The occurrence of microalbuminuria in a patient with type 1 diabetes is clearly indicative of an enhanced risk of nephropathy and cardiovascular disease1,2,3,4. The identification of such individuals is an important challenge to care providers. While microalbuminuria can be substantially delayed by tight glycemic control5, achieving the degree of control encouraged by the results of the Diabetes Control and Complications Trial (DCCT) remains impractical for many centers. Additionally, there are claims that there is a glycemic threshold below which the risk of progression to microalbuminuria remains static6, and although not confirmed by other reports7,8, such claims have important implications for care guidelines. It is still likely that progression to microalbuminuria will occur in a substantial proportion of patients, and therefore there is a need to obtain valid estimates of progression and to explore the role of risk factors other than glycemic control and diabetes duration that may provide further clues for novel interventions.
The EURODIAB Prospective Complications Study (PCS) is ideally placed to address some of these questions. To our knowledge this is the largest cohort study of people with type 1 diabetes, with standardized measures of albumin excretion rate (AER) at both baseline and follow-up. We determined the incidence of microalbuminuria, examined risk factors, and determined whether there was evidence for a glycemic threshold.
METHODS
A total of 3250 men and women with type 1 diabetes were recruited from 31 centers in 16 European countries, and were aged between 15 and 60 at the baseline investigation phase (1989 to 1991)9. The diagnosis of type 1 diabetes was a clinical one; diagnosis had to have occurred in the patient prior to age 36 and the patient had a continuous need for insulin within a year of diagnosis. Re-examination occurred on average of between six and eight years after baseline investigations.
At follow-up, the patient's complication status was again measured using the same protocol as baseline9. Two blood pressure measurements were taken at both visits, using the same standard random zero sphygmomanometer, and their mean value was used in all analyses. The previous two years (
8 visits) of values of locally measured HbA1c were recorded. Two 24-hour urine collections were performed (only one was performed at baseline). Patients performed collections on two consecutive days to minimize variability. Samples were tested for infection by dipstick (Nephur test), and if positive, the collections were discarded and a fresh collection was advised after any infection had been treated. If the dipstick was negative, aliquots were frozen and sent to London for analysis of urinary albumin, using an immunoturbidimetric method10 that included goat anti-human albumin antisera (Sanofi Diagnostics Pasteur Inc., MN, USA) and human serum albumin standards (ORHA 20/21 grade HSA; Behring Diagnostics, Hoechst UK Ltd., Hounslow, Middlesex, UK). The same laboratory was used for both baseline and follow-up studies. The coefficient of variation for the follow-up specimens was 34%, reflecting the known variability in this measure. Blood samples were taken, with the patient fasting if possible, for measurement of lipids. At both baseline and follow-up, these samples were sent to a central lab, and the methods used at baseline were standard enzymatic methods (Boehringer Mannheim, East Sussex, UK) on a cobas-bio centrifugal analyzer (Roche, Welwyn Garden City, Herts, UK)11,12,13. Low-density lipoprotein (LDL) cholesterol was calculated according to the Friedewald formula14. Glycosylated hemoglobin (HbA1c) was measured with an enzyme immunoassay using a monoclonal antibody against HbA1c (Dako, Ely, UK)15. The reference range for this assay is 2.9 to 4.8%. A sample was sent locally for measurement of HbA1c. At baseline, the coagulation factors fibrinogen and von Willebrand factor (vWF) were also measured16.
Statistical analysis
These analyses are restricted to those patients who were normoalbuminuric (AER <20 
g/min) at baseline. Microalbuminuria was defined as an AER of between 20 and 200 g/min, based on the arithmetic mean of the two follow-up urine collections. Two urine collections were available on 91% of patients; the value of the solitary specimen was used in the remainder. Centrally measured HbA1c was used in most analyses, but to get a measure of previous glycemic control, which was assessed locally, the following approach was used to standardize local clinic measures to the central measure. The central measurement and local measurement performed on aliquots of the same sample at follow-up were compared using linear regression by center. At worst, local measurements for an individual center were performed on 58% (35 out of 60) of all samples received centrally in London; this still was sufficient to estimate a slope between central and local HbA1c. For each center, the mean of the previous two years of data on HbA1c for each individual was calculated. The regression equation then was used to convert this local mean to a central, that is, London mean. If a center had changed lab methods for HbA1c, only those measurements using methods that we had been able to standardize were used in these analyses. For the assessment of a threshold effect and for comparisons between baseline and follow-up HbA1c, a conversion factor to the DCCT measure was calculated. This again was derived from a linear regression plot of measures of HbA1c comparing results from the London laboratory at baseline against those using the DCCT method. This was DCCT HbA1c = 1.0289 
London HbA1c + 1.5263. A similar technique was employed at follow-up so that direct comparisons between HbA1c at baseline and HbA1c could be made. This formula was: DCCT HbA1c = 0.9633 London HbA1c + 0.0709.
