Low Muscular Mass and Overestimation of Microalbuminuria by Urinary Albumin/Creatinine Ratio
Microalbuminuria is a mild urinary albumin elevation and is associated with cardiovascular disease. Urinary albumin/creatinine ratio is recommended for microalbuminuria assessment, because it reflects urinary albumin excretion. Muscular mass could affect albumin/creatinine ratio, because urinary creatinine reflects muscular mass. The study investigated high albumin/creatinine ratio attributed to low urinary creatinine without microalbuminuria. The Gubbio Population Study for ages 45 to 64 collected data on weight, skinfold, urinary albumin, urinary creatinine, and coronary heart disease. Weight and skinfold thickness were used to calculate fat and nonfat mass and urinary creatinine as a marker of muscular mass. Microalbuminuria was defined as urinary albumin of 20 to 199 μg/min and high albumin/creatinine ratio as a ratio of 17 to 250 μg/mg in men and of 25 to 355 μg/mg in women. Persons with macroalbuminuria (urinary albumin ≥200 μg/min) were excluded to focus analyses on microalbuminuria. Coronary heart disease was defined by ECG and questionnaire. The target cohort consisted of 1623 men and women, ages 45 to 64. Prevalence was 8.5% for high albumin/creatinine ratio (n=138), 4.3% for microalbuminuria (n=69), 5.2% for high albumin/creatinine ratio without microalbuminuria (n=85), and 1.0% for nonhigh albumin/creatinine ratio with microalbuminuria (n=16). High albumin/creatinine ratio without microalbuminuria was inversely associated with nonfat mass and urinary creatinine (P<0.04). Compared with persons with a nonhigh albumin/creatinine ratio, coronary heart disease was more prevalent in persons with a high albumin/creatinine ratio and microalbuminuria (18.9% and 7.1%; P=0.002), not in persons with a high albumin/creatinine ratio without microalbuminuria (8.2% and 7.1%; P=0.706). A high albumin/creatinine ratio in persons with low muscle mass indicates low urinary creatinine more often than microalbuminuria and cardiovascular disease.
A statement of the American Heart Association and guidelines of the National Kidney Foundation suggest that the moderate elevation in urinary albumin excretion (uAE) defined as microalbuminuria should be considered as a risk factor for cardiovascular disease.1,2 The urinary albumin/creatinine ratio (uACR) is recommended for the assessment of uAE, because uACR is an index of uAE than can be measured in untimed spot urine samples. In fact, the uACR can be calculated with use of the urinary concentration of albumin and creatinine without information on duration and volume of urine collection. The value of the uACR necessarily depends not only on the rate of uAE but also on the rate of urinary creatinine excretion that, in turn, reflects the interindividual differences in muscle mass.3,4 Gender-specific thresholds for the definition of high uACR were proposed to reduce the confounding because of an underestimate of microalbuminuria, that is, a normal uACR because of the combination of high uAE with high urinary creatinine.5,6 Research data are missing on the opposite confounding, that is, on the overestimation of microalbuminuria caused by a high uACR because of the combination of normal uAE with low urinary creatinine excretion (ie, low muscle mass). Significant and independent associations were reported previously among uAE, cardiovascular risk factors, and coronary heart disease (CHD) in middle-age persons participating in the Gubbio Population Study.4,7,8 Data of the Gubbio Study were analyzed to investigate persons with high uACR and normal uAE, that is, on the elevation of the uACR because of low urinary creatinine excretion rather than high uAE. Analyses were focused on the prevalence of this trait, its association with CHD, and indices of body mass.
