Diabetes mellitus accelerates the development

Transcription

1 O R I G I A L A R T I C L E Lower-Extremity Amputations in Diabetic and ondiabetic Patients A population-based study in eastern Finland OI I. SlITOE, MD, PHD LEO K. ISKAE, MD, PHD MARKKU LAAKSO, MD, PHD JUKKA T. SIITOE, BS KALEVI PYORAIA, MD, PHD OBJECTIVE To study the incidence of LEAs attributable to PVD in diabetic and nondiabetic patients. The age at first amputation, the level of amputation, the number of reamputations, and survival after amputation also were examined in the study populations. RESEARCH DESIG AD METHODS This retrospective study was based on a population of 253,000 inhabitants in eastern Finland. All patients with their first LEA performed during the period from 1 January 1978 to 31 December 1984 were identified from the registers of operation theaters in the study area. Furthermore, patient records and death certificates were reviewed. Amputations attributable to causes other than evident atherosclerotic vascular disease were excluded. RESULTS Altogether, 477 patients (85 diabetic men, 127 nondiabetic men, 169 diabetic women, and 96 nondiabetic women) were identified. The overall LEA rate was 26.9/100,000 per yr, and the incidence increased strongly with age in both diabetic and nondiabetic patients. The age-adjusted amputation incidence per yr was 349.1/100,000 for diabetic men, 33.9/100,000 for nondiabetic men, 239.4/100,000 for diabetic women, and 17.2/100,000 for nondiabetic women. The proportion of peripheral (toe, leg) amputations was markedly higher in diabetic patients who also tended to have more reamputations during the follow-up than did nondiabetic subjects. The diabetic status per se was a statistically significant risk factor for mortality in women, but not in men. COCLUSIOS Diabetic men and women had a and 13.8-fold higher risk, respectively, for LEA. FROM THE DEPARTTS OF MEDICIE AD CLIICAL UTRITIO, KUOPIO UIVERSITY HOSPITAL, AD THE UIVERSITY OF KUOPIO, KUOPIO, FILAD. ADDRESS CORRESPODECE AD REPRIT REQUESTS TO LEO ISKAE, MD, PHD, DEPARTT OF MEDICIE, KUOPIO UIVERSITY HOSPITAL, SF KUOPIO, FILAD. RECEIVED FOR PUBLICATIO 19 OVEMBER 1991 AD ACCEPTED I REVISED FORM 17 AUGUST LEA, LOWER-EXTREMITY AMPUTATIO; PVD, PERIPHERAL VASCULAR DISEASE; A OVA, AALYSIS OF VARIACE; TYPE II DIABETES, O-ISULI-DEPEDET DIABETES MELLITUS. Diabetes mellitus accelerates the development of all atherosclerotic complications, including PVD (1). LEA is a complication of peripheral arterial disease, and it is markedly more frequent among diabetic than nondiabetic patients (2-5). In a large autopsy study, gangrene of the lower extremity was 40 times more frequent in diabetic than in nondiabetic subjects (6). Diabetes is overrepresented among subjects who have PVD. In a study by Bell (7), 60-80% of patients with PVD and gangrene had diabetes. The prevalence of diabetes (8) and diabetic amputation rates increase with advancing age (5). LEAs imply significant morbidity, mortality, disability, and high costs in the health care system. As the proportion of elderly subjects in the population increases, the foot problems of diabetic patients can be considered a major health problem. Only a few population-based studies have investigated the frequency of LEAs in diabetic patients compared with those in nondiabetic subjects (5). Therefore, we conducted a study to estimate the incidence of LEAs attributable to atherosclerotic vascular disease and to assess the impact of diabetes on the risk of LEAs in the population of Kuopio province in eastern Finland. Furthermore, we analyzed the age at the time of first amputation, the level of amputation, the number of reamputations, and the survival after amputation both in diabetic and nondiabetic subjects. RESEARCH DESIG AD METHODS The study population comprised the inhabitants of Kuopio province in eastern Finland. In that area, the 1978 population was 253,157 and the 1984 population was 255,929. The 1978 population was used as a denominator for incidence calculations. To estimate the number of diabetic patients in each age-group in the study area, we used data from the register of drugtreated diabetic patients in All 16 DIABETES CARE, VOLUME 16, UMBER 1, JAUARY 1993

