Diabetic nephropathy (DN) is a primary cause of. Population-Based Assessment of Familial Clustering of Diabetic Nephropathy in Type 1 Diabetes

Transcription

1 Population-Based Assessment of Familial Clustering of Diabetic Nephropathy in Type 1 Diabetes Valma Harjutsalo, 1 Shuichi Katoh, 1,2 Cinzia Sarti, 1 Naoko Tajima, 2 and Jaakko Tuomilehto 1,3 We determined the magnitude of familial aggregation in the development of diabetic nephropathy (DN) among a population-based cohort of Finnish type 1 diabetic patients. Probands with type 1 diabetes were identified from the nationwide register of all Finnish cases diagnosed during By 1998, there were 537 families with at least two siblings with type 1 diabetes. These 537 probands and their 616 diabetic siblings were followed for a diagnosis of DN until the end of We identified 323 cases of DN in these families. If the proband had DN, 38% of the siblings also had DN, whereas out of the diabetic siblings of the probands without DN, only 17% had DN (P 0.001). Diabetic siblings of the nephropathic probands had a 2.3 times (95% CI ) higher risk of DN compared with siblings of probands free of DN. The presence of a severe form of DN in the proband increases the risk threefold for diabetic siblings. Sex, the DN of the proband, the age at the onset of diabetes, and parental type 2 diabetes were significant predictors of DN among diabetic siblings. Although the majority of sibpairs with type 1 diabetes are discordant for DN, its presence in one sibling doubles the risk for the other diabetic siblings. Diabetes 53: , 2004 Diabetic nephropathy (DN) is a primary cause of excess mortality in type 1 diabetic patients (1). It affects about one-third of diabetic patients (2,3), increases the risk of cardiovascular diseases, and may lead to renal failure (1,4). The incidence of DN increases linearly during the first 20 years of the duration of diabetes, but starts to decline thereafter (2,3,5), suggesting that there may be a subset of diabetic patients at an especially high risk for DN. Not all patients with poor glycemic control develop DN during their lifetimes. The observed incidence patterns suggest that genetic factors linked to a predisposition for DN play an important role in the regulating processes that lead to DN. Familial clustering of DN has been previously reported From the 1 Diabetes and Genetic Epidemiology Unit, Department of Epidemiology and Health Promotion, National Public Health Institute, Helsinki, Finland; the 2 Division of Diabetes, Metabolism, and Endocrinology, Department of Internal Medicine, Jikei University School of Medicine, Tokyo, Japan; and the 3 Department of Public Health, University of Helsinki, Helsinki, Finland. Address correspondence and reprint requests to Valma Harjutsalo, National Public Health Institute, Mannerheimintie 166, FIN Helsinki, Finland. valma.harjutsalo@ktl.fi. Received for publication 5 March 2004 and accepted in revised form 26 May CDR, Central Drug Register; DN, diabetic nephropathy; ESRD, end-stage renal disease; HDR, Hospital Discharge Register by the American Diabetes Association. (6 8). Whether it is a consequence of shared environmental risk factors, genetic factors, or both is unclear. Thus far, no population-based studies on the familial clustering of DN have been published. Previous studies on DN within families have been based on small, hospital-based series. The aim of this study was to analyze the familial aggregation of DN in a large, Finnish, population-based cohort of type 1 diabetic patients and their siblings, avoiding the bias of a family ascertainment. RESEARCH DESIGN AND METHODS We identified families with two or more siblings with type 1 diabetes and followed the diagnoses of DN in the families until the end of The original cohort consisted of diabetic subjects diagnosed before the age of 18 years between 1965 and 1979 (n 5,126), who were included in the nationwide register of Finnish type 1 diabetic patients (9,10). This register was initially based on the Social Insurance Institution Central Drug Register (CDR), which lists patients approved to receive free-of-charge medication for certain diseases, including diabetes (11,12). Their siblings (n 10,168) were identified through the National Population Register of Finland with a computer search using the unique personal identifier (ID) that is assigned to all residents of