sign in sign up
<<
Home , Diabetic dermopathy, Keratosis, Necrobiosis, Necrobiosis lipoidica

Clinical Care/Education/Nutrition/Psychosocial Research ORIGINAL ARTICLE

The Prevalence of Cutaneous Manifestations in Young Patients With Type 1

1 2 MILOSˇ D. PAVLOVIC´, MD, PHD SLAANA TODOROVIC´, MD tions, such as neuropathic foot ulcers; 2 4 TATJANA MILENKOVIC´, MD ZORANA ÐAKOVIC´, MD and 4) skin reactions to diabetes treat- 1 1 MIROSLAV DINIC´, MD RADOSˇ D. ZECEVIˇ , MD, PHD ment (1). 1 5 MILAN MISOVIˇ C´, MD RADOJE DODER, MD, PHD 3 To understand the development of DRAGANA DAKOVIC´, DS skin lesions and their relationship to dia- betes complications, a useful approach would be a long-term follow-up of type 1 OBJECTIVE — The aim of the study was to assess the prevalence of cutaneous disorders and diabetic patients and/or surveys of cuta- their relation to duration, metabolic control, and microvascular complications in chil- neous disorders in younger type 1 dia- dren and adolescents with . betic subjects. Available data suggest that skin dryness and -like RESEARCH DESIGN AND METHODS — The presence and frequency of skin mani- festations were examined and compared in 212 unselected type 1 diabetic patients (aged 2–22 changes of the hand represent the most years, diabetes duration 1–15 years) and 196 healthy sex- and age-matched control subjects. common cutaneous manifestations of Logistic regression was used to analyze the relation of cutaneous disorders with diabetes dura- type 1 diabetes seen in up to 49% of the tion, glycemic control, and microvascular complications. patients (3). They are interrelated and also related to diabetes duration. Timing RESULTS — One hundred forty-two (68%) type 1 diabetic patients had at least one cutaneous of appearance of various cutaneous le- disorder vs. 52 (26.5%) control subjects (P Ͻ 0.01). Diabetes-associated skin lesions were found sions in young patients with diabetes in 81 (38%) patients. Acquired , rubeosis faciei, diabetic hand, and li- might be potentially useful for the research poidica were seen in 22 vs. 3%, 7.1 vs. 0%, 2.3 vs. 0%, and 2.3 vs. 0% of type 1 diabetic and of their pathogenesis (i.e., derangement of control subjects, respectively. The frequency of cutaneous reactions to therapy was low epidermal lipid metabolism), therapeutic (–2.7%). The prevalence of fungal infections in patients and control subjects was 4.7% and 1.5%, respectively. pilaris affected 12% of our patients vs. 1.5% of control subjects. Diabetic intervention (i.e., application of moisturiz- hand was strongly (odds ratio 1.42 [95% CI 1.11–1.81]; P Ͻ 0.001), and rubeosis faciei weakly ers or antifibrosing agents), or predicting (1.22 [1.04–1.43]; P ϭ 0.0087), associated with diabetes duration. Significant association was microvascular complications. We decided also found between acquired ichthyosis and (1.53 [1.09–1.79]; P Ͻ 0.001). to examine an unselected young type 1 di- abetic population to see what kind of cuta- CONCLUSIONS — Cutaneous manifestations are common in type 1 diabetic patients, and neous manifestations develop at an earlier some of them, like acquired ichthyosis and keratosis pilaris, develop early in the course of the age and with a shorter duration of diabetes. disease. Diabetic hand and rubeosis faciei are related to disease duration.