The Gubbio Study is an epidemiological investigation ongoing in the Italian city of Gubbio.9,10 The study adheres to the principles of the Declaration of Helsinki and the Title 45 of US Code of Federal Regulation. Study activities were approved by the local institutional review committee, and an informed consent was given by participants. All of the procedures were in accordance with institutional guidelines. The protocol for participants 45 to 64 years of ages included the collection of timed overnight urine,4 standardized questionnaires on cardiovascular disease,11 blood pressure measurements, a 12-lead resting ECG read by the Minnesota Code,11 weight, and skinfold thickness.12
As reported previously,9 blood pressure was measured by trained medical doctors using mercury sphygmomanometers and cuffs of appropriate size with the participant in the sitting position. The first measurement was done after 5 minutes of quiet rest after application of the cuff. The mean of second and third measurements was used for analyses. Albumin, creatinine, and urea were measured in urine samples: albumin by ultrafiltration immunoturbidimetry4 and creatinine and urea by automated biochemistry.13 The rate of uAE (in μg/min) was defined as normal when <20, microalbuminuria when in the range of 20 to 199, and macroalbuminuria when ≥200.14 The uACR (in μg/mg) was defined as nonhigh when <17 in men and <25 in women and high when ≥17 in men and ≥25 in women.1,2,6 Persons with macroalbuminuria were excluded from analyses to focus on uAE in the range of normalcy and microalbuminuria. Hypertension was defined as systolic pressure ≥140 mm Hg and/or diastolic pressure ≥90 mm Hg and/or regular treatment with antihypertensive drug(s). CHD was analyzed as the most common clinical presentation of vascular disease11 and defined with use of ECG data combined with symptoms reported in the questionnaire as described previously.8 Myocardial infarction or ischemia were combined in analysis, because previous data showed that uAE is similarly and independently related to both disorders.8 Data of skinfold thickness (triceps and subscapular) and weight were used for the estimation of fat mass.12 Nonfat mass was calculated as weight minus fat mass.12 Urinary creatinine excretion (in mg/min) was used as index of muscle mass.15 Urinary urea (in mg/min) was used as index of dietary protein intake13,16 to control the confounding of diet composition on urinary creatinine.17
Statistical procedures included Pearson and rank correlation analysis, McNemar’s test for paired observations of categorical data, χ2 analysis, logistic regression analysis, ANOVA, and analysis of odds ratio (OR) with 95% CI. Data are reported as mean±SD unless otherwise indicated.
A total of 1684 persons were 45 to 64 years of age in the Gubbio population sample. Sixty-one persons were excluded from analyses because of missing ECG data (n=52) or macroalbuminuria (n=9). Thus, the study cohort was composed of 1623 persons. Table 1 shows descriptive statistics in the study cohort. Descriptive data on smoking, hypercholesterolemia, diabetes mellitus, infarction, and ischemia were reported previously.4,7,8 Differences between men and women were significant for weight, fat mass, nonfat mass, urinary creatinine, and microalbuminuria prevalence. Urinary creatinine excretion correlated with weight and nonfat mass (R=0.403 and 0.504; P<0.001) but not with fat mass (R=0.022; P= 0.379). Urinary creatinine also correlated with urinary urea (R=0.485; P<0.005). The correlation between urinary creatinine excretion and nonfat mass was also significant with control for urinary urea excretion (partial R=0.476; P<0.001). Findings were similar in analyses separate by gender (data not shown).
Agreement Between uACR and uAE
Individual values of uAE and uACR were significantly correlated in analysis for men and women combined together (rank correlation coefficient=0.884; P<0.001) and in analyses separate by gender (rank correlation coefficient=0.921 and 0.915; P<0.001). Table 2 reports data on the prevalence of normal uAE and microalbuminuria by the level of uACR. The overall accuracy of uACR for the prediction of uAE was 93.8%. The negative predictive power of uACR was high (98.9%), because there was a substantial agreement for definitions of nonhigh uACR and normal uAE. The positive predictive power of uACR was low (38.4%), because high uACR was much more prevalent than microalbuminuria (P<0.001 by McNemar test). The positive predictive power of uACR was similarly low in men and women (39.4% and 37.5%, respectively), as well as in analysis limited to hypertensive subjects (44.9%).
High uACR Without Microalbuminuria
Prevalence and Correlates in the Population
The combination of high uACR with normal uAE (defined as high uACR without microalbuminuria) was found in 5.2% of the cohort (n=85 of 1623). Gender did not correlate with high uACR without microalbuminuria, which was similarly prevalent in men and women (5.6% and 4.9%, respectively; P=0.595). Univariate logistic analyses were done to investigate the relation of high uACR without microalbuminuria with age and indices of body mass (Table 3). Prevalence of high uACR without microalbuminuria was significantly associated with age, nonfat mass, and urinary creatinine excretion but not with weight and fat mass. The association was positive for age and negative for nonfat mass and urinary creatinine excretion. Findings were similar in analyses separate by gender: high uACR without microalbuminuria was associated in men and women directly with age (OR for +5.6 years=1.23 and 1.66, respectively; P=0.205 and 0.002, respectively), inversely with nonfat mass (OR for +10.0 kg=0.56 and 0.82, respectively; P=0.021 and 0.037, respectively) and urinary creatinine excretion (OR for +0.44 mg/min=0.07 and 0.01, respectively; P<0.001). Findings were similar also in analyses limited to hypertensive subjects: high uACR without microalbuminuria was associated directly with age (OR for +5.6 years=1.42; P=0.042) and inversely with nonfat mass (OR for +10.0 kg=0.82; P=0.088) and urinary creatinine excretion (OR for +0.44 mg/min=0.09; P<0.001).