Finland. The diabetes status of the siblings was ascertained through several sources: from the nationwide Hospital Discharge Register (HDR) for the years , from the nationwide Finnish Diabetes Register for children and young adults for the years , and from the CDR through a record search using the ID. Because of these multiple data sources, the case ascertainment was virtually complete. The diagnosis of diabetes type and the date of onset were obtained by reviewing of the medical records for all probands and their diabetic siblings. By the end of the year 1998, a total of 537 families that included 616 siblings with type 1 diabetes were identified among the original Diabetes Epidemiology Research International (DERI) cases. In the present study, the proband was defined as the sibling who was first diagnosed with diabetes within the sibship. In order to identify patients with DN, copies of original medical records, death certificates, and autopsy data for the probands and siblings were systematically reviewed by one of the coauthors (V.H.). Overt nephropathy was defined based on the American Diabetes Association s criteria (13): when a patient repeatedly had either a urinary albumin excretion rate of 200 g/min or 300 mg/24 h, a 24-h urinary protein excretion rate of 0.5 g, or a positive urinalysis for protein using a reagent strip. Microalbuminuria was defined as a urinary albumin excretion rate of g/min or mg/24 h. Albumin elevations because of pregnancy, urinary tract infections, or other renal diseases alone were not considered as diagnostic for DN. The urinary albumin excretion rate had decreased in some patients because of the initiation of antihypertensive medication; in such cases, the classification of DN was based on findings before the initiation of drug treatment. Microalbuminuric patients were grouped together with normoalbuminuric subjects in all analyses. End-stage renal disease (ESRD) was diagnosed if the patient was undergoing dialysis or had received a kidney transplant. To study whether a parental history of type 2 diabetes and hypertension are the risk markers of DN in siblings, data from the parents was linked to the HDR, the CDR, and the National Death Registry. The date of the diagnosis of diabetes was defined as the date of the first hospital admission because of diabetes or the approval date for the free-of-charge medicine for diabetes, whichever was earlier. Parents with type 2 diabetes were defined as those whose age at the diagnosis of diabetes was 40 years. There were 85 parents out of 537 families with type 2 diabetes and 243 parents with hypertension. DIABETES, VOL. 53, SEPTEMBER

2 FAMILIAL CLUSTERING OF DIABETIC NEPHROPATHY TABLE 1 of DN in siblings of probands with type 1 diabetes per 1,000 person-years in each 5-year time band since the onset of diabetes according to sex Men Women Duration of diabetes (years) Men-to-women rate ratio 0 4 1, ( ) 1, ( ) , ( ) 1, ( ) , ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) 2.6 Statistical methods. For each sibling, the person-years at risk was calculated from the onset of diabetes to the date of DN diagnosis, death, or until the last urine screening test found in the medical records before the end of the year 2001 (the end of the follow-up). The 95% CIs for incidence rates (per 1,000 person-years) were calculated assuming a Poisson distribution of the cases. The longterm progression of DN according to the duration of diabetes was estimated by using the Kaplan-Meier survival analysis method. The log-rank test served to test the risk differences between subgroups. To identify the most important prognostic factors for the development of DN in diabetic siblings, Cox s proportional hazards model was fitted to the data. Sex, the age at the onset of diabetes, the calendar year at the onset of diabetes, the nephropathy status of the proband, the parental history of type 2 diabetes and hypertension, and maternal age at delivery were entered into the model. The first-order interactions between variables were also examined. In order to study the influence of the calendar period on the occurrence of DN, siblings were stratified according to the time of the onset of diabetes into three groups: in the year 1970 or before, from 1971 to 1979, and from 1980 onward. Additional stratified analyses were carried out for the age at the onset of diabetes in the siblings: 0 