2 Siitonen and Associates Finnish diabetic patients who need drug therapy receive it free of charge, according to Finland's Sickness Insurance Act. A central register of all patients who receive drug treatment is kept by the Social Insurance Institution in Helsinki. This register covers virtually all drug-treated diabetic patients, but does not include diet-treated diabetic patients. However, because the prevalence of diet-treated diabetic patients was known on the basis of the results from the population-based epidemiological surveys conducted by our department, the number of diettreated diabetic patients in the study area was estimated by using the data from this previous survey (8). The number of subjects in each age-group and for each sex was obtained from the population register. The numbers of nondiabetic subjects were obtained by subtracting the estimated numbers of diabetic patients from the total number of subjects in each age-group. During the period from 1 January 1978 to 31 December 1984, all patients with an LEA were identified from the registers of the operation theaters of all five hospitals with facilities for amputation in the study area (by codes indicating lower-extremity surgical procedure). Subjects who had an amputation as a result of acute thromboembolism, vasculitis, large ulcerations caused by venous varicose disease, or nonvascular indications (such as trauma, frostbite, or tumor) were excluded. Subjects who had undergone an earlier minor or major LEA before 1 January 1978 also were excluded; thus, our incidence calculations were based on the first amputation conducted during the 7-yr study period. Data concerning the patients' history of diabetes, mode of diabetes treatment, age at the time of first amputation, level of amputation, number of reamputations, and age at the time of death were obtained from hospital records and/or death certificates from Finland's Central Statistical Office. The cause of death was considered cardiovascular with ICD-9 codes (ischemic heart disease), Table 1umber of diabetic and nondiabetic patients with LEAs AGE (YR) DURATIO OF DIABETES (YR) TREATT OF DIABETES (%) ORAL DRUGS ISULI ± ± (23.5) 33 (38.9) 32 (37.6) Values are means ± SD. n, number of patients. *P = 03, diabetic versus nondiabetic (other heart diseases), (cerebrovascular diseases), and (diseases of the arteries). The survivors were followed until 31 December 1988, and the survival for nondiabetic and diabetic subjects was analyzed according to the Cox-survival model (9). The x 2 test was used to estimate statistical significances for the differences between the group frequencies, and AOVA was used for testing statistical differences of continuous variables between the groups. Age adjustment was performed by the direct method, which used the Finnish population >25 yr of age in 1980 as the standard population. RESULTS Altogether, 477 patients (85 diabetic men, 127 nondiabetic men, 169 diabetic women, and 96 nondiabetic women) with their first LEA fulfilling the criteria above were registered (Table 1). Diabetic men and women were younger at the time of amputation compared with corresponding nondiabetic subjects, but the difference was statistically significant only in women (P = 03). We noted no difference in known duration of diabetes between the sexes. Insulin treatment was more frequent among diabetic men than among diabetic women (P = 2). Table 2 presents the incidence of LEA by age and sex in diabetic and nondiabetic subjects. The incidence increased with age in both sexes and in O ± ±9.1* 12.0 ± (37.8) 66(39.1) 39(23.1) WO O ± 10.7 both groups (diabetic and nondiabetic), and was markedly higher in diabetic patients than in nondiabetic subjects. The overall amputation rate was 26.9/ 100,000 per yr. In comparing the figures of diabetic and nondiabetic subjects in age-groups >25 yr during which the first amputation had been conducted, the age-adjusted amputation incidence per 100,000 population per yr was in diabetic men, 33.9 in nondiabetic men, in diabetic women, and 