Baseline and follow-up characteristics for incidence were calculated using regression techniques for continuous variables and proportions for categorical variables. An adjustment was then made for those variables that were statistically significantly related to risk of microalbuminuria for HbA1c and diabetes duration, which a priori were thought to be the main risk factors for progression. A breakpoint or threshold effect for the relationship between HbA1c and microalbuminuria was tested for using a two-phase segmented-weighted regression analysis, which fits two straight lines through a series of defined points17. These points were calculated by logistic regression adjusted for diabetes duration. This segmented regression was compared with the line of best fit, using weighted linear regression. Logistic regression was used to test for a threshold effect18.
Multivariate regression models are often used to understand the relative importance of several predictive variables for a given outcome. However, the direct comparison of the size of the 
coefficients of these explanatory variables is problematic, as, for example, a one-year increase in age cannot be said to be equivalent to a 1% increase in HbA1c. Standardized regression effects (SREs) were used to overcome this limitation. SREs were calculated for continuous variables by multiplying the estimate from logistic regression models by the standard deviation of that variable; in this case, all log-transformed variables were not converted back. This allowed the direct comparison of the degree of importance of each variable by standardizing for population variance. Multivariate models were restricted to those individuals who had complete data on all included risk factors (N = 715). Much of the missing data was due to the lack of a fasting triglyceride measurement at baseline (760 out of 1115 fasting samples available), but this ratio did not differ by progressor status. Other risk factors were compared in those who did and did not have a fasting triglyceride measurement, and the risk factor relationships were found to be identical.
RESULTS
The mean follow-up was 7.3 years. Out of the 3250 patients examined at baseline, 1865 were normoalbuminuric and available for follow-up examination. Of these, data on AER at follow-up were available on 1134 patients Figure 1. Follow-up rates by center varied from 44 to 94%, but bore no association with incidence of microalbuminuria. Baseline data were compared in those who were potentially available for follow-up, but on whom there were no follow-up AER data (that is, 550 + 6 + 135, excluding those who had died) with those who did provide a urine collection at follow-up Table 1. There were no significant differences in risk factors at baseline, apart from HbA1c, which was 6.7% in those with no follow-up data and 6.3% in those with follow-up data (P = 0.0001), and fasting triglyceride (0.96 vs. 0.89 mmol/L, P = 0.004).
Table 1 - Comparison of baseline data for those with and without an albumin excretion rate (AER) measurement at follow-up.
Full table The incidence of microalbuminuria was 12.6% (143/1134, 95% CI, 10.7 to 14.7%), a rate of 1.8 per 100 person-years. This did not differ by sex, being 13.2% (73 out of 554) in men and 12.1% (70 out of 580) in women (P = 0.5), equivalent to rates of 1.9 and 1.7 per 100 person-years, respectively. The risk of progression to macroalbuminuria (an AER> 200 g/min) was 1.7% (19 out of 1134, 95% CI, 1.0 to 2.6%). Risk factors in these 19 patients were similar to those for progressors to microalbuminuria, however, with such small numbers it would be wrong to attach too much significance to these findings. On the other hand, it may be misleading to simply combine these patients with the group progressing to microalbuminuria. These 19 patients therefore were excluded from subsequent analyses.
Risk factors at baseline for progression to microalbuminuria did not include diabetes duration, but did include HbA1c and AER Table 2. A model fitting a breakpoint to the association between HbA1c and risk of microalbuminuria was not superior to a log linear model Figure 2. Other univariate associations with progression to microalbuminuria were noted for cholesterol, fasting triglyceride, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, body mass index (BMI), and waist to hip ratio (WHR). In a sex-specific analysis for the latter, WHR was greater at baseline in those who progressed to microalbuminuria compared with those who did not for both men (0.89 vs. 0.87, P = 0.07) and women (0.82 vs. 0.79, P = 0.06). The presence of peripheral neuropathy and retinopathy was also strongly predictive of progression to microalbuminuria.
Figure 2.
Relationship between glycemic control at baseline and risk of microalbuminuria at follow-up; testing for a threshold effect (numbers of patients progressing to microalbuminuria/total at risk). Symbols are: (
) Odds ratio estimated with logistic regression, adjusted for baseline duration; (
) line of best fit, using weighted linear regression; (
) segment 1 of the threshold model (at 8.5% London HbA1c); (
) segment 2 of the threshold model.
Even in the univariate analysis, baseline blood pressure was not associated with progression, whether it was calculated as a mean for the whole population (systolic 118 mm Hg and diastolic 73 mm Hg in both groups), the mean for all those not on antihypertensive therapy (as shown in Table 2), or as a median, with all treated hypertensives assigned to the upper decile of the blood pressure distribution (systolic 114 mm Hg in progressors, 117 in nonprogressors, P = 0.3, diastolic 73 mm Hg in both progressors and nonprogressors, P = 0.5)19. At baseline, 5% (N = 58) of these patients were on antihypertensive therapy. This proportion did not differ between those who went on to progress to microalbuminuria compared with those who remained normoalbuminuric (5 vs. 6%, respectively, P = 0.9). At follow-up, 12.6% of persistently normoalbuminuric and 18% of microalbuminuric patients were on antihypertensive therapy. Of the 122 normoalbuminuric patients on antihypertensive therapy, 88 (72%) were on angiotensin-converting enzyme inhibitors, and 29 (24%) on calcium channel blockers. The rest of the subjects were on other medications. Indications for treatment of these patients were a history of hypertension at baseline (N = 61) or follow-up (an additional 34). Of the remaining patients with no history of hypertension, two were prescribed therapy because of persistent microalbuminuria. Thus, we could find reasons for therapy in 80% of patients on antihypertensive medication.