4 years, 5 9, 10 14, and 15. RESULTS The average duration of diabetes (until death or to the end of the follow-up) was years among the probands and years among the diabetic siblings. Of the probands, 97% had suffered from diabetes for 20 years and none of those alive at the end of the follow-up had a duration of diabetes of 22 years. Among siblings, 60% have had diabetes for at least 20 years and 87% have had a duration of diabetes of 10 years. The total number of person-years with diabetes in siblings during the follow-up was 10,913. By the end of 2001, 323 cases of DN were identified, 176 among the probands (33%) and 147 among diabetic siblings (24%), with a total of 274 sibpairs affected. Of these, 77 were concordant sibpairs (both affected with DN) and 197 pairs were discordant sibpairs. Of the 204 siblings from families in which the proband had DN, 77 (38%) also had nephropathy, whereas among the 409 siblings of the probands without DN only 70 (17%) diabetic siblings had nephropathy. Seventy-seven probands were microalbuminuric, and 44 (57%) of them were treated with antihypertensive medications. Among the siblings, the corresponding numbers were 65 and 36 (55%). Fourteen sibpairs were concordant for microalbuminuria. The first case of DN in both sexes occurred after 5 years from the onset of diabetes. The peak incidence was seen at years after the onset of diabetes among women (24.2 per 1,000 person-years) and at years among men (41.1 per 1,000 person-years); it decreased thereafter (Table 1). Table 2 shows the age-specific incidence rates of DN by the nephropathy status of the proband. The risk of DN among diabetic siblings was significantly increased (P ) when the proband also suffered from nephropathy. The incidence of DN rose gradually and reached the peak TABLE 2 of DN per 1,000 person-years for the siblings of probands with type 1 diabetes according to nephropathy status in the proband and attained age Proband without DN Proband with DN Attained age (years) DN -to-dn rate ratio , ( ) , ( ) ( ) , ( ) ( ) , ( ) ( ) , ( ) ( ) ( ) ( ) ( ) ( ) 4.6 All 6, ( ) 3, ( ) 2.3 DN, presence of nephropathy and includes ESRD; DN, absence of nephropathy DIABETES, VOL. 53, SEPTEMBER 2004

3 V. HARJUTSALO AND ASSOCIATES TABLE 3 of DN per 1,000 person-years in the diabetic siblings of probands with type 1 diabetes according to sex and attained age Men Women Attained age (years) Men-to-women rate ratio ( ) ( ) ( ) ( ) ( ) , ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) 5.6 All 5, ( ) 4, ( ) 1.6 at ages of years, decreasing thereafter in the siblings of the probands without DN, whereas new nephropathy cases continued to appear beyond these ages in the siblings of the probands with DN (Table 2). The male-tofemale rate ratio for DN was close to 1 up to the age of 29 years. After this age, it began to increase to 5.6 at age 40 years and above (Table 3). Overall, the male-to-female rate ratio was 1.6. The independent prognostic factors for the development of DN were sex, the age at the diagnosis of diabetes, the nephropathy status of the proband, and parental type 2 diabetes in the multivariate analysis using Cox s regression model. Maternal age at delivery, parental hypertension, and the year of the diagnosis of diabetes were not significant predictors of DN in this model. The presence of type 2 diabetes in parents increased the risk of DN in siblings 1.6-fold (95% CI ). Even though not statistically significant, perhaps due to small numbers, there was some evidence of the interaction between sex and the age at diagnosis of diabetes in siblings (P 0.07). The risk pattern was different according to onset age of diabetes in men and women. The male-to-female rate ratio was equal in the age of onset of diabetes groups 0 4 and 5 9 years, increasing with increasing age at onset of diabetes after that. No other first-order interactions were detected. The overall 25-year cumulative risk of DN was 34.5% (95% CI ), but nephropathy status and severity of DN in probands significantly influenced the risk of DN in siblings (P , log-rank test). The corresponding risk for the siblings of probands without nephropathy was 24.8% ( ), for the siblings of the probands with nephropathy 43.2% ( ), and for the siblings of the probands with ESRD 58.0% ( ) (Fig. 1). The risk ratio for DN in the siblings of the probands with DN was 2.0 ( ) and in the siblings