17.3 in nondiabetic women. Thus, diabetic men and women had a and fold higher risk, respectively, for LEA compared with nondiabetic men and women. Amputations were performed mostly at the thigh level (Fig. 1). The proportion of peripheral (toe and leg) amputations was higher in diabetic patients than in nondiabetic subjects (P < 01 for men, P = 02 for women). During the 11-yr follow-up, diabetic men and women tended to have more reamputations compared with nondiabetic men and women (Table 3). The incidence of LEAs by mode of treatment in diabetic patients is presented in Table 4. The high frequency of insulintreated elderly subjects is explained, in part, by the transfer to insulin because of deterioration of metabolic control before the amputation. The crude number of deaths during the follow-up in both groups are DIABETES CARE, VOLUME 16, UMBER 1, JAUARY

3 Amputations in diabetic and nondiabetic patients Table 2Population of the study area (diabetic subjects estimated), incidence of LEAs/100,000 nondiabetic and diabetic subjects peryr, and incidence ratio by sex and age AGE-GROUP (YR) <24 75 > TOTAL AGE ADJUSTED WO <24 75 > TOTAL AGE ADJUSTED ICIDECE OF AMPUTATIOS/ 100, ,248 50,780 10, ,750 52,136 48,316 13,347 10, ,407 O ICIDECE OF AMPUTATIOS/ 100, ICIDECE RATIO ( VERSUS O) Diabetes (11). In that study, the prevalence of leg amputation varied widely between diabetic populations, and was lowest in Oriental diabetic populations, which are known to have the lowest overall rates for atherosclerotic vascular disease. Lindega'rd et al. (12) investigated the incidence of all lower leg and thigh amputations associated with diabetes in two Swedish counties, Gotland and UmeS, between 1971 and The estimated incidence rates of amputations in diabetic patients were 20.5/100,000 per yr in Gotland and 6.5/100,000 per yr in UmeS. The lower amputation rate in Umea 1 was probably attributable to more effective preventive procedures in this district. Similarly estimated, the annual lower leg and thigh amputation rate per 100,000 total population was 10.1 in our diabetic population. LEAs among diabetic Pima Indians have been reported at 206/100,000 per yr (4), the highest incidence rate reported in any diabetic shown by sex in Table 5. The main cause of death was, according to the death certificates, cardiovascular, in 40 (64%) diabetic men, in 63 (67%) nondiabetic men (S), in 89 (59%) diabetic women, and in 52 (80%) nondiabetic women (P < 5), respectively. The other major cause of death in diabetic patients was related to diabetes and its complications (24 and 23% in men and women, respectively). Survival after the first LEA (Fig. 2) was analyzed by sex and diabetic status by the Cox model, which included age as an independent variable. In men, diabetes status per se was not associated with mortality, but in women, the coefficient/se for diabetes status was 2.60 (P < 5). This implies that diabetes independently contributed to total mortality in women. COCLUSIOS This populationbased study from eastern Finland showed that diabetic men and women had a and a 13.8-fold higher risk for LEA, respectively, compared with nondiabetic subjects. Furthermore, the overall amputation rate (26.9/100,000 per yr) can be considered high. However, this figure is comparable with previous southern Finnish data showing about the same overall incidence of vascular LEAs (10). Only a few previous reports have been published on the incidence rate of LEAs among the diabetic population. Most and Sinnock (5) reported an ageadjusted LEA rate based on hospital discharge data for diabetic individuals in the U.S. The rate increased in diabetic men from 27.1/10,000 in the age-group <45 yr to 132.1/10,000 in those >65 yr, and in diabetic women from 10.3 to 90.9/10,000, respectively. Somewhat higher incidence rates in our study may be explained, in part, by the high prevalence of atherosclerotic vascular disease in eastem Finland (1). This explanation is supported by the findings of the WHO Multinational Study of Vascular Disease in yrs yrs yrs > 75 yrs I II III I II III I II III I i DM- C3 DM* WO yrs yrs yrs > 75 yr il I II III I II III I II III I II III IH DM- [Q DM- Figure 1Level of first LEA by age-group and by sex for diabetic (DM+) and nondiabetic (DM-) individuals. (I), Foot level; (II), leg level; (III), thigh level. 18 DIABETES CASE, VOLUME 16, UMBER 1, JAUARY 1993