Markers of endothelial function such as vWF and presence of cardiovascular disease at baseline were not associated with progression of renal disease.
Once duration, HbA1c, and baseline AER were accounted for, the risk factors that remained significantly related to risk of progression to microalbuminuria were fasting triglyceride, HDL cholesterol, LDL cholesterol, BMI, and WHR Table 3. The presence of any retinopathy at baseline increased the risk of progression to microalbuminuria 1.8-fold (95% CI, 1.1 to 2.8, P = 0.02). Standardized regression estimates were calculated, entering those significant risk factors that appeared to be independent of baseline AER and HbA1c simultaneously into a model, but not including the presence of other complications Table 4. The factors that remained statistically significant in this multivariate model were HbA1c, AER, fasting triglyceride, and WHR. When LDL was added to this model, the SRE was 1.24 (95% CI, 0.95 to 1.62, P = 0.1), and the effect for triglyceride was attenuated to 1.26 (95% CI, 1.01 to 1.56, P = 0.08).
Table 3 - Baseline risk factors for progression to microalbuminuria after adjusting for duration, HbA1c and AER.
Full table Table 4 - Standardized estimates of relative risk (SERR) for progression to microalbuminuria.
Full table Univariate standardized regression effects for triglycerides were 1.52 (95% CI, 1.23 to 1.87, P < 0.001) and for WHR 1.32 (95% CI, 1.08 to 1.61, P = 0.008). Both of these associations were attenuated but not abolished in the multivariate model. Use of lipid-lowering therapy was confined to 2.7% of nonprogressors (26 out of 972) and 4.9% (7 out of 143) of progressors. Their exclusion did not alter the results shown here.
When risk factors at follow-up between those who became microalbuminuric and those who remained normoalbuminuric were compared, a different risk profile associated with microalbuminuria emerged. Thus, both systolic and diastolic blood pressure values at follow-up were higher in the incident cases than in those who remained normoalbuminuric, even when baseline AER was taken into account (systolic 123 vs. 118 mm Hg, P = 0.0005, diastolic 75 vs. 73 mm Hg, P = 0.02; Table 5). These results did not change qualitatively when those on antihypertensive therapy at either baseline or follow-up were removed or when the median blood pressure was used. Furthermore, follow-up HbA1c, the previous two years worth of HbA1c, and fasting triglyceride were higher at follow-up in progressors compared with nonprogressors. Change in risk factor status over the follow-up period was also calculated. Thus, while systolic blood pressure increased by 5.8 
17.6 (SD) mm Hg in progressors, there was little change in nonprogressors (0.5 16.9, P = 0.0004). Similarly, using the HbA1c conversion to the DCCT values at both baseline and follow-up, the change in HbA1c was 2.1 13.9% in progressors and –0.7 12.6% in nonprogressors (P = 0.02).
Table 5 - Risk factors at follow-up by progression to microalbuminuria, adjusting for duration, HbA1c, and AER at baseline.
Full table DISCUSSION
Over seven years, the incidence of microalbuminuria in the EURODIAB cohort of type 1 patients was 12.6% (1.8 per 100 person-years). To our knowledge this is the largest study examining microalbuminuria incidence. Comparisons between studies are difficult, as potential determinants of microalbuminuria, such as age, duration, blood pressure, and glycated hemoglobin inclusion criteria may differ, but our estimate falls within the range of 1 to 3% per annum anticipated by previous studies20,21,22, and is close to the incidence of 14.5% over seven years found by the Microalbuminuria Collaborative Study Group23. The incidence in our study is probably underestimated, as those who were lost to follow-up had worse glycemic control at baseline. However, even taking this into account, our estimates are much lower than those derived from the U.S. Epidemiology of Diabetes Complications Study (EDC), which had an incidence of microalbuminuria over two years of 9.4%24. The most likely explanation for this discrepancy is the poorer glycemic control in the U.S. cohort compared to EURODIAB at baseline9.
One of the main factors accounting for risk of progression to microalbuminuria was glycemic control. Others have vigorously argued for a glycemic threshold for risk of microalbuminuria6. This finding is not supported by larger cross-sectional studies, and the DCCT provides no support for this contention7,8. Here, in the largest observational follow-up study, we also demonstrate that there is no glycemic threshold for risk of microalbuminuria, and efforts to reduce HbA1c should therefore be continued at all levels. Clearly, there may be limitations to our threshold analysis. A post hoc power calculation indicates that if the difference in the odds ratio for progression to microalbuminuria from the first to the median value of HbA1c (median HbA1c = 6%) is 1 and the difference in the odds ratio from the median to the last HbA1c value is 6, this latter difference can be shown with the numbers we have with P < 0.016 and a power of 76% (these power calculations have been corrected to take into consideration multiple significance testing that has to be performed in the process of the segmented regression analysis)17,25. Thus, should a threshold exist, we had sufficient power to detect it. While a single baseline measurement of HBA1c may not represent the total glycemic control that an individual is exposed to, as control "tracks" within individuals, it is likely that the relative ranking of each individual would not change had we taken several baseline measurements. Furthermore, we cannot be sure that glycemic control was captured at a time when the initiation of microalbuminuria would take place, but again, our argument regarding tracking still applies. These limitations would apply to the previous analyses of a threshold effect.
