IJE Advance Access originally published online on June 16, 2006
International Journal of Epidemiology 2006 35(5):1331-1335; doi:10.1093/ije/dyl115
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Article |
Albuminuria as a marker of the risk of developing type 2 diabetes in non-diabetic Aboriginal Australians
Centre for Chronic Disease, Department of Medicine—Central Clinical Division, The University of Queensland, Brisbane, Queensland, Australia
* Corresponding author. Centre for Chronic Disease, Department of Medicine—Central Clinical Division, The University of Queensland, Level H, Clinical Sciences Building, Royal Brisbane Hospital, Herston Qld 4029, Australia. E-mail: zwang{at}ccs.uq.edu.au
 |
Abstract |
|---|
 Top Abstract Subjects, materials, and methodsResultsDiscussionReferences |
|---|
Background Aboriginal Australians experience a higher risk of diabetes than the general Australian population. In this paper, we conducted a nested case–control study to determine whether the presence of microalbuminuria and macroalbuminuria is associated with the development of diabetes among diabetes-free Aboriginal people at baseline.
Methods Urine albumin to creatinine ratios (ACRs) were obtained from 882 Aboriginal people aged 20–74 years from one community. Among them 750 were free of either clinical known diabetes or newly diagnosed diabetes according to WHO 1999 criteria. Over an 11 year follow-up period, 117 participants developed diabetes. They were defined as cases. Each case was matched by an individual control with same sex and body mass index (BMI) category, and age within 2 years. Conditional logistic regression was used to assess the association between albuminuria and diabetes.
Results The baseline level of ACR was significantly higher among cases than among controls. The odds ratios for future diabetes were 2.36 [95% confidence interval (95% CI) 1.01–5.50] and 3.27 (95% CI 1.38–7.77) for middle and upper tertiles, respectively, with adjustment for age, BMI, serum total cholesterol, serum C-reactive protein values, and fasting plasma glucose at the baseline. The adjusted odds ratios were 1.90 (95% CI 0.88–4.06) and 2.51 (95% CI 1.08–5.87) for those with microalbuminuria and macroalbuminuria, respectively.
Conclusions The presence of microalbuminuria and macroalbuminuria predicts diabetes independent of other known risk markers of development of type 2 diabetes in Aboriginal people.
Keywords diabetes in minorities, albuminuria, case–control, aboriginal health, epidemiology
Accepted 4 May 2006
Aboriginal Australians have a higher prevalence of diabetes than the general Australian population.1 Physical inactivity and obesity are independent predictors of type 2 diabetes.2 However, current physical inactivity and obesity may not fully explain the high risk of type 2 diabetes in Aboriginal people. There has been considerable focus on the concept of microalbuminuria, not only because it predicts renal disease in diabetes but also because it relates to premature mortality in the diabetic and in the general population. Albuminuria in Aboriginal people is a major risk factor for renal disease3 and cardiovascular disease.4 Albuminuria can also be present at the time of diagnosis of diabetes. An increased prevalence of microalbuminuria has been observed among those with impaired fasting glucose before developing clinically diagnosed diabetes.5 Mykkanen et al.6 reported microalbuminuria predicted the development of diabetes independently of blood pressure level. Brantsman et al.7 hypothesized that the presence of increased urine albumin excretion in the non-diabetic population is associated with an increased risk of development of type 2 diabetes. Their 4 year follow-up data suggest that urinary albumin excretion is an independent predictor of diabetes.7 However, their data were mainly from a Caucasian population in The Netherlands. Their findings need to be confirmed in other populations. Confounding is a major concern in most observational studies. People with diabetes have significantly higher levels of most risk factors such as body mass index (BMI), waist circumference, glucose, and lipids than those without diabetes. In this study, we examine the association between albuminuria and type 2 diabetes in an Aboriginal population using an individually matched nested case–control study design.
|
Subjects, materials, and methods |
|---|
|
TopAbstractSubjects, materials, and methodsResultsDiscussionReferences |
|---|
In 1992 a community-wide renal disease screening programme was started in a remote Northern Territory Aboriginal community in Australia. Participants were offered a baseline examination and testing between 1992 and 1995. Eight hundred and eight two (882) adults aged 20–74 years were followed until April 30, 2005.
