Prospectively Clinical Study in 81 Patients with Diabetic Foot
from 1990 to 2003: a Follow-up of 13 Years
Xueyi Ma*, Danqing Jing, Jingsheng Hu, Hua Bai, Yan Sun
Abstract: To investigate the clinical characteristics and mortality of clinical endpoint events of the patients with diabetic foot(DF), we prospectively studied 81 cases with DF. According to the foot ulcer classifications system of university of Texas, the 81 cases were divided into four groups, A, B, C and D, their clinical data from inpatient records and outpatient follow-up records were analyzed from 1 Nov. 1990 to 31 Dec. 2003. The levels of blood glucose and lipids, the accumulated amputations rates, mortalities and death reasons were compared. All the patients in 4 groups had a bad glucose control (HbA1c>8.6%), although there were no significant differences in BP, TC, TG and LDL levels and smokers (p>0.05), the cases with hyperlipemia, hypertension and hyperglycemia in all the 4 groups were above 50%. The creatinine clean rate decreased gradually from group D to group A (A: 98.8¡À33.4 ml/min, B: 68.7¡À48.4 ml/min, C: 67.2¡À25.4 ml/min, D: 55.5 ¡À 24.3 ml/min, p<0.05). In group C with severe infections, the levels of CRP and ESR, the count of WBC and PLT significantly increased comparing with the other 3 groups (p<0.01), while the HGB count and HCT decreased significantly in group C and D. The accumulated amputation rate in group C and D were 22.2% and 47.1% respectively and the accumulated mortality in D group after 13 years was 58.8%, significantly higher than the 7.1% in group A. Among the deaths in group D, 95% of cases died of cerebro-cardiovascular endpoint events. In conclusion, good control of hyperglycemia, hyperlipidemia and hypertension in diabetics was the basal way to prevent DF. Diabetes education and protecting foot from trauma and infection can effectively avoid DF, as well as decrease the mortality of clinical endpoint events.
Key words: diabetic foot; risk factors; mortality
Diabetic foot (DF) is a serious end-stage complication in diabetic patients, both neuropathy and ischemic or obstructive arteriosclerosis disease act as permissive factors .The trauma and infection are the main startup or inducement factors. DF include intermittent claudication, chronic ulcer and infections in foot and/or in low limbs, charcot's joints, dry and/or moisture gangrenes, which are the most difficult clinical syndromes to deal with, and they even results in amputations and deformities in diabetic patients. DF also severely depresses the life quality of diabetic patients, consuming high costs of medical and health care. It not only threatens the patients' lives, but also brings them a heavy psychological pressure and distresses. Prevention of DF now is much more important and imperative than treatment of this disease. In order to investigate the clinical progression and prognosis characteristics in Chinese patients with DF, and to help both doctors and patients find and treat in advance the risk factors which may cause amputations or deaths for those with foot problems, we have done a prospective clinical study to follow up 81 patients with DF for 13 years (from 1990 to 2003), analyzing the possible risk factors which might deteriorate the foot progress and all the clinical endpoint events.
94 inpatients with type 2 diabetes mellitus (T2DM) and DF in our hospital since 1st Jan. 1990 were included in our study. According to foot ulcer classification of TEXAS university , they were divided into A, B, C and D 4 groups (see Table 1). The TEXAS foot ulcer classification divided the foot ulcers into 1- 4 degrees and A,B,C,D 4 stages: TEXAS-1 degree: with foot ulcer history; 2 degree: superficial ulcer in foot or leg skin; 3 degree: ulcer depth to muscle tendon and ligament; 4 degree: ulcer depth to bone and joint. TEXAS-A stage: without infection and ischemia; B stage: with infection; C stage: with ischemia; D stage: with infection and ischemia. So in the group A: 14 of the patients only had TEXAS-1 degree foot problems but might in the TEXAS A or C stages; in group B: 15 patients with foot ulceration were in TEXAS-2 degrees but might in A, B, C or D 4 different stages; in group C: 18 of patients had diabetic feet in TEXAS-3 degrees and in B, C or D 3 different stages; in group D: 34 of patients with gangrene in their foot or leg were in TEXAS-4 degrees and in C or D 2 stages (ulceration in foot or leg were deep to bone with osteoarthritis and with ischemia vascular disease). Their clinical data included their hospitalization medical records firstly diagnosed for DF (ulcer, infection or gangrenes), and the 2nd or others several hospitalizations' medical records for any reasons. If they died, the death medical records were collected and analyzed. We also took telephone call to those patients to follow-up their health conditions every year if they did not came to our hospital again. This follow-up continued for 13 years (median follow-up durations were 35 months) until to 31, Dec. 2003. Among the 94 cases, 81cases finished all the data collections, while 13 cases lost. The follow-up rate was 86.2% and the lost rate was 13.8%.