Apart from well-known risk factors such as HbA1c and AER at baseline5,24,26,27, independent associations were also observed with fasting triglyceride and WHR. Data from previous cross-sectional studies show that lipids are abnormal in patients with microalbuminuria28,29,30,31 and, more importantly in terms of assessing causality, that apolipoprotein B levels32 and LDL cholesterol24 are elevated in those at risk of subsequent microalbuminuria. Interestingly, while many of the lipid parameters measured here are highly correlated, it is striking that only triglyceride remains significant, despite being less precisely measured than total and HDL cholesterol. To our knowledge, no previous study has explored the predictive role of central obesity in microalbuminuria or examined whether these factors are independent of disturbances in blood glucose. In EURODIAB, both triglyceride and WHR were independent of glycated hemoglobin and AER at baseline. Furthermore, when the strength of key risk factors was compared by using standardized estimates of risk, we observed that both fasting triglyceride and WHR were nearly as strong as baseline AER in predicting microalbuminuria.
Both of these factors are features of the insulin-resistance syndrome33, and there is evidence that type 1 diabetes is associated with insulin resistance34. This is predictive of vascular disease35, is more pronounced in those with microalbuminuria36, and is also evident in family members of patients with type 1 diabetes37. Insulin resistance and at least overall obesity is associated with microalbuminuria in some38 but not all studies in the general population39, and in type 1 diabetes, insulin resistance predates the onset of microalbuminuria40. Although there were no direct measures of insulin resistance in our study, triglyceride and WHR are key features of the syndrome, while factors such as blood pressure are less consistently associated. Our demonstration of an impact of these factors on AER, independent of glycemic control, emphasizes their importance as risk factors and further indicates that the mechanism of action is not entirely via glycemic control.
The mechanisms by which insulin resistance and microalbuminuria may be linked in type 1 diabetes are not clear. There are two categories of explanation: either insulin resistance has a direct pathological effect on the kidney to cause microalbuminuria41,42,43,44, or the two factors cluster together and are in turn caused by another factor. An interesting candidate for this latter role is endothelial dysfunction, which is related to reduced insulin action on the one hand and enhanced capillary leakage of albumin on the other45. To support the latter, there is evidence that endothelium-dependent vasodilation is impaired in people with type 1 diabetes, and again more marked in those with microalbuminuria46. This has also been shown in insulin resistance47. The presence of other complications at baseline, such as peripheral neuropathy and retinopathy, also predicted the incidence of microalbuminuria at follow-up. Again, this supports the notion of a generalized state of endothelial dysfunction, which enhances the risk of all complications48.
However, there was no difference at baseline in vWF, which is commonly used as a marker of endothelial dysfunction, in those who progressed to microalbuminuria compared with those who remained normoalbuminuric. This is in contrast to some49 but not all50 previous findings. It is difficult to account for this discrepancy, particularly as we have shown a relationship between vWF and microalbuminuria cross-sectionally51. It may be that changes in endothelial dysfunction before development of microalbuminuria are too subtle to be detected by vWF levels at this early stage.
Interestingly, blood pressure and smoking did not appear to be risk factors, even in the univariate models. However, when follow-up data were compared, resting blood pressure was higher in incident cases. As AER at baseline was significantly greater in progressors versus nonprogressors, a tentative interpretation is that the increase in AER either precedes or is concomitant with that of blood pressure. This has been a controversial topic, with some arguing in favor of this temporal sequence20,24,27 and some against it23,52. All centers were trained and tested to the same protocol for the study so that our measurement of resting blood pressure was well standardized. However, resting blood pressure cannot capture the whole picture of blood pressure occurring over a 24-hour period, and considerable differences in 24-hour profiles have been observed, even in normoalbuminuric patients53. Interestingly, several normoalbuminuric patients were on antihypertensive therapy at follow-up (12.6%). Approximately 80% of this therapy could be accounted for by a diagnosis of hypertension at either baseline or follow-up.