Diabetes Care 30:1964–1967, 2007 RESEARCH DESIGN AND METHODS — Two hundred and twelve children, adolescents, and young hough it is well known that diabe- first presenting sign or even precede the adults with type 1 diabetes (113 male and tes is associated with a number of diagnosis by many years. The cutaneous 99 female subjects), with disease onset at cutaneous manifestations (1–3), findings can be classified into four ma- T age Յ15 years (Table 1) and consecu- there is a relative paucity of studies jor groups: 1) skin associated tively attending the outpatient diabetes looking at the prevalence of skin with diabetes, such as scleroderma-like clinic at the Mother and Child Healthcare changes in young patients with type 1 changes of the hand, necrobiosis li- Institute of Serbia over a 5-month period diabetes. Cutaneous manifestations poidica, and diabetic dermopathy; 2) (April through August 2005), were exam- generally appear subsequent to the de- cutaneous infections; 3) cutaneous ined by two dermatologists. They took a velopment of diabetes but may be the manifestations of diabetes complica- medical history about skin diseases and ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● performed the whole-body cutaneous ex- amination, including visible mucosal sur- From the 1Department of , Military Medical Academy, Belgrade, Serbia; the 2Department of Endocrinology, Mother and Child Healthcare Institute of Serbia “Vukan E` upic´,” Belgrade, Serbia; the 3De- faces. During the same time frame, 196 partment of Dental Medicine, Military Medical Academy, Belgrade, Serbia; the 4Institute of Dermatovene- healthy children and adolescents (115 reology, Clinical Center of Serbia, Belgrade, Serbia; and the 5Department of Gastroenterology, Military male and 81 female subjects, aged 3–21 Medical Academy, Belgrade, Serbia. years, mean 11.5 Ϯ 4.2) attending the Address correspondence and reprint requests to Dr. Milosˇ D. Pavlovic´, Dermatology, Military Medical Academy, Crnotravska 17, 11002, Belgrade, Serbia. E-mail: [email protected]. dental medicine service of the Military Received for publication 8 February 2007 and accepted in revised form 12 May 2007. Medical Academy for a routine dental Published ahead of print at http://care.diabetesjournals.org on 22 May 2007. DOI: 10.2337/dc07-0267. check-up also underwent dermatological A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion examination by the same physicians. All factors for many substances. clinically definable cutaneous lesions © 2007 by the American Diabetes Association. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby were recorded in both populations. marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Scleroderma-like skin changes of the

1964 DIABETES CARE, VOLUME 30, NUMBER 8, AUGUST 2007 Pavlovic´ and Associates

Table 1—Characteristics of the study popu- Due to a well-known ␣-inflation associ- ence was highly significant (Table 2; P Ͻ lation of 212 young patients with type 1 dia- ated with stepwise logistic regression and 0.01). Thyroid-stimulating hormone lev- betes several regressions performed, a cutoff for els were normal in all patients with ich- significance level for variables in the thyosiform skin changes. Rubeosis, Յ Characteristics model was set at P 0.01. diabetic hand (scleroderma-like changes and/or limited joint mobility), and necro- Age (years) 12.5 Ϯ 3.7 (2–22) RESULTS — The clinical data of the biosis lipoidica were found only in pa- Sex (male/female) 113/99 study population are given in Table 1. A tients with type 1 diabetes (Table 2). All Duration of diabetes 4.2 Ϯ 3.0 (1–15) total of 142 patients (68%) had at least five subjects with diabetic hand were (years) one cutaneous disorder, and 81 patients boys, and two of them also had Du- Age at onset (years) 8.3 Ϯ 3.6 (0.5–15) (38%) had skin lesions considered to be puytren’s contracture. The prevalence of Diabetes complications associated with diabetes (Table 2). The in our type 1 dia- Nephropathy 15 (7) most prevalent cutanous manifestation betic population was 2.3%. Though the Retinopathy 0 (0) was xerosis, found in 22% of type 1 dia- prevalence of fungal, viral, and bacterial Neuropathy 0 (0) betic patients. In control subjects, the ich- infections was higher in the study popu- Hypertension 4 (1.9) thyosiform changes affected only 3% of lation (4.3, 4.3, and 3.0%, respectively) Cumulative A1C 9.1 Ϯ 1.6 the children and adolescents. The differ- than in control subjects, the difference Data are means Ϯ SD (range) or n (%).