4 Siitonen and Associates Table 3umber of diabetic and nondiabetic patients with one or more LEAs WO O O UMBER OF AMPUTATIOS OE Two THREE FOUR OR MORE 52 (61.2) 23 (27.9) 7 (8.2) 3 (3.5) 92 (72.4) 27(21.3) 8 (6.3) 91 (53.8) 49 (29.0) 22 (13.0) 7(4.1) 76 (79.2) 18 (18.8) 2 (2.0) population. In our study, the amputation incidence rate among diabetic patients was similar to that found in the Pima study. Several factors, such as the selection of the study population, the age and the race of the patients, regional facilities for preventive care and advanced vascular surgery or angioplasty may explain the variation in incidence rates. In our study, the relative risk for LEA was greatest among diabetic patients <54 yr of age and was decreased with age. This change of relative risk with age probably reflects the increasing prevalence of atherosclerotic vascular disease in the nondiabetic population. Furthermore, the frequency of smokers was likely higher in the nondiabetic population, but the role of this important and potentially preventable risk factor could not be estimated reliably. Other risk factors with respect to the development of PVD are elevated blood pressure, abnormalities in serum lipids and lipoproteins, and known duration of diabetes (11,13-15). Recently, we have shown that in addition to smoking and elevated systolic blood pressure, abnormalities in very-lowdensity lipoprotein metabolism and hyperinsulinemia were associated independently with the development of PVD during the 5-yr follow-up of patients with newly diagnosed type II diabetes and in nondiabetic control subjects (16). It is generally assumed that, in diabetic individuals, amputations are the ultimate result of various factors, primarily advanced PVD and neuropathy. Other factors, such as minor trauma or failure of cutaneous wound healing, also may play a part (17). In contrast, in nondiabetic subjects, nontraumatic amputations result primarily from severe atherosclerosis. In this study, we could not analyze the impact of neuropathy or other unitary causes that could have lead to amputation. Based on patient medical histories obtained from hospital records, severe peripheral arterial disease was a common feature of the patients. ote that subjects with an amputation attributable to nonvascular reasons and to other vascular diseases were excluded. A potential bias for incidence calculations may result from the different treatment policies of lower-extremity rest pain or gangrene in various hospitals. This population-based study included all hospitals with surgical facilities for amputations in the study area. During the study period, the potential 'on-dlabetlc Figure 2Estimated survivor functionthe survival histogram by sex for diabetic and nondiabetic patients after the first LEA. role of surgery in limb salvage, especially in elderly diabetic patients, was neglible. However, since the mid 1980s, vascular surgery and, especially, angioplasty, has played an increasingly important role in this respect. The prevalence of drugtreated diabetes in the population was known from the drug register of the Social Insurance Institution in Finland. The total number of diabetic patients, including those treated with diet alone, was estimated by using a previously published population-based prevalence Table 4Incidence of LEAs in diabetic patients per 100 patient-yr by age and mode of treatment AGE-GROUP (YR) >75 OVERALL ISUU ORAL ISULI WO ORAL DIABETES CARE, VOLUME 16, UMBER 1, JAUARY