All the clinical data including blood cell counts, blood glucose, hemoglobin A1c (HbA1c), lipid profile, hepatic and renal functions, serum calcium and phosphate, C reactive protein (CRP), fibrinogen (Fib), erythrocyte sedimentation rate (ESR), 24 hours Urine Albumin excretion Rate (UAR)(except urine macroprotein), endogenetic clean creatinine rate (Ccr), ophthalmoscopy and Dopple B type ultrasonic arterial examination for lower limbs (made by doctors of special department) were analyzed and compared in 4 groups.
1. Diagnosis standards for diabetic microvascular complications
Diabetic retinopathy and nephropathy were diagnosed according to Clinical Diabetes Practice Guide . Diabetic neuropathy had been diagnosed if 4 items showed positive in following 7 items: ¢Å Patients complained for feet insensitive/impassible/abnormal feeling (numb, formication, pricking pain);¢ÆVibrancy threshold value of tuning fork¡Ü4(normal¡Ý5);¢ÇAchilles' tendon reflect delayed or disappeared (except the hypothyroidism, edema in low limbs, electrolyte turbulence);¢È retention of urine confirmed by B type ultrasonic examination (postvoid residual urine more than 100 ml in bladder); ¢Éalternation of the diarrhea and constipation without any reasonable reasons;¢Ê gastric stasis and gastroparesis; ¢Ë positional hypotension (the difference of Bp between stand and decubitus¡Ý30/15mmHg)
2. Score of angiopathy
The final scores were the total scores of arterial endangium thickness, degrees of arteriosclerosis and narrow radio in artery diameter. For example: ¢Å When the thickness of endangium in vessel < 1mm were recorded as 0 score, then 1¡«1.2mm recorded as 1 score, while > 1.3mm recorded as 2 score; ¢Æ Stage of arteriosclerosis: no arteriosclerosis in vessel was recorded as 0 score; mild arteriosclerosis (no hyperplasia and no plaque in endangium but increase of ultrasonic resonance which suggested calcification or fibrosis inside arterial) was recorded as 1 score; moderate/sever arteriosclerosis (with mild narrow and plaque in lumen) were recorded as 2 scores; ¢Ç Plaque characteristics: no plaque in vessel were recorded as 0 score; single plaque was recorded as 1 score; ¡Ý2 plaques in a vessel were recorded as 2 scores; diffuse hyperplastic plaques were recorded as 3 scores. ¢È The narrow degree(%) in vessel lumen: 0 score: without narrow in vessel lumens; 1 score: narrow rates in lumen were 30%¡«50%; 2 scores: narrow rates in lumen were 50%¡«70%; 3 score: narrow rates in lumen were ¡Ý76%. The B type Doppler's ultrasonic diagnosis apparatus was HP 8500 IPHX (USA). All the serum biochemical assays were measured by automatic biochemical instruments (Hitachi 7600 Japan).
3. Endpoint events
The accumulated amputation patient numbers and amputation numbers, accumulated death cases, death causes such as cerebro-cadiovascular endpoint events (including myocardial infarction; heart failure; cardiogenic sudden death; stroke) as well as other reasons were collected and compared.