Table 2—Distribution of cutaneous lesions in 212 young type 1 diabetic patients and 196 age- hand were diagnosed and assessed ac- and sex-matched control subjects cording to the criteria of Seibod (4). Xerosis (acquired ichthyosis) was clini- Patients Control subjects cally diagnosed based on the palpatory (n ϭ 212) (n ϭ 196) feeling of dry and rough skin accompa- Male/ Male/ nied by visible squames. The disorder was Lesions n (%) female n (%) female typically most severe over shins. In a few cases in which the diagnoses of the two Skin manifestations associated with diabetes examiners were discordant, they exam- Xerosis (acquired ichthyosis) 47 (22.2)* 24/23 6 (3) 3/3 ined the patient together with a third der- Diabetic hand 5 (2.3)† 5/0 0 (0) 0 matologist and reached a consensus. Rubeosis 15 (7.1) 6/9 0 (0) 0 Medical files were reviewed for data on Necrobiosis lipoidica 5 (2.3) 1/4 0 (0) 0 diabetes duration, A1C levels, fasting tri- Infections glycerides, cholesterol and thyroid- Fungal 10 (4.7) 3/7 3 (1.5)‡ 3/0 stimulating hormone levels, and renal Tinea pedis 4 (1.9) 2/2 1 (0.5) 1/0 function. Cumulative A1C values as a 2 (0.9) 2/0 0 (0) 0/0 measure of glucose control were ex- Candidosis 4 (1.9) 1/3 0 (0) 0/0 pressed as a mean of the yearly A1C levels. Viral 8 (3.7) 4/4 4 (2) 2/2 Retinopathy was diagnosed by an experi- Bacterial 7 (3.3)§ 5/2 2 (1) 0/2 enced ophthalmologist using direct and in- 3 (1.4) 1/2 1 (0.5) 0/1 direct ophthalomoscopy, nephropathy was 3 (1.4) 2/1 1 (0.5) 0/1 assessed by means of albumin excretion rate Skin reactions to insulin therapy using three consecutive timed overnight Lipohypertrophy 4 (1.8) 2/2 0 (0) 0/0 urine collections (albumin excretion rate Lipoatrophy 2 (0.9) 0/2 0 (0) 0/0 Ն20 ␮g/min in at least two of three mea- Other skin disorders surements), and peripheral neuropathy was 41 (19.3) 20/23 31 (15.5) 17/14 assessed by means of a positive diabetic Keratosis pilaris 27 (11.7)* 13/14 3 (1.5) 3/0 neuropathy index (5) and, when indicated, Pityriasis versicolor 2 (0.8) 2/0 5 (2.5) 4/1 electromyoneurography. Café-au-lait macules 6 (2.6) 4/2 0 (0) 0/0 Halo nevi 4 (1.7) 2/2 0 (0) 0/0 Statistical analysis Eczema 8 (3.5) 3/5 2 (1) 1/1 Statistical tests were performed using 1 (0.4) 1/0 1 (0.5) 1/0 SPSS 8.0 for Windows (StatSoft). The ␹2 2 (0.9) 1/1 0 (0) 0/0 test was used to assess differences be- 1 (0.4) 1/0 2 (1) 2/0 tween the prevalence of cutaneous lesions 2 (0.9) 2/0 0 (0) 0/0 in patients and control subjects (Table 2). Seborrheic dermatitis 6 (2.6) 3/3 7 (3.5) 5/2 P values were based on a two-sided test Striae dystensae 3 (1.3) 0/3 0 (0) 0/0 and considered significant if Ͻ0.01. Step- /gluten 1 (0.4) 0/1 0 (0) 0/0 wise logistic regression was used to assess enteropathy significant associations of cutaneous le- Purpuric dermatosis 1 (0.4) 1/0 0 (0) 0/0 sions with clinical and metabolic param- *P Ͻ 0.01 vs. control patients. †Two patients also had Dupuytren contracture. ‡Two patients had tinea eters of the patients with type 1 diabetes. corporis (M. canis) and (M. canis), respectively. §A patient with an ingrowing toenail.