A multivariate logistic analysis was done to assess whether the association of high uACR without microalbuminuria with age and nonfat mass was independent of urinary creatinine excretion (index of muscular mass). In the logistic model with age, nonfat mass, and urinary creatinine excretion included together as independent variables (men and women combined, whole cohort), high uACR without microalbuminuria was associated with urinary creatinine excretion (OR for +0.44 mg/min=0.03; P<0.001) but not with nonfat mass and age (P>0.4).
Relation of uACR to CHD
Differences by Level of uAE
Figure shows data for men and women combined on CHD prevalence in the group of persons with nonhigh uACR and the group of persons with high uACR. Persons with high uACR were also analyzed as 2 subgroups: the subgroup without microalbuminuria and the subgroup with microalbuminuria. CHD prevalence was significantly higher in the high uACR group than in the nonhigh uACR group. Findings were similar for myocardial infarction (prevalence in high uACR group and nonhigh uACR group=4.3% and 2.4%, respectively; OR, 1.83; 95% CI, 0.76 to 4.42) and ischemia (prevalence in high uACR group and nonhigh uACR group=8.0% and 5.0%, respectively; OR, 1.65; 95% CI, 0.86 to 3.19). The excess of CHD prevalence in the high uACR group reflected the combination of dissimilar data between the subgroups with and without microalbuminuria. In comparison to the nonhigh uACR group, CHD prevalence was significantly increased in the high uACR subgroup with microalbuminuria but not in the high uACR subgroup without microalbuminuria. CHD was 2.59-times more prevalent in the high uACR subgroup with microalbuminuria than in the high uACR subgroup without microalbuminuria (95% CI, 0.92 to 7.30; P=0.065). Findings were similar in analyses by gender: CHD prevalence was higher in the high uACR subgroup with microalbuminuria than in the high uACR subgroup without microalbuminuria among men (19.2% and 2.5%, respectively; OR=9.29; P=0.021) and women (18.5% and 13.3%, respectively; OR=1.48; P=0.554). Findings were also similar in analyses limited to hypertensive subjects: CHD prevalence was higher in hypertensive subjects with high uACR and microalbuminuria than in hypertensive subjects with high uACR without microalbuminuria (25.2% and 8.2%, respectively; OR=3.75; P=0.030).
Analyses for Persons With High uACR
Differences by Level of uAE and by Nonfat Mass
Table 4 shows gender, age, and indices of body mass in the group persons with high uACR (n=138) divided in the subgroups without microalbuminuria and with microalbuminuria. In comparison to the subgroup with microalbuminuria, the subgroup without microalbuminuria differed for higher age (difference with borderline significance), lower weight, lower nonfat mass, and lower urinary creatinine excretion but not for gender and fat mass.
The relation between nonfat mass and uAE in persons with high uACR was additionally investigated with use of nonfat mass tertiles. To control for the effect of gender on nonfat mass (shown in Table 1), cutoff points of tertiles were defined by analysis of nonfat mass distribution separately in men and women with high uACR. Table 5 shows that, among persons with high uACR, nonfat mass was linearly related to uAE (both prevalence of normal uAE and mean uAE), urinary creatinine excretion, and CHD in the absence of significant differences for gender and age. The relation of nonfat mass to CHD among persons with high uACR was explained by the confounding of uAE, because it was significant in univariate analysis (OR of CHD for +10 kg in nonfat mass, 2.24; P=0.017; 95% CI, 1.15 to 4.34) but not with control for uAE (OR of CHD for +10 kg in nonfat mass, 1.19; P=0.500, 95% CI, 0.71 to 2.01).