4. Statistic analysis
All the data were represented by ¡À S, the significant differences were at the level of p<0.05 and analyzed by analysis of variance and Dunnett test, the comparison of the morbidity rate among the 4 groups were studied by ¦Ö2 test.
1. The clinical data of patients with diabetic foot
All the subjects in 4 groups showed no significant differences in age, sex, diabetes histories, smokers and morbility of diabetic complications. The numbers of all the patients in 4 groups whose blood pressure, lipid levels beyond the target arrange exceeded 50% . But the cases numbers with higher levels of total cholesterol (TC>4.5mmol/L), triglycerides (TG>1.5mmol/L), low density lipoprotein (LDL>2.6mmol/L) were similar(p>0.05)£¬only the numbers of patients with low HDL (high density lipoprotein) level (<1.0mmol/L) were much more in D group than in other 3 groups (72.2% vs 20%¡«30%) (p<0.05 and p<0.01£¬see Table 1).
Table 1. Comparison of the clinical characteristics of 4 groups' patients with DF
(y)Uncontrolled cases (%)Bp>140/90
mmHgTC > 4.6
n=18GroupsSmoker (%)DMH (y)Stroke (%)CVD (%)DR (%)DN (%)DNN (%)A group
(n=34)13(38.2 )12¡À811(32.4)14(41.2)13(39.4)16(48.5)21(61.8) DR: Diabetic retinopathy; DN: Diabetic Nephropathy; DNN: Diabetic neuropathy; CVD: Coronary Heart Disease; DMH: DM history. ¡ï: In comparison of group D and group A: p<0.05, In comparison of group D and group B: p<0.01, In comparison of group D and group C: p<0.05
2. Comparison of metabolic profile, blood pressure and renal functions among the 4 groups
All the patients showed worse control of blood glucose. Their blood glucose and HbA1c levels increased significantly, especially in D group, the postprandial glucose levels were much higher than that in other 3 groups, their HDL levels were significantly lower than that in B group. The Ccr level reduced significantly along with the increase of the serious degree of DF conditions (see Table 2: the Ccr level in groups D, C and B reduced significantly than in group A, p<0.05). The calcium level in group D were higher than that in both group A and B, which suggested that the serious degrees of metabolism disturbance positively related with serious degrees of deterioration of DF (see Table 2).
Table 2. In comparison of the metabolism profiles, renal functions and blood pressure among the 4 groups
GroupsnFB (mmol/L)P2hBG(mmol/L)HbA1c (%)TG (mmol/L)TC (mmol/L)A group1410.2¡À2.914.6¡À4.18.9¡À1.81.9¡À1.75.2¡À0.8B group1510.1¡À3.713.5¡À5.58.3¡À2.52.1¡À1.45.6¡À1.1C group1811.9¡À4.516.1¡À7.18.9¡À4.02.7¡À3.55.2¡À2.2D group3411.5¡À5.119.9¡À7.8¡ï9.5¡À3.01.4¡À0.95.1¡À1.5HDL (mmol/L)LDL (mmol/L)UA (umol/L)Ca (mmol/L)P (mmol/L)A group141.22¡À0.273.3¡À0.8279¡À812.16¡À0.151.23¡À0.12B group151.33¡À0.213.8¡À1.8311¡À972.14¡À0.171.25¡À0.21C group181.24¡À0.403.6¡À18277¡À952.12¡À0.231.38¡À0.34D group341.05¡À0.45¡ø3.1¡À0.9265¡À1142.31¡À0.22¡÷1.32¡À0.28UAR (mg/24h)Ccr (ml/min)SBP (mmHg)DBP (mmHg)A group1493¡À8799¡À33¡ò147.¡À1778¡À9B group1588¡À12369¡À48148¡À3482¡À9C group1880+8167¡À25140¡À2977¡À15D group3466¡À756¡À24¡ö140¡À2580¡À13 FBG: fasting blood glucose. P2hBG: postprandial blood glucose. HbA1c: glycated hemaglobineA1c. TG: triglyceride. TC: total cholesterol. HDL: high Density lipoprotein cholesterol. LDL: low Density lipoprotein cholesterol. UA: Urine Acid. UAR:Urine Albumin excretion Rate. ¡ï: group D vs. group A, p =0.0227, group D vs. group B: p =0.0111. ¡ø: group D vs. group B:p =0.00304. ¡÷: group D vs. group A: p=0.034, group D vs. group B: p =0.0208. ¡ö: group D vs. group A: p =0.00005. ¡ò: group A vs. group B: p£½0.0384, group A vs. group C: p =0.0439
3. The comparison of the angiopathy and pathological foot
The comparison results of the angiopathies' degrees and pathological feet among the 4 groups were shown in Table 3. All the 34 cases in D group had feet gangrenes; among which the dry, moisture and mixture kind accounted one third, respectively. In D group the angiopathies were the most serious and the scores of angiopathy were the highest among the 4 groups. It was notable that all scores of vascular lesions in both lower extremities were similar in all 4 groups no matter whether the DF was in only one leg, with or without gangrene or which stage of TEXAS classification the pathological changes were in. This meant all the vascular lesions were almost similar in both legs in patients with DF (see Table 3).