DIABETES CARE, VOLUME 30, NUMBER 8, AUGUST 2007 1965 Cutaneous manifestations in young patients was not significant and the numbers were causes of acquired ichthyosis (e.g., atopic type 1 diabetic patients, and although generally low. Keratosis pilaris, though dermatitis), nutritional deficiencies, or microangiopathy has been proposed as not considered a diabetes-related cutane- chronic renal failure. a causative factor for necrobiosis (10), ous manifestation, was significantly more The prevalence of rubeosis faciei in none of our patients had retinopathy or common (11.7%) in type 1 diabetic pa- our patients with type 1 diabetes was nephropathy. tients than in control subjects (1.5%) rather high (7%) and was twice as much The prevalence of cutaneous infec- (P Ͻ 0.01). The prevalence of acne was as in a similar previous study (3), and it tions, though somewhat higher in the comparable in both populations (Table was found to be weakly (P ϭ 0.0087) re- population of type 1 diabetic patients, 2). Other cutaneous disorders were un- lated to the disease duration. The preva- was not statistically different from the common and listed in Table 2. lence in most previous studies in patients healthy subjects (Table 2). Stepwise logistic regression was used to with was estimated at 21– infections (tinea pedis and onychomyco- assess the influence of diabetes risk factors 59% (8,9). It is presumed that venular sis) were diagnosed in 2.8% and candido- and late complications (disease duration, dilation in the cheeks of diabetic patients sis in 1.9% of our diabetic patients. The metabolic control, and nephropathy) on underlies rubeosis faciei and is caused by figures are very close to those found skin lesions. We found that diabetic hand -induced sluggish micro- among 64 type 1 diabetic patients in the was significantly related to the disease du- circulation (10). Hypertension may exac- study of Romano et al. (2), where the ration (odds ratio 1.42 [95% CI 1.11–1.81]; erbate the capillary damage (10). We prevalence of tinea infections was 3% P Ͻ 0.001) with no evidence to relate it to could not demonstrate any relation of ru- and of Candida infections 5%. The fig- either metabolic control or nephropathy. beosis faciei in our young patients with ures were higher for their type 2 dia- Rubeosis faciei was less strongly related to type 1 diabetes with metabolic control, betic patients (ϳ20%). Two larger the disease duration (1.22 [1.04–1.43]; P ϭ and only one patient had concomitant hy- studies in nonselected adult diabetic pa- 0.0087). Ichthyosiform skin changes were pertension. It is interesting that in the Ital- tients demonstrated mycologically significantly related only to keratosis pilaris ian series (2) of 64 patients with type 1 proven dermatophyte infections in 4 (1.53 [1.09–1.79]; P Ͻ 0.001). There diabetes, rubeosis faciei was not seen in and 31% of patients, respectively, but was no association of the mean values of any of the patients. the prevalence was similar or higher in triglycerides, fasting cholesterol, thy- Diabetic hand, a variety of scleroder- control subjects (15,16). However, Yo- roid-stimulating hormone levels, or moid lesions in diabetic patients, was seen sipovitch et al. (3) reported tinea pedis blood pressure with any of the cutane- in only 2.3% of our patients but was in 36% of their young type 1 diabetic ous disorders. strongly related to diabetes duration (P Ͻ patients versus only 7% in control subjects. 0.001). In young patients with type 1 di- On balance, contrary to a common belief, it CONCLUSIONS — This study was abetes with a three-times-longer disease seems that dermatophyte and canidida in- performed within a young population of duration, the prevalence of scleroderma- fections are not more common in patients type 1 diabetic patients. Disease duration like changes of the hand was 39% (3). All with diabetes in general as compared with and age of patients were lower than in two of our five patients with diabetic hand healthy individuals. similar cross-sectional studies (4 vs. 10 were boys aged Ͼ14 years with mean di- Cutaneous reactions to insulin ther- and 13 years and 12 vs. 22 and 23 years, abetes duration of 9 years. Previous stud- apy (lipoatrophy and lipohypertrophy) respectively). This is certainly a reason for ies (10) have shown a correlation of these were observed in only six patients in our the low prevalence of cutaneous disorders changes with the duration of diabetes and series (four with lipohypertrophy associated with diabetes. The exception is increasing age. The association with mi- [1.8%]). In the largest study of skin dis- xerosis, found in 22% of patients vs. 3% crovascular complications could not be orders in young type 1 diabetic patients, of control subjects. In