During the baseline visit, a fasting venous blood sample was drawn for plasma glucose, and serum total cholesterol and triglycerides measurements. For those free from diabetes at baseline, plasma glucose levels were also measured 2 h after a 75 g oral glucose challenge. An individual with a 2 h plasma glucose level of 7.8–11.0 mmol/l was considered as having impaired glucose tolerance (IGT).
Urinary albumin concentration was measured by the Beckman immunoassay. A urine albumin to creatinine ratio (ACR) was calculated (mg/mmol). Previous studies in Aboriginal people used 3.4 and 34 mg/mmol as cut-off for microalbuminuria and macro (or overt) albuminuria.8–10 However, to be consistent with recent studies, this study defined normoalbuminuria as a urine ACR <2.5 mg/mmol, microalbuminuria as a urine ACR 2.5–25 mg/mmol, and macroalbuminuria as a urine ACR >25 mg/mmol.11–13 In addition, baseline urine ACR values were categorized into tertiles based on the sex-specific distribution of the total study population.
High sensitivity C-reactive protein concentrations were measured using the immunoturbidimetric C-reactive protein assay on a Hitachi 917 analyser (Roche Diagnostics Australia) with a detection limit of 0.03 mg/l. The assay's analytical range was from 0.1 to 20 mg/l. Samples with values >20 mg/l were measured using dilution techniques. The imprecision of the assay is <5%. Other risk factors were measured, such as blood pressures, serum cholesterol, triglycerides, BMI, cigarette smoking, alcohol drinking, and the presence of diabetes, and have been described elsewhere.14–16
A total of 132 participants who either had known diabetes before the baseline examination or were newly diagnosed as having diabetes at the baseline examination according to the World Health Organisation (WHO) 1999 criteria17 on fasting and 2 h post-load glucose values were excluded. During a median 11 year follow-up period, newly diagnosed cases of diabetes were determined through clinical and hospital records among 750 participants free from diabetes at baseline. A total of 117 participants developed clinically diagnosed diabetes according to their symptoms, fasting and/or 2 h post-load glucose values. Each case was matched by a control of the same sex and BMI category of <25, 25–29, and >30 kg/m2 and the closest age (within 2 years). Participants who died or had cardiovascular disease during the follow-up period before the diagnosis of diabetes were also excluded.
The project was approved by the Joint Institutional Ethics Committee of the Menzies School of Health Research and Territory Health Services and The Behavioural and Social Science Ethical Review Committee of the University of Queensland.
Statistical analysis
We used a matched t-test to assess differences in means and the McNemar test to assess for differences in proportions. Since the urine ACR, plasma glucose, and C-reactive protein values were skewed, the logarithmic transformed values were used and their geometric means were presented. Conditional logistic regression was used to examine the association between diabetes and urine ACR values. Crude and adjusted odds ratios and their 95% confidence intervals (95% CIs) were calculated with the lowest urine ACR group as reference. Since each case was matched by a control with the same sex, sex cannot be a possible confounder. However, the potential residual confounding effects of age and BMI were further assessed in the multiple conditional logistic regressions along with other non-matched variables such as serum cholesterol, C-reactive protein, fasting plasma glucose, and impaired plasma glucose status. Interactions between ACR and other variables were tested. We also used a logarithmic transformation of ACR as a continuous variable to assess its association with diabetes and its interactions with other factors as conducted in one of previous studies.7 All analyses were performed using Stata 9.0.18
|
Results |
|---|
|
TopAbstractSubjects, materials, and methodsResultsDiscussionReferences |
|---|
Baseline characteristics of study participants who were subsequently diagnosed with diabetes (cases) and those remaining free from diabetes (controls) are shown in Table 1. Since sex, age, and BMI are matching variables, as expected for successful matching, cases and controls are similar in those variables. Cases had higher serum cholesterol, plasma glucose, and C-reactive protein levels, and a higher proportion of participants with IGT than controls. There were no statistically significant differences in blood pressures, triglycerides, smoking and drinking, and waist circumferences.
|
The geometric means of urine ACR was significantly higher among cases than among controls. Table 2 shows higher proportions of microalbuminuria, macroalbuminuria, and participants in higher tertile groups among cases than among controls.
