that the baseline levels of acute‐phase reactants and decreased haemoglobin levels were significantly associated with increased amputation risk; multivariate analysis also showed that baseline CRP and ESR levels were independent predictors of overall and major amputations, respectively.37 According to Akinci et al, post‐treatment CRP levels, ESR, and WBC counts were significantly associated with increased amputation risk.68 In our study, several variables that were significantly associated with amputation in univariate analysis became insignificant predictive factors after adjustment through multivariate analysis; one explanation might be that these features were not independent of each other. We also failed to show any association between amputation and blood glucose levels, age, gender, duration of diabetes, LDL‐C, total cholesterol, and smoking or drinking status.

Various factors could accelerate the formation of foot ulcers, including peripheral neuropathy, foot deformity, trauma, presence of PVD, and peripheral oedema.69 Calle‐Pascual et al70 reported that PVD was associated with all major amputations in their studies. Although the presence of PAD has been cited by many authors as a risk factor for amputations in diabetics,40, 41, 43, 71 lower limb amputation was not significantly associated with the presence of PVD or abnormal TBPI in present study. Our study also failed to show a significant association between ABPI and amputation, possibly because these patients were not stratified according to ABPI values; another possibility was that some individuals might have avoided amputation by receiving endovascular therapy. ABPI ≤0.5 usually indicated severe poor perfusion and non‐healing ulcers because of ischaemia, which required surgical or interventional revascularisation therapy to avoid amputation. As inflammation markers were proven to be associated with LEA in our study, and PVD, ABPI, or TBPI was not related with amputation risk, it was reasonable to suggest that these amputations were more likely to be caused by infections and inflammation.

Most of our results are compatible with those of previous studies. We have also noted the following limits and shortcomings: (a) lack of details about the exact kind, location, and extension of foot ulcers, and imaging examinations of the limb arterial axis were also not provided; (b) we could not examine the risk factors for major and minor amputations separately because of limited data; (c) medication used was not analysed in the present study, which might explain why these patients with comorbidities were less likely to undertake amputation; and (d) we only used the Wagner classification system, which provided little information about ulcer size and infection status. Other classification systems, such as the University of Texas (UT); the perfusion, extent, depth, infection, and sensation (PEDIS); and ulcer size (area and depth), sepsis, arteriopathy, and denervation (S[AD]SAD) systems, could provide more detailed information about the foot ulcer;72 (e) being a retrospective study and not involving long‐term outcome, a prospective multicentre research on a larger scale will be much more helpful for preventing amputations because of DFU.

5. CONCLUSION

In summary, this study has shown that independent risk factors for diabetic foot‐related amputations in the south‐central part of China included elevated HbA1c level, lower triglyceride level, and higher Wagner grades. After diagnosis of DFU, intensive monitoring and appropriate medical care may improve the prognosis and reduce the amputation risk of diabetic patients. Emphasising clinical management and nutritional support are also strongly recommended. Efforts to promote medical treatment and prevent LEAs of diabetics, especially among high‐risk individuals, should be supported as a long‐term sustainable component of China's health care system.

CONFLICT OF INTEREST

The authors declared no conflicts of interest in this study.

Guo Z, Yue C, Qian Q, He H, Mo Z. Factors associated with lower‐extremity amputation in patients with diabetic foot ulcers in a Chinese tertiary care hospital. Int Wound J. 2019;16:1304–1313. 10.1111/iwj.13190