the series of Ro- assessed, as they were extremely rare in these changes were seen in 6.5% of pa- mano et al. (2), only 6% of their patients our young patient population. It seems tients (3). The prevalence of lipohypertro- had xerosis. Yosipovitch et al. (3) regis- that scleroiderma-like lesions of the hand phy in type 1 diabetic patients in two tered 48% of the patients with ichthyosi- are a late of type 1 diabetes, studies specifically looking at the disorder form shin changes in their series of at least if compared with ichthyosiform was very high (29 and 48%) (17,18). de patients. Despite possible definition dis- skin changes. Advanced glycosylation is Villiers (19), in his 33 young patients with crepancies across the studies, it is clear believed to underlie the connective tissue type 1 diabetes, using careful palpation, that the skin dryness is one of the earliest changes in this disorder (3,11). not only visible lesions, found the preva- and most common manifestations of type We found necrobiosis lipoidica in lence of 52% and even 80% when only 1 diabetes. The clinical observations are 2.3% (n ϭ 5) of our patients with type 1 patients examined on several occasions supported by objective findings of a re- diabetes (mean age 14 years, mean disease were counted. On the other hand, Ro- duced hydration state of the stratum cor- duration 7 years; four girls), but no signif- mano et al. (2) in his series of type 1 dia- neum and decreased sebaceous gland icant relationship to disease duration, age, betic patients did not mention a single activity in patients with diabetes, without or metabolic control was confirmed. patient with either lipohypertrophy or li- any impairment of the stratum corneum Studies (8,12–14) indicated the preva- poatrophy. Lypohypertrophy is a conse- barrier function (6). Even in the absence lence of 0.3–1.2% among diabetic pa- quence of repeated insulin injections at of clinically apparent xerosis, patients tients, and two-thirds of them had type 1 the same sites and may complicate treat- with diabetes have an impaired desqua- diabetes. In young subjects with type 1 ment leading to delayed insulin absorp- mation process (7). Similarly to a previ- diabetes, necrobiosis lipoidica was found tion. The low number of patients with ous series of young patients with type 1 in 1.6% of the patients (3). We cannot lipohypertrophy in our series may result diabetes (3), ichthyosiform lesions in our account for the higher prevalence of from a failure to recognize subtle changes patients were not associated with known necrobiosis lipoidica in our population of amenable only to careful palpation, but

1966 DIABETES CARE, VOLUME 30, NUMBER 8, AUGUST 2007 Pavlovic´ and Associates still the prevalence of visible lesions course of the disease and do not correlate athy Committee. Diabetes Care 20:836– should range from 15 to 50% according to with its duration. In our study, the type 1 843, 1997 previous studies (18,19). The low preva- diabetic patient population was younger 6. Sakai S, Kikuchi K, Satoh J, Tagami H, lence of cutaneous reactions to insulin and had shorter diabetes duration than Inoue S: Functional properties of the stra- therapy in our patients might be a result those described in previous studies, did tum corneum in patients with diabetes mellitus: similarities to senile xerosis. Br J of regular education of our patients and not have retinopathy and neuropathy, Dermatol 153:319–323, 2005 their parents about adequate injection and had a very low prevalence of ne- 7. Yoon HS, Baik SH, Oh CH: Quantitative technique with a division of main injec- phropathy and hypertension. This may measurement of desquamation and skin tion areas in several zones (e.g., abdomen explain the rarity of other diabetes- elasticity in diabetic patients. Skin Res in six to nine zones). associated cutaneous manifestations. In Technol 8:250–254, 2002 Keratosis pilaris proved to be signifi- comparing mean disease duration in our 8. Huntley AC: Cutaneous manifestations in cantly more prevalent among our type 1 population and the Israeli population (3) diabetes: general considerations. In The diabetic patients than in control subjects of type 1 diabetic patients (4.2 vs. 13 Skin in Diabetes. Jelinek JE, Ed. Philadel- (11.7 vs. 1.5%; 2P Ͻ 0.01). This disorder years), it may be supposed that many of phia, Lea & Febiger, 1986, p. 23–30 the diabetes-related skin disorders begin 9. Jelinek JE: The skin in diabetes. Diabet showed, in a stepwise logistic regression Med 10:201–213, 1993 analysis, a highly significant association to develop after the 5th year of diabetes 10. Ngo BT, Hayes KD, DiMiao DJ, Srinivasan with ichthyosiform skin changes. This is (In 71% of our patients, the duration of Ͻ SK, Huerter