|
|
In age, sex, and BMI matched analysis, the odds ratios of incident diabetes were 2.22 (95% CI 1.11–4.48) and 3.32 (95% CI 1.61–6.85) for middle and upper tertiles, respectively, with the lower tertile group as reference. Similarly, odds ratios for microalbuminuria and macroalbuminuria were 2.22 (95% CI 1.15–4.29) and 2.69 (95% CI 1.30–5.56), respectively, relative to the normal albuminuria group. Interactions between albuminuria and other variables were examined and no significant interaction terms were observed. Adjusting for age and BMI had little impact on the effect estimate, as both were matching variables. Even after adjusting for possible confounding factors collected in this study, the associations between diabetes and elevated levels of ACR remained statistically significant. The adjusted odds ratios were 2.36 (95% CI 1.01–5.50) and 3.27 (95% CI 1.38–7.77) for middle and upper tertiles, and 1.90 (95% CI 0.88–4.06) and 2.51 (95% CI 1.08–5.87) for microalbuminuria and macroalbuminuria groups, respectively. Crude and adjusted odds ratios for log (ACR) as a continuous variable were significantly higher than the null effect.
|
Discussion |
|---|
|
TopAbstractSubjects, materials, and methodsResultsDiscussionReferences |
|---|
In this matched nested case–control study, we found that an elevated urine ACR value is associated with the development of diabetes in Aboriginal people. This association is independent of age, sex, BMI, serum cholesterol, C-reactive protein, and fasting plasma glucose. Our findings suggest that albuminuria can precede and predict the development of diabetes in Aboriginal people.
Albuminuria has usually been considered a consequence of diabetes.19 It has been associated with increased cardiovascular risk in populations with and without diabetes.13,20–22 In Aboriginal people, albuminuria predicts the risk of coronary heart disease independent of traditional risk factors.4 Several studies have demonstrated that microalbuminuria occurs in individuals without diabetes.19,23
Only two studies have examined whether it predicts the development of diabetes among people without diabetes.6,7 Both studies were conducted in Caucasian populations. Diabetes cases were determined during 4.2 years of follow-up in Brantsma et al.'s study7 and over 3.5 years in the study by Mykkanen et al.6 This nested case–control study is the first one in a high-risk indigenous population and has the longest follow-up of 11 years.
The mechanism of the association is not clear. Confounding effects are a major concern for an observational epidemiological study. Albuminuria is associated with the insulin resistance and metabolic syndrome,23,24 and the presence of the metabolic syndrome and insulin resistance increases the risk of diabetes.25–27 In this study, we did not identify individuals with metabolic syndrome, but the major risk components of metabolic syndrome were adjusted for, either through the matched study design or multiple conditional logistic regressions. After such adjustment, the association between diabetes and albuminuria remained. Albuminuria and diabetes may also share other risk factors that are not currently known. Since the onset of type 2 diabetes is generally held to be at least several years before clinical diagnosis,28 the pathophysiology of diabetes related vascular complications probably begins before the diagnosis of diabetes. Microalbuminuria is a marker of endothelial dysfunction, which may occur with early renal disease29 and before the current diagnosis of diabetes.
Although the underlying mechanism of the association is still not clear, it is important to know that the presence of microalbuminuria and macroalbuminuria can precede diabetes and independently predict the risk of future diabetes. Improving kidney function may be potentially beneficial to preventing diabetes. Our results suggest that microalbuminuria and macroalbuminuria can at least be useful for identifying persons at increased risk of diabetes regardless of whether the association is causal.
Our findings have particular clinical and public health implications for Aboriginal Australians since a high prevalence of albuminuria has also been found in many different Aboriginal communities.8,10,30 We have already reported that, in this study population, >30% of adults have microalbuminuria and 20% have macroalbuminuria.4
Our study had several limitations. First, the diabetes events were identified through hospital records and clinical records. It is likely that some undetected cases were included in the control group. The effect is likely to be underestimated owing to this bias. Second, those with albuminuria may be more likely to visit their doctors for other health problems. Therefore, the presence of diabetes among them is more likely to be identified. If this bias is present, it will overestimate the true association between albuminuria and diabetes, although there is no evidence on this. Third, albuminuria in this study was defined using random urine ACR values rather than 24 h urine collection for urinary albumin excretion. However, it has been found that spot ACR is a good screening test for microalbuminuria.31 Fourth, because the study participants were Aboriginal adults from a remote community in the Northern Territory of Australia, the results may not be generalizable to other populations. The findings in the two Caucasian populations already cited, however, suggest otherwise.