There are limitations to this study. Follow-up data were obtained in 62% of potentially available patients. Given the wide European base for this study, this is a highly respectable follow-up rate for such a cohort study. Patients who are lost to follow-up tend to have a worse risk factor profile, and we showed this to be the case for glycated hemoglobin and AER. Therefore, the risk of microalbuminuria may be underestimated, but it is unlikely that our conclusions on risk factor relationships, which are the main findings of this analysis, would differ, as it would be hard to hypothesize a situation where, for example, WHR was positively related to risk of progression in responders and negatively related in nonresponders. The strengths of this analysis are that all measurements were performed using a standard protocol, which investigators were well acquainted with from the baseline investigations until the end of the study. Furthermore, while the number of observers for some of the measures, in particular blood pressure and WHR, may have introduced precision variability, this would be nondifferential. In other words, it is unlikely that investigators consistently underestimated WHR in those who were persistently normoalbuminuric and overestimated in those with incident microalbuminuria. Given this variability, it is even more surprising that such a strong association is noted between WHR and microalbuminuria, and with a more precise measure this association may be even stronger. A similar argument applies to the urine collections. Investigators were trained to a standard protocol, and any inaccuracies in collection would be nondifferential, that is, not systematically biased by risk factor status or outcome. The same laboratory was used for measurement of urinary albumin at both baseline and follow-up; the fact that we had one urine collection at baseline and two at follow-up should not seriously affect the assessment of risk factor relationships.
We conclude that the risk of microalbuminuria remains high in type 1 diabetes and that glycemic control is one of the strongest predictors of this risk, with no evidence of a threshold effect, supporting efforts to keep HbA1c levels as low as possible. Microalbuminuria is strongly determined by markers of the insulin resistance syndrome, particularly triglyceride and WHR. The close relationship between lipids and microalbuminuria has important implications for treatment. A study in type 2 diabetes indicated that statin therapy reduced albuminuria by 25%54, although this could not be replicated in type 1 patients, possibly because of an inadequate sample size55. Other indications that both insulin resistance and microalbuminuria share a common antecedent are tantalizing and require further investigation.
References
| 1. | Viberti GC, Hill RD & Jarrett RJ et al. Microalbuminuria as a predictor of clinical nephropathy in insulin-dependent diabetes mellitus. Lancet 1982; 1: 1430−1432. | PubMed | ISI | ChemPort | |
| 2. | Messent JWC, Elliott TG & Hill RD et al. Prognostic significance of microalbuminuria in insulin-dependent diabetes mellitus: A twenty-three year follow-up study. Kidney Int 1992; 41: 836−839. | PubMed | ISI | ChemPort | |
| 3. | Jensen T, Borch-Johnsen K, Kofoed-Enevoldsen A & Deckert T. Coronary heart disease in young type 1 (insulin-dependent) diabetic patients with and without diabetic nephropathy: Incidence and risk factors. Diabetologia 1987; 30: 144−148. | Article | PubMed | ISI | ChemPort | |
| 4. | Deckert T, Yokoyama H & Mathiesen E et al. Cohort study of predictive value of urinary albumin excretion for atherosclerotic vascular disease in patients with insulin dependent diabetes. Br Med J 1996; 312: 871−874. | ISI | ChemPort | |
| 5. | The 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 1993; 329: 977−986. | PubMed | ISI | |
| 6. | Krolewski AS, Laffel LMB & Krolewski M et al. Glycosylated hemoglobin and the risk of microalbuminuria in patients with insulin-dependent diabetes mellitus. N Engl J Med 1995; 332: 1251−1255. | Article | PubMed | ISI | ChemPort | |
| 7. | The Diabetes Control and Complications Trial Research Group. The absence of a glycaemic threshold for the development of long-term complications: The perspective of the diabetes control and complications trial. Diabetes 1996; 45: 1289−1298. | ISI | |
| 8. | Chaturvedi N & Fuller JH. Glycosylated haemoglobin and the risk of microalbuminuria in insulin dependent diabetes mellitus. N Engl J Med 1995; 333: 940−940. | Article | PubMed | ISI | ChemPort | |