CJ, Rendell MS: Manifesta- in line with findings of the Israeli group diabetes was 5 years.) tions of cutaneous diabetic microangiopa- (3,20) linking keratosis pilaris to a high A prospective arm of the study ana- thy. Am J Clin Dermatol 6:225–237, 2005 BMI, dry skin, and atopic eczema. In our lyzing cutaneous manifestations in the 11. Lyons TJ, Kennedy L: Nonenzymatic gly- series, only one patient had atopic ec- same cohort of young patients after sev- cosylation of skin in patients zema, and, unfortunately, we did not cal- eral years would offer additional data on with type I (insulin dependent) diabetes culate the BMI of our patients. The age of the natural history of skin lesions in pa- mellitus and limited joint mobility. Dia- betologia 28:2–5, 1985 a majority of our type 1 diabetic patients tients with type 1 diabetes in relation to 12. Ahmed I, Goldstein B: Diabetes mellitus. with keratosis pilaris (21 of 27) was Ͼ10 the disease duration and microvascular complications. The frequency and type of Clin Dermatol 24:237–246, 2006 years. There was no preponderance of ad- 13. Lowitt MH, Dover JS: Necrobiosis li- olescent girls in our patients with kerato- cutaneous lesions justify the early inclu- poidica. J Am Acad Dermatol 25:735, 1991 sis pilaris, as suggested in some studies sion of a dermatologist in the manage- 14. Muller SA, Winkelman RK: Necrobiosis (21). It seems that skin xerosis plays a ment of type 1 diabetic patients in order lipoidica diabeticorum, a clinical and significant role in the development of pi- to recognize and treat the disorders and pathological investigation of 171 cases. lar keratosis. direct further research into prevention Arch Derm 93:272–281, 1966 Frequencies of other dermatologic some of the lesions, especially in those 15. Lugo-Somolinos A, Sanchez JL: Preva- conditions, unrelated to diabetes, were with an underlying fibrosing process. lence of dermatophytosis in patients with diabetes. J Am Acad Dermatol 26:408– not different between patients and con- 410, 1992 trol subjects (Table 2). When, like with References 16. Romano C, Massai L, Asta F, Signorini halo nevi, a condition was present only 1. Ferringer T, Miller F 3rd: Cutaneous AM: Prevalence of dermatophytic skin in type 1 diabetic patients, the numbers manifestations of diabetes mellitus. Der- and infections in diabetic patients. were too small to draw any conclusion. matol Clin 20:483–492, 2002 Mycoses 44:83–86, 2001 We have not observed many other skin 2. Romano G, Moretti G, Di Benedetto A, 17. Hauner H, Stockamp B, Haastert B: Prev- disorders associated with diabetes, such Giofre` C, Di Cesare E, Russo G, Califano alence of lipohypertrophy in insulin- as generalized thick skin, diabetic der- L, Cucinotta D: Skin lesions in diabetes treated diabetic patients and predisposing mopathy, yellow hands, acanthosis nig- mellitus: prevalence and clinical correla- factors. Exp Clin Endocrinol Diabetes 104: 106–110, 1996 ricans, annulare, multiple tions. Diabetes Res Clin Pract 39:101–106, 1998 18. Kordonouri O, Lauterborn R, Deiss D: Li- skin tags, or ulcers. The most likely ex- 3. Yosipovitch G, Hodak E, Vardi P, Shraga pohypertrophy in young patients with planation is the relatively short duration I, Karp M, Sprecher E, David M: The prev- type 1 diabetes (Letter). Diabetes Care 25: of diabetes in our pediatric and adoles- alence of cutaneous manifestations in 634, 2002 cent patients with type 1 diabetes com- IDDM patients and their association with 19. de Villiers FP: Lipohypertrophy: a com- pared with previous cross-sectional diabetes risk factors and microvascular plication of insulin injections. S Afr Med J studies (2,3). complications. Diabetes Care 21:506– 95:858–859, 2005 Results of the cross-sectional study 509, 1998 20. Yosipovitch G, Mevorah B, Mashiach J, confirm that the most prevalent cutane- 4. Seibold JR: Digital sclerosis in children Chan YH, David M: High body mass in- ous manifestation in young patients with with insulin dependent diabetes mellitus. dex, dry scaly leg skin and atopic condi- Arth Rheum 25:1355–1361, 1982 tions are highly associated with keratosis type 1 diabetes is dry skin. As suggested 5. Fedele D, Comi G, Coscelli C, Cucinotta pilaris. Dermatology 201:34–36, 2000 by Yosipovitch et al. (3), keratosis pilaris D, Feldman EL, Ghirlanda G, Greene DA, 21. Barth JH, Wojnarowska F, Dawber RP: Is is also a condition disproportionately rep- Negrin P, Santeusanio F: A multicentre keratosis pilaris another androgen depen- resented in the type 1 diabetic popula- study on the prevalence of diabetic neu- dent dermatosis. Clin Exp Dermatol 13: tion. Both conditions appear early in the ropathy in Italy: Italian Diabetic Neurop- 240–241, 1988

DIABETES CARE, VOLUME 30, NUMBER 8, AUGUST 2007 1967