Branstma et al. found that the predictive value of urine albumin excretion for diabetes was modified by the level of C-reactive protein. Urinary albumin excretion predicts the development of diabetes most strongly when the level of C-reactive protein is low.7 No significant interaction between C-reactive protein and urine ACR was observed in our study. This may be due to a relatively small sample size.
Compared with two previous studies,6,7 this study used a different study design and different measuring method for albuminuria in a different population. However, the findings that the presence of microalbuminuria and macroalbuminuria predicts the future development of diabetes are consistent.
|
Acknowledgments |
|---|
This work was funded by the National Health and Medical Research Council (NHMRC) of Australia (301024 and 320860). We especially thank the Tiwi people who participated in this study; the Tiwi Land Council and the local clinics for their help and support. The baseline data were collected by the renal research team at the Menzies School of Health Research, Darwin. David Ung at the Northern Territory Department of Health and Community Services assisted in the hospital data collection. We thank Mark Shephard from the Community Point-of-Care Services unit at the Flinders University Rural Clinical School for his contribution to CRP testing.
KEY MESSAGES
|
|
References |
|---|
|
TopAbstractSubjects, materials, and methodsResultsDiscussionReferences |
|---|
1 Wang Z, Hoy WE. Hypertension, dyslipidemia, body mass index, diabetes and smoking status in Aboriginal Australians in a remote community. Ethn Dis 2003;13:324–30.[Web of Science][Medline]
2 Weinstein AR, Sesso HD, Lee IM, et al. Relationship of physical activity vs body mass index with type 2 diabetes in women. JAMA 2004;292:1188–94.
3 Hoy WE, Wang Z, VanBuynder P, Baker PR, McDonald SM, Mathews JD. The natural history of renal disease in Australian Aborigines. Part 2. Albuminuria predicts natural death and renal failure. Kidney Int 2001;60:249–56.[CrossRef][Web of Science][Medline]
4 Wang Z, Hoy WE. Albuminuria and incident coronary heart disease in Australian Aboriginal people. Kidney Int 2005;68:1289–93.[CrossRef][Web of Science][Medline]
5 Damsgaard EM, Mogensen CE. Microalbuminuria in elderly hyperglycaemic patients and controls. Diabet Med 1986;3:430–35.[Web of Science][Medline]
6 Mykkanen L, Haffner SM, Kuusisto J, Pyorala K, Laakso M. Microalbuminuria precedes the development of NIDDM. Diabetes 1994;43:552–57.[Abstract]
7 Brantsma AH, Bakker SJ, Hillege HL, de Zeeuw D, de Jong PE, Gansevoort RT. Urinary albumin excretion and its relation with C-reactive protein and the metabolic syndrome in the prediction of type 2 diabetes. Diabetes Care 2005;28:2525–30.
8 Rowley KG, Iser DM, Best JD, O'Dea K, Leonard D, McDermott R. Albuminuria in Australian Aboriginal people: prevalence and associations with components of the metabolic syndrome. Diabetologia 2000;43:1397–403.[CrossRef][Web of Science][Medline]
9 McDonald SP, Maguire GP, Hoy WE. Renal function and cardiovascular risk markers in a remote Australian Aboriginal community. Nephrol Dial Transplant 2003;18:1555–61.
10 Hoy W, McDonald SP. Albuminuria: Marker or target in indigenous populations. Kidney Int Suppl 2004;S25–31.
11 Mogensen CE. Microalbuminuria and hypertension with focus on type 1 and type 2 diabetes. J Intern Med 2003;254:45–66.[CrossRef][Web of Science][Medline]
12 Mogensen CE, Keane WF, Bennett PH et al. Prevention of diabetic renal disease with special reference to microalbuminuria. Lancet 1995;346:1080–84.[CrossRef][Web of Science][Medline]
13 Yuyun MF, Khaw KT, Luben R et al. A prospective study of microalbuminuria and incident coronary heart disease and its prognostic significance in a British population: the EPIC-Norfolk study. Am J Epidemiol 2004;159:284–93.