of the probands with ESRD was 2.9 ( ) compared with the siblings of normo- or microalbuminuric probands. The risk ratio was 2.3 ( ) when the proband s nephropathy status and the ESRD groups were combined. Figure 2 shows the development of DN in siblings stratified by age at the onset of diabetes. Subjects diagnosed with diabetes between the ages of 5 9 years and of years had the highest risk of developing DN, whereas subjects diagnosed at a very young age or after puberty had a lower risk. The 20-year progression rate of DN for the siblings diagnosed with diabetes at the ages of 4 years, 5 9, 10 14, and 15 were 18.7% ( ), 28.2 ( ), 30.7 ( ), and 15.9 ( ), respectively (P 0.001, log-rank test). ESRD was present in 46 (8.6%) probands and in 36 (5.9%) diabetic siblings. There were 10 concordant pairs for ESRD (13.9% of all sibpairs with ESRD). The risk of ESRD in siblings after a 25-year duration of diabetes was 8.9% (95% CI ). No significant difference was found in the ESRD risk between male and female siblings. DISCUSSION Our study represents the first longitudinal populationbased study of DN among siblings of type 1 diabetic patients. It covered 10,937 person-years of observation and 23 years of the median duration of diabetes in diabetic siblings. We estimated for the first time the magnitude of the familial clustering of DN free of a family-ascertainment bias and the influence of the severity of the disease in the proband. This study confirms a previous suggestion that a strong familial clustering exists for DN. DN in the proband increases the risk twofold for the other diabetic sibling. This estimate is slightly lower than those suggested by other smaller studies (7,8), in which the odds ratios were between 2.5 and 4.9. Another study that was restricted to severe cases of DN in the proband (selected from the kidney transplantation register) showed a remarkable 24-times higher risk of DN for diabetic siblings (6). This high odds ratio may have been influenced by a small sample size and a selection bias toward sibpairs concordant for DN and probably the aggregation of disadvantageous genotypes associated with the severity of the renal involvement (14). When analyses were confined to the siblings of the probands with ESRD in the present study, the recurrence risk increased but still was only 2.9-fold. In the present study, we categorized all subjects with microalbuminuria as unaffected and only subjects with macroalbuminuria were categorized as nephropathic. If the proband had microalbuminuria, the risk estimate for siblings progressing to overt proteinuria was similar to the siblings of the normoalbuminuric probands (data not shown), in agreement with the results of Quinn et al. (7). Although microalbuminuria has a predictive value for the risk of nephropathy, recent investigations suggest that it is not an inexorable process leading to overt proteinuria; only 10 30% of microalbuminuric patients develop macroalbuminuria during a long follow-up (15 18). In light of these recent findings, supported also by our analyses, we DIABETES, VOL. 53, SEPTEMBER

4 FAMILIAL CLUSTERING OF DIABETIC NEPHROPATHY FIG. 1. Kaplan-Meier plot of the probability of remaining free of DN in diabetic siblings of type 1 diabetic probands according to the nephropathy status of probands. The difference between groups was statistically highly significant (P < , log-rank test). *Siblings with unknown DN status (n 3) and siblings with unknown DN status in the proband (n 13). DN, absent nephropathy; DN, present nephropathy and excludes ESRD. concluded that we can pool microalbuminuric patients together with normoalbuminuric ones. Familial aggregation has been shown for many chronic noncommunicable diseases. The nature of the factors responsible for the familial aggregation of DN is unclear. Despite inherited genetic factors, there might be more similar lifestyles, eating and smoking habits, and glycemic control among diabetic siblings than among unrelated patients. On the basis of a simulation study done by Khoury et al. (19), it was concluded that without any genetic susceptibility, familial clustering of high environmental risk is unlikely to fully explain the aggregation of a disease among the siblings and that genetic factors are the most likely cause for familial aggregation. Glycemic control has only been considered to be a nongenetic risk factor for