| 9. | The EURODIAB IDDM Complications Study Group. Microvascular and acute complications in insulin dependent diabetes mellitus: The EURODIAB IDDM Complications Study. Diabetologia 1994; 37: 278−285. | PubMed | ISI | |
| 10. | Kearney EM, Mount JN & Watts GF et al. Simple immunoturbidometric method for determining urinary albumin at low concentrations using centrifugal analyser. J Clin Pathol 1987; 40: 465−468. | PubMed | ISI | ChemPort | |
| 11. | Siedel J, Hagele EO, Ziegenhorn J & Wahlefeld AW. Reagent for the enzymatic determination of serum total cholesterol with improved lipolytic efficiency. Clin Chem 1983; 29: 1075−1080. | PubMed | ISI | ChemPort | |
| 12. | Bucolo G & David H. Quantitative determination of serum triglycerides by the use of enzymes. Clin Chem 1973; 19: 476−482. | PubMed | ISI | ChemPort | |
| 13. | Warnick GR & Albers JJ. A comprehensive evaluation of the heparin-manganese precipitation procedure for estimating high density lipoprotein cholesterol. J Lipid Res 1978; 19: 65−76. | PubMed | ISI | ChemPort | |
| 14. | Friedewald WT, Levy RI & Fredrickson DS. Estimation of the concentration of low density lipoprotein in cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972; 18: 499−502. | PubMed | ISI | ChemPort | |
| 15. | John GW, Gray MR, Bates DL & Beacham JL. Enzyme immunoassay: A new technique for estimating HbA1c. Clin Chem 1993; 39: 663−666. | PubMed | ISI | |
| 16. | Ford I, Malik RA & Newrick PG et al. Relationships between haemostatic factors and capillary morphology in human diabetic neuropathy. Thromb Haemost 1992; 68: 628−633. | PubMed | ISI | ChemPort | |
| 17. | Jones RH & Molitoris BA. A statistical method for determining the breakpoint of two lines. Ann Biochem 1984; 141: 287−290. | Article | ChemPort | |
| 18. | Ulm K. A statistical method for assessing a threshold in epidemiological studies. Stat Med 1991; 10: 341−349. | PubMed | ISI | ChemPort | |
| 19. | White IR, Chaturvedi N & McKeigue PM. Median analysis of blood pressure for a sample including treated hypertensives. Stat Med 1994; 13: 1635−1641. | PubMed | ISI | ChemPort | |
| 20. | Mathiesen ER, Ronn B & Storm B et al. The natural course of microalbuminuria in insulin-dependent diabetes: A 10 year prospective study. Diabet Med 1995; 12: 482−487. | PubMed | ISI | ChemPort | |
| 21. | Andersen AR, Christiansen JS & Andersen JK et al. Diabetic nephropathy in type 1 (insulin-dependent) diabetes: An epidemiological study. Diabetologia 1983; 25: 496−501. | PubMed | ISI | ChemPort | |
| 22. | Kofoed-Enevoldsen A, Borch-Johnsen K & Kreiner S et al. Declining incidence of persistent proteinuria in type 1 (insulin-dependent) diabetic patients in Denmark. Diabetes 1987; 36: 205−209. | PubMed | ChemPort | |
| 23. | The Microalbuminuria Collaborative Study Group. Predictors of the development of microalbuminuria in patients with type 1 diabetes mellitus: A seven-year prospective study. Diabet Med 1999; 16: 918−925. | ISI | |
| 24. | Coonrod BA, Ellis D & Becker DJ et al. Predictors of microalbuminuria in individuals with IDDM. Diabetes Care 1993; 16: 1376−1383. | PubMed | ISI | ChemPort | |
| 25. | Hilsenbeck RH & Clark G. Practical p-value adjustment for optimally selected cutpoints. Stat Med 1996; 15: 103−112 10.1002/(sici)1097-0258(19960115)15:1<103::aid-sim156>3.0.co;2-y. | Article | PubMed | ISI | ChemPort | |
| 26. | Yip J. Microalbuminuria collaborative study group: United Kingdom: Risk factors of microalbuminuria in insulin dependent diabetic patients: A cohort study. Br Med J 1993; 306: 1235−1239. | ISI | |
| 27. | Gilbert RE, Tsalamandris C & Bach LA et al. Long-term glycemic control and the rate of progression of early diabetic kidney disease. Kidney Int 1993; 44: 855−859. | PubMed | ISI | ChemPort | |
| 28. | Jones SL, Close CF & Mattock MB et al. Plasma lipid and coagulation factor concentrations in insulin dependent diabetics with microalbuminuria. Br Med J 1989; 298: 487−489. | ISI | ChemPort | |
| 29. | Vannini P, Ciavarella A & Flammini M et al. Lipid abnormalities in insulin-dependent diabetic patients with albuminuria. Diabetes Care 1984; 7: 151−154. | PubMed | ISI | ChemPort | |
| 30. | Jensen T, Stender S & Deckert T. Abnormalities in plasma concentrations of lipoproteins and fibrinogen in type 1 (insulin-dependent) diabetic patients with increased urinary albumin excretion. Diabetologia 1988; 31: 142−145. | Article | PubMed | ISI | ChemPort | |
| 31. | Sibley SD, Hokanson JE & Steffes MW et al. Increased small dense LDL and intermediate-density lipoprotein with albuminuria in type 1 diabetes. Diabetes Care 1999; 22: 1165−1170. | PubMed | ISI | ChemPort | |