Table 3. Comparison of the sever-degrees of diabetic foot and the angiopathy in low limbs of patients in 4 groups
ClassificationTEXAS stage classification of DM footA stageB stageC stageD stageA group1413(21.4)0(0)11(78.6)0(0)B group1522(13.3)3(20)4(26.7)6(40)C group1830(0)2(11.1)0(0)16(67)D group3440(0)2(5.9)4(11.8)28(82.4)GroupsDry
low limbs)Number of
with DM footNumber
low limbsA group
(n=13) TEXAS-1 degree: with ulcer history; 2 degree: skin surface ulcer; 3 degree: ulcer depth to muscle tendon; 4 degree: Ulcer depth to bone and joint. TEXAS-A stage: without infection and ischemia; B stage: with infection; C stage: with ischemia; D stage: with infection and ischemia. The score methods of angiopathy see "study method" in above article. ¡ï: group D vs. group A: p£½0.0008, group D vs. group B: p =0.00003, group D vs. group C: p = 0.0391. ¡ñ: group D vs. group C: p =0.03955, group D vs. group B: p =0.001798, group D vs. group A: p =0.0008. ¡ø: group D vs. group C: p = 0.02858, group D vs. group B: p = 0.00351, group D vs. group A: p = 0.017668
4. The comparison of inflammation factors in 4 groups
With the DF deteriorated according to the TEXAS classifications in 4 groups, patients' CRP, white blood cells count (WBC), blood platelet (PLT), ESR also rose. Especially in group C, the subjects had most serious acute infections, whose hemoglobin (HG) and hematocrit (HCT) decreased significantly. This also suggested the patients' nutritional condition deteriorated when degrees of their DF became worse (see Table 4).
Table 4. The comparison of inflammation factors and blood counting in 4 groups
(%)A group143.0¡À1.31.8¡À3.1¡ï37¡À386.8¡À1.9179¡À35135¡À1640¡À5B group153.3¡À0.75.8¡À7.641¡À448.1¡À2.5234¡À90¡÷128¡À1936¡À5C group183.1+1.321.7¡À34.289¡À20££12.4¡À4.1¡ñ301¡À135¡ò115¡À21¡ø34¡À6¡îD group343.8¡À1.311.5¡À7.0655¡À09.0¡À3.1218+80113¡À20¡ó36¡À5¡ô Fib: Fibrinogen; CRP: C-Reactive protein; ESR: Erythrocyte sedimentation rate; WBC: White blood cell; PLT: Platelet; HGB: Hemoglobin; HCT: Hematocrit. ¡ï: In comparison of group A and group B: p£½0.03499, group A vs. group C: p£½0.0492, group A vs. group D: p = 0.0107. ¡ñ: group C vs. group A: p£½0.0005, group C vs. group B: p£½0.00085. ¡÷: group B vs. group A: p£½0.02344. ££: group C vs. group A: p£½0.023916. ¡ò: group C vs. group A: p£½0.002416, group C vs. group B: p£½0.0454, group C vs. group D: p£½0.0092. ¡ø: group C vs. group A: p£½0.004376, group C vs. group B: p£½0.0488. ¡î: group C vs. group A: p£½0.00918. ¡ô: group D vs. group A: p£½0.016537. ¡ó: group D vs. group A: p£½0.0000337, group D vs. group B: p£½0.010759
5. The comparison of morbility of clinical endpoint events in 4 groups
Along with the deterioration of infections and ischemia proceeding in the leg with DF in different groups, the patients' clinical endpoint events also increased significantly. The numbers of accumulate amputation and amputees were the highest in group D (47.1% and 35.3% respectively). In this group, 4 cases had twice amputations for 2 times in one leg; 2 were amputated both low limbs. The 81 cases with DF were followed up for 35 months (media duration)(2 months¡«17 years). Until the end of study, total mortality of each group respectively was as following: group A: 7.1%; group B: 26.7%; group C: 38.9% and group D: 58.8%(see Table 5). This suggested that with the serious degree of DF and vascular lesions rose, the fatality rate significantly increased. In 4 groups the main reasons of death were cerebro-cadiovascular endpoint events, such as myocardial infarction, heart failure, cardiogenic sudden death and stroke.