DN in general, and thus the similarity in glycemic control among sibpairs is interpreted as a shared environmental factor. Snieder et al. (20) have shown in a recent twin study that HbA 1c levels are genetically determined; additive genetic effects explained 62% of the population variance of HbA 1c, with the remaining effects being influenced by a unique environmental effect and age. Genes associated with HbA 1c may play a role in a background of DN predisposition. On the other hand, observations that some patients with relatively good glycemic control may develop DN and some patients with poor glycemic control do not (21,22) support the hypothesis FIG. 2. Kaplan-Meier plot of the probability of remaining free of DN in diabetic siblings of type 1 diabetic probands according to age at onset of diabetes in siblings. The difference between groups was statistically highly significant (P 0.001, log-rank test). *Siblings with unknown DN status (n 3). that additional factors unrelated to glycemia must be operating in the development of DN (23). Parental history of hypertension, type 2 diabetes, cardiovascular diseases, and insulin resistance appear to be risk factors of DN in diabetic patients, suggesting that DN may be linked by a complex interrelated genetic predisposition to those disorders (24 26). A biopsy study done by Fioretto et al. (23) also highlighted the importance of the genetic basis for DN; they found a strong concordance in the severity and patterns of glomerular lesions between diabetic siblings, despite a lack of concordance for glycemia. However, in several studies the role of most gene candidates for DN has remained controversial, and the etiology of DN is still unclear. We did not genotype these patients, but various types of studies searching for the genes for a predisposition for DN are ongoing within our group. It has been speculated that the peripubertal years with diabetes may have a greater impact on the risk of developing microvascular complications (27) and premature death (28) than the years before puberty. Few studies (2,27,29 33) have distinguished the early childhood years from the years during the transition from prepuberty to puberty or have focused on early signs of microvascular complications. In our study population, diabetic siblings diagnosed at a very early age, 4 years, had the lowest risk of developing DN, whereas the risk for siblings whose age 2452 DIABETES, VOL. 53, SEPTEMBER 2004

5 V. HARJUTSALO AND ASSOCIATES of onset of diabetes was between 5 and 9 years was similar to the peripubertal risk. In a recent study, Donaghue et al. (34) reported that subjects diagnosed before the age of 5 years had a delayed onset of early retinopathy and microalbuminuria compared with those diagnosed after 5 years of age, in keeping with our results. It has been reported that glycemic control tends to deteriorate during the pubertal years due to dramatic physiological and psychosocial changes (35). Hormonal changes start their influence years before visible pubertal changes. Among others, growth hormone and adrenal androgen levels are relatively low before the age of 5 years, after which the latter gradually increases while growth hormone levels rise after the onset of gonadarche (36,37). However, children who were younger than 5 years at the onset of diabetes experience the same hormonal changes during puberty as those diagnosed with diabetes later in childhood. Additionally, it seems that children with an onset of diabetes at a very young age do not maintain good glycemic control during the pubertal years (38), although there are few longitudinal studies that address this question. It is, however, unclear if hormonal changes of the years before and during puberty occurring simultaneously with the onset of diabetes have an impact on the increased risk of DN later in life. The incidence of DN reached a peak around the age of years for both sexes. There were only a few cases of DN in diabetic sisters after the age of 35 years, in contrast to diabetic brothers, for whom DN occurred throughout the entire age range covered by our study. This difference in the age at onset of DN between sexes was not a consequence of a different duration of diabetes; it was slightly but not significantly longer in women (median 20) than in men (median 18) at the attained age of 35 years. Women tended to develop DN earlier after