| 32. | Watts GF, Powrie JK, O'Brien SF & Shaw KM. Apolipoprotein B independently predicts progression of very-low-level albuminuria in insulin-dependent diabetes mellitus. Metabolism 1996; 45: 1101−1107. | Article | PubMed | ISI | ChemPort | |
| 33. | Reaven GM. Role of insulin resistance in human disease. Diabetes 1988; 37: 1595−1607. | PubMed | ISI | ChemPort | |
| 34. | DeFronzo RA, Hendler R & Simonson D. Insulin resistance is a prominent feature of insulin-dependent diabetes. Diabetes 1982; 31: 795−801. | PubMed | ISI | ChemPort | |
| 35. | Martin FI & Hopper JL. The relationship of acute insulin sensitivity to the progression of vascular disease in long-term type 1 (insulin-dependent) diabetes mellitus. Diabetologia 1987; 30: 149−153. | Article | PubMed | ISI | ChemPort | |
| 36. | Yip J, Mattock MB & Morocutti A et al. Insulin resistance in insulin-dependent diabetic patients with microalbuminuria. Lancet 1993; 342: 883−887. | Article | PubMed | ISI | ChemPort | |
| 37. | Yip J, Mattock M & Sethi M et al. Insulin resistance in family members of insulin-dependent diabetic patients with microalbuminuria. Lancet 1993; 341: 369−370. | Article | PubMed | ISI | ChemPort | |
| 38. | Mykkanen L, Zaccaro DJ & Wagenknecht LE et al. Microalbuminuria is associated with insulin resistance in nondiabetic subjects: The insulin resistance atherosclerosis study. Diabetes 1998; 47: 793−800. | PubMed | ISI | ChemPort | |
| 39. | Zavaroni I, Bonini L & Gasparini P et al. Dissociation between urinary albumin excretion and variables associated with insulin resistance in a healthy population. J Intern Med 1996; 240: 151−156. | Article | PubMed | ISI | ChemPort | |
| 40. | Ekstrand AV, Groop PH & Gronhagen-Riska C. Insulin resistance precedes microalbuminuria in patients with insulin-dependent diabetes mellitus. Nephrol Dial Transplant 1998; 13: 3079−3083 10.1093/ndt/13.12.3079. | PubMed | ISI | ChemPort | |
| 41. | Nestler JE, Barlascini CO & Tetrault GA et al. Increased transcapillary escape rate of albumin in non-diabetic men in response to hyperinsulinaemia. Diabetes 1990; 39: 1212−1217. | PubMed | ISI | ChemPort | |
| 42. | Feldt-Rasmussen B. Increased transcapillary escape rate of albumin in type 1 (insulin-dependent) diabetic patients with microalbuminuria. Diabetologia 1986; 29: 282−286. | Article | PubMed | ChemPort | |
| 43. | Cohen AJ, McCarthy DM & Stoff JS. Direct haemodynamic effect of insulin in the isolated perfused kidney. Am J Physiol 1989; 257: F580−F585. | PubMed | ISI | ChemPort | |
| 44. | Tucker BJ, Anderson CM & Thies RS et al. Glomerular hemodynamic alterations during acute hyperinsulinemia in normal and diabetic rats. Kidney Int 1992; 42: 1160−1168. | PubMed | ISI | ChemPort | |
| 45. | Yudkin JS. Hyperinsulinaemia, insulin resistance, microalbuminuria and the risk of coronary heart disease. Ann Med 1996; 28: 433−438. | PubMed | ISI | ChemPort | |
| 46. | Zenere BM, Arcaro G & Saggiani F et al. Noninvasive detection of functional alterations of the arterial wall in IDDM patients with and without microalbuminuria. Diabetes Care 1995; 18: 975−982. | PubMed | ISI | ChemPort | |
| 47. | Steinberg HO, Chaker H & Leaming R et al. Obesity/insulin resistance is associated with endothelial dysfunction: Implications for the syndrome of insulin resistance. J Clin Invest 1996; 97: 2601−2610. | PubMed | ISI | ChemPort | |
| 48. | Stehouwer CD, Fischer HR & van Kuijk AW et al. Endothelial dysfunction precedes development of microalbuminuria in IDDM. Diabetes 1995; 44: 561−564. | PubMed | ISI | ChemPort | |
| 49. | Stehouwer CD, Stroes ES & Hackeng WH et al. Von Willebrand factor and development of diabetic nephropathy in IDDM. Diabetes 1991; 40: 971−976. | PubMed | ISI | ChemPort | |
| 50. | Myrup B, Mathiesen ER, Ronn B & Deckert T. Endothelial function and serum lipids in the course of developing microalbuminuria in insulin-dependent diabetes mellitus. Diabetes Res 1994; 2: 33−39. |
| 51. | Greaves M, Malia RG & Goodfellow K et al. Fibrinogen and von Willebrand factor in IDDM. Relationships to lipid vascular risk factors, blood pressure, glycaemic control and urinary albumin excretion rate: The EURODIAB IDDM complications study. Diabetologia 1997; 40: 698−705 10.1007/s001250050736. | Article | PubMed | ISI | ChemPort | |
| 52. | Microalbuminuria Collaborative Study Group UK. Risk factors for development of microalbuminuria in insulin dependent diabetic patients: A cohort study. Br Med J 1993; 306: 1235−1239. | ISI | |
| 53. | Nielsen S, Schmitz A & Poulsen PL et al. Albuminuria and 24-h ambulatory blood pressure in normoalbuminuric and microalbuminuric NIDDM patients: A longitudinal study. Diabetes Care 1995; 18: 1434−1441. | PubMed | ISI | ChemPort | |