Table 5. The comparison of the clinical endstage events incidents in 4 groups
Clinical end-stage events incidentsAmputaition
numbers (%)Persons numbers of
Amputation (%)Total persons of death (%)Cerebro-cardiovascular end-stage events (%)Renal failure
shock (%)A group(n=14)0(0)0(0)1(7.1)1(7.1)0(0)0 (0)B group(n=15)0(0)0(0)4(26.7)3(20.0)0(0)1(6.7)C group(n=18)4(22.2)4(22.2)7(38.9)4(22.2)1(5.5)2(11.1)D group(n=34)16(47.1)¡ï12(35.3)20(55.9)19(52.9)¡ñ0 (0)1(2.9) ¡ï: In comparison of group D and A¡¢B: p <0.01, group D vs. group C: p <0.02. ¡ñ: group D vs. group A: p <0.01, group D vs. group B: p <0.05
1. Metabolic disturbance is an essential reason results in DF
The 81 cases with DF were divided into 4 groups according to the TEXAS foot ulcer' classification system. We did not found significant differences among 4 groups in patients' age, sex, blood pressure, smokers, incidences of diabetic macro-complications and micro-complications. All their blood glucose were worse controlled: the fasting blood glucose >10mmol/L, postprandial 2h BGs>14mmol/L , HbA1c>8.6%. Along with the degree of DF deteriorated£¬the level of HbA1c rose. According to the target goals of serum lipid for T2DM by Western Pacific Asia T2DM Group Police in 1999, diabetic patients and patients with peripheral angiopathy should had their blood lipids controlled as following: TC<4.5mmol/L, TG<1.5mmol/L, LDL<2.5mmol/L, HDL¡Ý1.1mmol/L. Unfortunately half of these patients in all 4 groups did not meet above target goals. However their TC, TG, LDL levels did not show significant differences, only HDL level in group D decreased significantly. Group D had the most patients (72.2%), whose lipid levels did not fall into above standard, and they were significantly higher than that in other 3 groups (p<0.05 and p<0.01). One study in China also observed the negative relationship between incidences of angiopathy in low limbs and serum HDL level at first onset of diabetes. The ankle/brachial pressure index (ABI) was not related to TC, TG levels, but it only negatively related to HDL level . VAHIT study also found in patients with coronary heart disease and lower HDL normal LDL levels, the decrease of incidence of CVD did not depend on the changes of LDL or TG levels, and risk of lower HDL level was also not changed when the hyperlipemia treated by drugs . FREMINGHAM heart study also showed the HDL level was an independent risk factor for CVD. The CVD did not dependent on plasma LDL, TG and TC levels or other non-lipid factors but only on HDL level. It was especially important for diabetics to know that lower HDL level could cause dangers . Recently some medical studies confirmed the HDL had a function of reversional transplantation of cholesterol. Some anti-oxidation enzymes carried by HDL could stop the original inflammation in the process of arteriosclerosis, preventing the congregation and oxidation of LDL, accelerating the fibrin melt. Therefore HDL had an anti-thrombosis and anti-inflammation function . Our study results showed that during the deteriorating procedure of diabetic foot, the decrease of HDL level played much more important promoting rule than TC and TG did, but whether the increase of the plasma calcium level also accelerated the arteriosclerosis procedure was still not clear.