the onset of diabetes compared with men; the peak in incidence occurred during the first 20 years of diabetes in women, whereas in men it occurred 5 years later. This may be due in part to an earlier growth spurt in women than in men. It is also possible that factors for DN are to some extent different in men than in women (39,40); for example, sex hormones might be relevant, as indicated in animal models (41). In conclusion, we have confirmed in the populationbased sample that the majority of sibpairs with type 1 diabetes are discordant for DN. Its presence in one sibling increases the risk twofold for the other diabetic siblings. If the proband had an advanced stage of DN, the risk of DN in siblings was threefold. Although the causes of familial clustering of DN still remain unclear, genetic factors seem to be strongly involved. Being diagnosed with diabetes during puberty or a few years before, being of male sex, and having a sibling already diagnosed with DN are markers of an increased susceptibility for DN in siblings of type 1 diabetic patients. ACKNOWLEDGMENTS This study was supported by grants from the Academy of Finland (46558), the National Institutes of Health (DK/ AG63045), the Novo Nordisk Foundation, and Munuaissäätiö (V.H.). The work of S. K. was supported by a grant from Uehara Memorial Foundation in Japan. REFERENCES 1. Borch-Johnsen K, Kreiner S: Proteinuria: value as predictor of cardiovascular mortality in insulin dependent diabetes mellitus. Br Med J (Clin Res Ed) 294: , Krolewski AS, Warram JH, Christlieb AR, Busick EJ, Kahn CR: The changing natural history of nephropathy in type I diabetes. Am J Med 78: , Andersen AR, Christiansen JS, Andersen JK, Kreiner S, Deckert T: Diabetic nephropathy in type 1 (insulin-dependent) diabetes: an epidemiological study. Diabetologia 25: , Tuomilehto J, Borch-Johnsen K, Molarius A, Forsen T, Rastenyte D, Sarti C, Reunanen A: of cardiovascular disease in type 1 (insulindependent) diabetic subjects with and without diabetic nephropathy in Finland. Diabetologia 41: , Diabetes Control and Complications Trial Research Group: The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 329: , Seaquist ER, Goetz FC, Rich S, Barbosa J: Familial clustering of diabetic kidney disease: evidence for genetic susceptibility to diabetic nephropathy. N Engl J Med 320: , Quinn M, Angelico MC, Warram JH, Krolewski AS: Familial factors determine the development of diabetic nephropathy in patients with IDDM. Diabetologia 39: , Borch-Johnsen K, Norgaard K, Hommel E, Mathiesen ER, Jensen JS, Deckert T, Parving HH: Is diabetic nephropathy an inherited complication? Kidney Int 41: , Diabetes Epidemiology Research International Mortality Study Group: International evaluation of cause-specific mortality and IDDM. Diabetes Care 14:55 60, Diabetes Epidemiology Research International Mortality Study Group: Major cross-country differences in risk of dying for people with IDDM. Diabetes Care 14:49 54, Tuomilehto J, Rewers M, Reunanen A, Lounamaa P, Lounamaa R, Tuomilehto-Wolf E, Åkerblom HK: Increasing trend in type 1 (insulin-dependent) diabetes mellitus in childhood in Finland: analysis of age, calendar time and birth cohort effects during 1965 to Diabetologia 34: , Lounamaa R: Mortality in Finnish Patients With Insulin-Dependent Diabetes Mellitus: A Follow-Up Study of Patients Diagnosed When Under Twenty Years of Age. Helsinki, Finland, Social Insurance Institution, American Diabetes Association: Diabetic nephropathy (Position Statement). Diabetes Care 26 (Suppl. 1):S94 S98, Rogus JJ, Moczulski D, Freire MB, Yang Y, Warram JH, Krolewski AS: Diabetic nephropathy is associated with AGT polymorphism T235: results of a family-based study. Hypertension 31: , Forsblom CM, Groop PH, Ekstrand A, Groop LC: Predictive value of microalbuminuria in patients with insulin-dependent diabetes of long duration. BMJ 305: , Tabaei BP, Al-Kassab AS, Ilag LL, Zawacki CM, Herman WH: Does microalbuminuria predict diabetic nephropathy? Diabetes Care 24: , Almdal T, Norgaard K, Feldt-Rasmussen B, Deckert T: The predictive value of microalbuminuria in IDDM: a 5-year follow-up study. Diabetes Care 17: , Perkins BA, Ficociello LH, Silva KH, Finkelstein DM, Warram JH, Krolewski AS: Regression of microalbuminuria in type 1 diabetes. N Engl J Med 348: , Khoury MJ, Beaty TH, Liang KY: Can familial aggregation of disease be explained