| 54. | Tonolo G, Ciccarese M & Brizzi P et al. Reduction of albumin excretion rate in normotensive microalbuminuric type 2 diabetic patients during long-term simvastatin treatment. Diabetes Care 1997; 20: 1891−1895. | PubMed | ISI | ChemPort | |
| 55. | Zhang A, Vertommen J & Van-Gaal L et al. Effects of pravastatin on lipid levels, in vitro oxidizability of non-HDL lipoproteins and microalbuminuria in IDDM patients. Diabetes Res Clin Pract 1995; 29: 189−194 10.1016/0168-8227(95)01138-2. | Article | PubMed | ISI | ChemPort | |
Appendices
APPENDIX
Centers and staff involved in the study include the following: B. Karamanos, A. Kofinis, and K. Petrou (Hippokration Hospital, Athens, Greece); R. Giorgino, F. Giorgino, G. Picca, A. Angarano, and G. De Pergola (Istituto di Clinica Medica, endocrinologia e Malattie Metaboliche, Univesita di Bari, Bari, Italy); C. Ionescu-Tirgoviste and A. Coszma (Clinic of Diabetes, Nutrition and Metabolic Diseases, Bucharest, Romania); M. Songini, A. Casu, M. Pedron, and M. Fossarello (Department of Internal Medicine, Ospedale San Michele, Cagliari, Italy); J.B. Ferriss, G. Grealy, D.O. Keefe, A. White, and P.E. Cleary (Cork University Hospital, Cork, Ireland); M. Toeller and C. Arden (Diabetes Research Institute, Heinrich-Heine University, Dusseldorf, Germany); R. Rottiers, C. Tuyttens, and H. Priem (University Hospital of Gent, Gent, Belgium); P. Ebeling, M. Kylliäinen, and T. Kyostio-Renvall (University Hospital of Helsinki, Helsinki, Finland); B. Idzior-Walus, J. Sieradzki, and K. Cyganek (Department of Metabolic Diseases, Jagiellonian University, Krakow, Poland); H.H.P.J. Lemkes and C. Roest (University Hospital of Leiden, Leiden, The Netherlands); J. Nunes-Correa, M.C. Rogado, L. Gardete-Correia, and M.C. Cardoso (Portuguese Diabetic Association, Lisbon, Portugal); G. Michel, R. Wirion, and S. Cardillo (Center Hospitalier, Luxembourg); G. Pozza, R. Mangili, V. Asnaghi, Rosangela Lattanzio, and G. Galardi (Ospedale San Raffaele, Milan, Italy); E. Standl, B. Schaffler, H. Brand, and A. Harms (City Hospital Schwabing, Munich, Germany); D. Ben Soussan and O. Verier-Mine (Center Hospitalier de Valenciennes, France); J.H. Fuller, J. Holloway, L. Asbury, and D.J. Betteridge (University College London, UK); G. Cathelineau, A. Bouallouche, and B. Villatte Cathelineau (Hôpital Saint-Louis, Paris, France); F. Santeusanio, G. Rosi, V. D'Alessandro, and C. Cagini (Instituto di Patologia Medica, Policlinico, Perugia, Italy); R. Navalesi, G. Penno, S. Bandinelli, and R. Miccoli (Dipartimento di Endocrinologia e Metabolismo, Pisa, Italy); G. Ghirlanda, C. Saponara, P. Cotroneo, A. Manto, and A. Minnella (Universita Cattolica del Sacro Cuore, Rome, Italy); J.D. Ward, S. Tesfaye, S. Eaton, and C. Mody (Royal Hallamshire Hospital, Sheffield, UK); M. Porta, P. Cavallo Perin, M. Borra, and S. Giunti (Clinica Medica B, Patologia Medica, Ospedale Molinette, and Ospedale "Agnelli," Turin, Italy); N. Papazoglou and G. Manes (General Hospital of Thessaloniki, Greece); M. Muggeo and M. Iagulli (Cattedra di Malatties del Metabolismo, Verona, Italy); K. Irsigler and H. Abrahamian (Hospital Vienna Lainz, Vienna, Austria); S. Walford, E.V. Wardle, J. Sinclair, and S. Hughes (New Cross Hospital, Wolverhampton, UK); G. Roglic, Z. Metelko, and Z. Resman (Vuk Vrhovac Institute for Diabetes, Zagreb, Croatia).
The Steering Committee Members: J.H. Fuller (London), B. Karamanos, Chairman (Athens), A.-K. Sjolie (Aarhus), N. Chaturvedi (London), M. Toeller (Dusseldorf), G. Pozza, co-chairman (Milan), B. Ferriss (Cork), M. Porta (Turin), R. Rottiers (Gent), and G. Michel (Luxembourg).
Co-ordinating Center: J.H. Fuller, N. Chaturvedi, J. Holloway, D. Webb, and L. Asbury (University College London).
The Central Laboratories: G.-C. Viberti, R. Swaminathan, P. Lumb, A. Collins, and S. Sankaralingham (Guy's and St. Thomas' Hospital, London, UK).
The Retinopathy Grading Center: S. Aldington, T. Mortemore, and H. Lipinski (Imperial College, London, UK).
Acknowledgments
This study was supported by project grants from the Wellcome Trust and the European Union. Some of these data were presented orally at the 1999 meeting of the European Association for the Study of Diabetes. We thank Ms. Lynda Stevens, statistician at University College (London) for assistance with the breakpoint analysis, Ms. Sue Manley from the Radcliffe Infirmary (Oxford) and Ms. Suky Sankaralingham for the standardization of follow-up HbA1c performed by the London method to the DCCT method, and Mr. Gary John from St. Bartholomew's Hospital (London) for baseline standardization. We would also like to thank all staff and patients who took part in the study.