Compared of the peripheral blood cell counts among 4 groups, the HGB and HCT counts were significantly lower in groups C and D than that in groups A and B, which suggested that these patients not only had turbulence metabolic state, but also had malnutrition and anemia which farther aggravated the blood and oxygen supports to tissues. Therefore it may be an important cause or result for patients with DF had very bad prognosis and the low probability of surviving.
2. Inflammation procedure promotes the clinical endpoint events for patients with DF
In all the 81 patients, the DF ulcers, infections and gangrenes were triggered by the same inducement such as hot-scald (because using hot-water bag, treated low limbs problems by infrared ray thermometry and Sauna bath); scrape the skin when cutting toe nails; puncture the skin by foreign nails or some materials; fungal infections in foot skin and Bulla lesion broken, etc. In group A, B and C, although the scores of vascular lesion were similar in both low limbs, the low limbs opposite to the limbs with foot ulcer/infection/gangrenes did not suffer from any same troubles or amputations because the skin were intact. The incidence of diabetic peripheral neuropathy in 4 groups were more than 60%(61.8%¡«71.4%) without any significant differences among each other. But in group C (100% of cases had gangrenes with infections), the inflammation situation was the most serious. The patients' serum CRP¡¢PLT¡¢WBC and ESR were significantly higher than that in the other groups. In group D the patients suffered from infection were also up to 67.7%, and residual patients suffered from only dry gangrene. The patients in group D with DF in TEXAS stage D (had both infection and gangrene) were the most (82.4%) among the 4 groups. Their vascular lesions were also the most serious, and their accumulative total mortality was the highest one (58.8%). While in group A patients without any foot ulcers infected or gangrenes had the highest accumulated surviving rate (92.9%) after 13years. This result suggested that infection plus ischemia in low limbs were the mostly causes resulted in the clinical endpoint events for patients of diabetic foot. The neuropathy was also related to infection because the patients' feet were easy to be wounded with week feeling in feet. Inflammation may deteriorate the abroad inflammation reactions in vessels and in tissues. Each of the inflammation factors accelerated the deterioration of endothelium functions in whole body's vessels and resulted in abnormal blood rheologic changes. The blood hyper-coagulation and hyper-viscidity led the vessels obstruction, the whole metabolic and nutrition situation farther worsen. All of above changes in patients' body increased the amputation rate, and decreased the accumulated surviving rate for DF patients . This observation was similar with that from Faglia . The accumulated mortality after 6 years from DF occurrence in group D was 44.1%, during 7¡«9 years it was 50%, after 10 years it left for 53%, and after 10¡«13 years it left only for 47.2%. The mortality was significantly higher than that in patients of group A who also had angiopathy in low limbs but without wound and infection on feet skin (after 13 years the mortality was only 7.1%). The 94.6% of decedent in D group caused by cerebro-cardiovascular endpoint events, which suggested diabetic patients had not only angiopathy in low limbs but also similarly in whole arterial system. Correcting the metabolic turbulence is the basal method to protect the patients from diabetic macrovascular complications development and progress. It is especially important for those already had angiopathies in low limbs to avoid the DF and to decrease the clinical endpoint events through the diabetes education about how to prevent from foot trauma, wound and infection.
(to be continued to Page 33)
* Corresponding to Xueyi Ma, female, professor, director of the Department of Endocrinology; Address: Beijing 304 Hospital of PLA, Beijing, Postcode: 100037; Tel: 010-66867321; E-mail: firstname.lastname@example.org.