by familial aggregation of environmental risk factors? Am J Epidemiol 127: , Snieder H, Sawtell PA, Ross L, Walker J, Spector TD, Leslie RD: HbA 1c levels are genetically determined even in type 1 diabetes: evidence from healthy and diabetic twins. Diabetes 50: , Kullberg CE, Arnqvist HJ: Glycemic control in patients with type I diabetes and normoalbuminuria after long diabetes duration. J Diabetes Complications 11: , Zhang L, Krzentowski G, Albert A, Lefebvre PJ: Factors predictive of nephropathy in DCCT type 1 diabetic patients with good or poor metabolic control. Diabet Med 20: , Fioretto P, Steffes MW, Barbosa J, Rich SS, Miller ME, Mauer M: Is diabetic nephropathy inherited? Studies of glomerular structure in type 1 diabetic sibling pairs. Diabetes 48: , Fagerudd JA, Pettersson-Fernholm KJ, Grönhagen-Riska C, Groop PH: The impact of a family history of type II (non-insulin-dependent) diabetes mellitus on the risk of diabetic nephropathy in patients with type I (insulindependent) diabetes mellitus. Diabetologia 42: , 1999 DIABETES, VOL. 53, SEPTEMBER

6 FAMILIAL CLUSTERING OF DIABETIC NEPHROPATHY 25. Rippin JD, Patel A, Bain SC: Genetics of diabetic nephropathy. Best Pract Res Clin Endocrinol Metab 15: , Chowdhury TA, Dyer PH, Kumar S, Barnett AH, Bain SC: Genetic determinants of diabetic nephropathy. Clin Sci (Lond) 96: , Kostraba JN, Dorman JS, Orchard TJ, Becker DJ, Ohki Y, Ellis D, Doft BH, Lobes LA, LaPorte RE, Drash AL: Contribution of diabetes duration before puberty to development of microvascular complications in IDDM subjects. Diabetes Care 12: , Nishimura R, Tajima N, Matsushima M, LaPorte RE: Puberty, IDDM, and death in Japan: Diabetes Epidemiology Research International Study Group. Diabetes Care 21: , Donaghue KC, Fung AT, Hing S, Fairchild J, King J, Chan A, Howard NJ, Silink M: The effect of prepubertal diabetes duration on diabetes: microvascular complications in early and late adolescence. Diabetes Care 20:77 80, Schultz CJ, Konopelska-Bahu T, Dalton RN, Carroll TA, Stratton I, Gale EA, Neil A, Dunger DB: Microalbuminuria prevalence varies with age, sex, and puberty in children with type 1 diabetes followed from diagnosis in a longitudinal study: Oxford Regional Prospective Study Group. Diabetes Care 22: , McNally PG, Raymond NT, Swift PG, Hearnshaw JR, Burden AC: Does the prepubertal duration of diabetes influence the onset of microvascular complications? Diabet Med 10: , Holl RW, Lang GE, Grabert M, Heinze E, Lang GK, Debatin KM: Diabetic retinopathy in pediatric patients with type-1 diabetes: effect of diabetes duration, prepubertal and pubertal onset of diabetes, and metabolic control. J Pediatr 132: , Lane H: Diabetic kidney disease: impact of puberty. Am J Physiol Renal Physiol 283:F589 F600, Donaghue KC, Fairchild JM, Craig ME, Chan AK, Hing S, Cutler LR, Howard NJ, Silink M: Do all prepubertal years of diabetes duration contribute equally to diabetes complications? Diabetes Care 26: , Hamilton J, Daneman D: Deteriorating diabetes control during adolescence: physiological or psychosocial? J Pediatr Endocrinol Metab 15: , Korth-Schutz S, Levine LS, New MI: Serum androgens in normal prepubertal and pubertal children and in children with precocious adrenarche. J Clin Endocrinol Metab 42: , Albertsson-Wikland K, Rosberg S, Karlberg J, Groth T: Analysis of 24-hour growth hormone profiles in healthy boys and girls of normal stature: relation to puberty. J Clin Endocrinol Metab 78: , Kordonouri O, Danne T, Enders I, Weber B: Does the long-term clinical course of type I diabetes mellitus differ in patients with prepubertal and pubertal onset? Results of the Berlin Retinopathy Study. Eur J Pediatr 157: , Monciotti CG, Semplicini A, Morocutti A, Maioli M, Cipollina MR, Barzon I, Palaro C, Brocco E, Trevisan M, Fioretto P, Crepaldi G, Nosadini R: Elevated sodium-lithium countertransport activity in erythrocytes is predictive of the development of microalbuminuria in IDDM. Diabetologia 40: , Connell SJ, Hollis S, Tieszen KL, McMurray JR, Dornan TL: Gender and the clinical usefulness of the albumin:creatinine ratio. Diabet Med 11:32 36, Williamson JR, Rowold E, Chang K, Marvel J, Tomlinson M, Sherman WR, Ackermann KE, Berger RA, Kilo C: Sex steroid dependency of diabetesinduced changes in polyol metabolism, vascular permeability, and collagen cross-linking. Diabetes 35:20 27, DIABETES, VOL. 53, SEPTEMBER 2004