14 Hoy WE, Wang Z, VanBuynder P, Baker PR, Mathews JD. The natural history of renal disease in Australian Aborigines. Part 1. Changes in albuminuria and glomerular filtration rate over time. Kidney Int 2001;60:243–48.[CrossRef][Web of Science][Medline]
15 Wang Z, Hoy WE. Association between diabetes and coronary heart disease in Aboriginal people: are women disadvantaged? Med J Aust 2004;180:508–11.[Web of Science][Medline]
16 Wang Z, Hoy WE. Waist circumference, body mass index, hip circumference and waist-to-hip ratio as predictors of cardiovascular disease in Aboriginal people. Eur J Clin Nutr 2004;58:888–93.[CrossRef][Web of Science][Medline]
17 World Health Organization. Definition, Diagnosis and Classification of Diabetes Mellitus and its Complications; Part 1: Diagnosis and Classification of Diabetes Melitus. Geneva: Department of Noncommunicable disease Surveillance, WHO, 1999.
18 StataCorp. Stata Statistical Software: Release 9.0. College Station. TX: StataCorp LP, 2005.
19 Lane JT. Microalbuminuria as a marker of cardiovascular and renal risk in type 2 diabetes mellitus: a temporal perspective. Am J Physiol Renal Physiol 2004;286:F442–50.
20 Corona AJ, Martinez DR, Avila MH et al. Microalbuminuria as a predictor of myocardial infarction in a Mexican population: the Mexico City Diabetes Study. Kidney Int Suppl 2005;S34–39.
21 Haffner SM, Stern MP, Gruber MK, Hazuda HP, Mitchell BD, Patterson JK. Microalbuminuria. Potential marker for increased cardiovascular risk factors in nondiabetic subjects? Arteriosclerosis 1990;10:727–31.
22 Yudkin JS, Forrest RD, Jackson CA. Microalbuminuria as predictor of vascular disease in non-diabetic subjects. Islington Diabetes Survey. Lancet 1988;2:530–33.[Web of Science][Medline]
23 Mykkanen L, Zaccaro DJ, Wagenknecht LE, Robbins DC, Gabriel M, Haffner SM. Microalbuminuria is associated with insulin resistance in nondiabetic subjects: the insulin resistance atherosclerosis study. Diabetes 1998;47:793–800.[Abstract]
24 Palaniappan L, Carnethon M, Fortmann SP. Association between microalbuminuria and the metabolic syndrome: NHANES III. Am J Hypertens 2003;16:952–58.[CrossRef][Web of Science][Medline]
25 Lorenzo C, Okoloise M, Williams K, Stern MP, Haffner SM. The metabolic syndrome as predictor of type 2 diabetes: the San Antonio heart study. Diabetes Care 2003;26:3153–59.
26 Ferrannini E. Insulin resistance versus insulin deficiency in non-insulin-dependent diabetes mellitus: problems and prospects. Endocr Rev 1998;19:477–90.
27 Gerich JE. Contributions of insulin-resistance and insulin-secretory defects to the pathogenesis of type 2 diabetes mellitus. Mayo Clin Proc 2003;78:447–56.
28 Harris MI, Klein R, Welborn TA, Knuiman MW. Onset of NIDDM occurs at least 4-7 yr before clinical diagnosis. Diabetes Care 1992;15:815–19.[Abstract]
29 Jensen T, Bjerre-Knudsen J, Feldt-Rasmussen B, Deckert T. Features of endothelial dysfunction in early diabetic nephropathy. Lancet 1989;1:461–63.[Web of Science][Medline]
30 McDonald SP, Maguire GP, Duarte N, Wang XL, Hoy WE. Carotid intima-media thickness, cardiovascular risk factors and albuminuria in a remote Australian Aboriginal community. Atherosclerosis 2004;177:423–31.[CrossRef][Web of Science][Medline]
31 Houlihan CA, Tsalamandris C, Akdeniz A, Jerums G. Albumin to creatinine ratio: a screening test with limitations. Am J Kidney Dis 2002;39:1183–89.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||