5 Amputations in diabetic and nondiabetic patients Table 5umber of deaths among diabetic and nondiabetic patients in study with LEA and the time interval between the first amputation and death during follow-up DEATHS (%) TIME ITERVAL (MO)* * Values are means ± SD. 63 OF 85 ( ± 22.3 O 1) 94 OF 127(74.0) 24.7: ±26.8 study (8). Thus, the calculated overall prevalence of diabetes in the area of the Kuopio University Hospital was 3.0%, which is in agreement with earlier studies conducted in Finland (8,18). The prognosis after amputation has reportedly been poor in diabetic patients (19,20). Survival of a diabetic amputee was significantly reduced in female diabetic patients compared with female nondiabetic patients. This accords with other studies indicating that diabetes increases the risk for atherosclerotic complications, particularly in women (1). In conclusion, the risk for amputation of lower extremities as a result of peripheral arterial disease increases significantly with advancing age in the nondiabetic population. In spite of that, the risk for amputation is markedly increased in diabetic individuals. Because an LEA is a serious, but potentially preventable vascular complication among elderly people, more attention should be given to risk factors of PVD, to preventive foot care, and to invasive interventions among high-risk patients to reduce the number of LEAs. References 1. Pyorala K, Laakso M, Uusitupa M: Diabetes and atherosclerosis: an epidemiological view. Diabetes Metab Rev 2: , Silverstein MJ: A study of amputations of WO '. O 150 OF 169 (88.8) 65 OF 96 (67.7) 27.3 ± ±25.0 the lower extremity. Surg Gynecol Obstet 137:579-80, Cotton LT, Higton DIR, Berry HE: Diabetes and vascular surgery. Postgrad Med J 47:84-85, elson RG, Gohdes DM, Everhart JE, Hartner JA, Zwemer FL, Pettitt DJ, Knowler WC: Lower extremity amputations in IDDM: 12-yr follow-up study in Pima Indians. Diabetes Care 11:8-16, Most RS, Sinnock P: The epidemiology of lower extremity amputations in diabetic individuals. Diabetes Care 6:87-91, Bell ET: A postmortem study of vascular disease in diabetics. Arch Pathol 53:444-55, Bell ET: Atherosclerotic gangrene of the lower extremities in diabetic and nondiabetic persons. Am J Clin Pathol 28:27-36, Laakso M, Siitonen O, Savolainen K, Kansanen A, Aberra E, Pyorala R Prevalence of different types of diabetes and the mode of treatment in elderly subjects in eastern Finland (in Finnish) Suom Laakaril 42: , Cox DR: Analysis of Binary Data. London, Methuen, Pohjolainen T, Alaranta H: Lower limb amputations in southern Finland Prosthet Orthotics lnt 12:9-18, West KM, Ahuja MMS, Bennett PH, Qyzyk A, De Acosta OM, Fuller JH, Grab B, Grabauskas V, Jarrett J, Kosaka K, Keen H, Krolewski AS, Miki E, Schliak V, Teuscher A, Watkins PJ, Stober JA: The role of circulating glucose and triglyseride concentrations and their interactions with other "risk factors" as determinants of arterial disease in nine diabetic population samples from the WHO Multinational Study. Diabetes Care 6:361-69, Lindega'rd P, Jonsson B, Lithner F: Amputations in diabetic patients in Gotland and Umea Counties Ada Med Scand (Suppl.) 687:89-93, Janka HU: Five-year incidence of major macrovascular complications in diabetes mellitus. Horm Metab Res (Suppl.) 15: 15-19, Beach KW, Bedford GR, Bergelin RO, Martin DC, Vandenberge, Zaccardi M, Strandness DE Jr: Progression of lowerextremity arterial occlusive disease in type II diabetes mellitus. Diabetes Care 11:464-72, Palumbo PJ, O'Fallon WF, Osmundson PJ, Zimmerman BR, Langworthy AL, Kazmier FJ: Progression of peripheral occlusive arterial disease in diabetes mellitus: what factors are predictive? Arch Intern Med 151:717-21, Uusitupa MI, iskanen LK, Siitonen O, Voutilainen E, Pyorala K: 5-year incidence of atherosclerotic vascular disease in relation to general risk factors, insulin level and abnormalities in lipoprotein composition in non-insulin-dependent diabetic and nondiabetic subjects. Circulation 82: 27-36, Pecoraro RE, Reiber GE, Burgess EM: Pathways to diabetic limb amputation: basis for prevention. Diabetes Care 13: , Laakso M, Reunanen A, Klaukka T, Aromaa A, Maatela J, Pyorala K: Changes in the prevalence and incidence of diabetes mellitus in Finnish adults Am J Epidemiol 133:850-57, Silbert S: Amputation of the lower extremity in diabetes mellitus. Diabetes 1:297-99, Cameron HC, Lennard-Jones JE, Robinson MD: Amputations in the diabetic: outcome and survival. Lancet 2: , DIABETES CARE, VOLUME 16, UMBER 1, JAUARY 1993