The study reports new findings on the confounding because of the use of the uACR for assessment of microalbuminuria. First, the prevalence of microalbuminuria was overestimated with use of uACR, because &60% of persons with high uACR do not have microalbuminuria. Second, persons with high uACR but without microalbuminuria did not have the excess of CHD prevalence typical of persons with microalbuminuria. Third, the combination of high uACR with normal uAE was associated with low nonfat mass or low muscular mass in analyses for the whole population sample and in analyses limited to persons with high uACR. Fourth, in persons with low nonfat mass, high uACR reflected low urinary creatinine excretion rather than microalbuminuria and vascular disease.
Various methodological limitations could have affected the results of the study. The errors in urine collection precision, certainly frequent in a population-based study, should have diluted the results for uAE but not those for uACR, because the uACR calculation is independent of duration and volume of urine collection. Thus, because of this limitation, the results might underestimate the strength of the association between high uAE and CHD but not between high uACR and CHD. The use of a single urine sample likely caused errors in the classification of normal/high values because of daily variability in urinary albumin.14 This misclassification should have similarly influenced the results for uAE and uACR, because these 2 indices were measured in the same urine sample. The same should be true for the misclassification attributed to absence of urinalysis that could be of help for definition of persons with high urinary albumin secondary to urinary tract infection.14 The exclusion from analysis of data for other vascular diseases should have had only a minor role, unless urinary albumin and urinary creatinine (ie, muscular mass) were associated with those diseases differently than CHD. The use of skinfold data and of urinary creatinine could have caused an inaccurate assessment of muscular mass, although the finding of an independent relation between nonfat mass and urinary creatinine is against this possibility. A different preexamination mortality between persons with low and high muscular mass could have confounded some of the results. Finally, the cross-sectional design of the analysis did not exclude the possibility of changes in urine composition secondary to CHD,18 a point that was behind the aims of the study yet was important for the comprehension of the mechanisms underlying the association urinary albumin and CHD.
Taken together, study results indicate that interindividual differences in muscular mass substantially influence the value of the uACR via urinary creatinine excretion. Because of this influence, persons with low muscular mass often have the moderate uACR elevation used for the definition of microalbuminuria in the absence of a true elevation in absolute uAE. Thus, the uACR has a low predictive power for the definition of microalbuminuria implying the overestimation of the true prevalence of microalbuminuria in persons with low urinary creatinine because of low muscular mass. The finding that the group of persons with high uACR and low muscular mass was without an excess of CHD prevalence additionally supports this interpretation, because present and previous data consistently show an independent association between microalbuminuria and CHD.8,19 Within the middle-age sample of the Gubbio Population Study, the overestimation of microalbuminuria because of low urinary creatinine was found in &60% of persons with high uACR. Muscular mass and urinary creatinine excretion undergo parallel reductions during aging.20,21 Thus, a reasonable inference is that the frequency of the overestimation of microalbuminuria could progressively increase from young to older ages. The finding in this study of a positive relationship between age and the prevalence of overestimation of microalbuminuria supports this possibility. At variance with the low value of the positive predictive power, uACR appears to have a high negative predictive power, because a person with nonhigh uACR was unlikely to have microalbuminuria.
A practical implication of the study is that low muscle mass is an important confounder in the use of uACR as marker of microalbuminuria and cardiovascular disease in the population. Physicians should know that in persons with low muscular mass, a high uACR may be because of low urinary creatinine excretion rather than high uAE and may not be associated with vascular disease. The strong cross-sectional association between microalbuminuria and CHD in the middle-age population might explain, at least in part, the predictive power of microalbuminuria for the incidence of cardiovascular disease, because the presence of CHD is, per se, a predictor of cardiovascular events.
The Gubbio Population Study was carried out in collaboration with Merck Sharp and Dohme, Rome, Italy; Gubbio Center of Preventive Medicine, Gubbio, Italy; Gubbio Hospital, Gubbio, Italy; “Federico II” University of Naples, Italy; University of Milan, Italy; Northwestern University of Chicago, Chicago, IL; Istituto Superiore di Sanità, Rome, Italy; Second University of Naples, Naples, Italy.
- Received September 23, 2005.
- Revision received October 23, 2005.
- Accepted November 17, 2005.
Sarnak MJ, Levey AS, Schoolwerth AC, Coresh J, Culleton B, Hamm LL, McCullough PA, Kasiske BL, Kelepouris E, Klag MJ, Parfrey P, Pfeffer M, Raij L, Spinosa DJ, Wilson PW. Kidney disease as a risk factor for development of cardiovascular disease - A statement from the American Heart Association Councils on kidney in cardiovascular disease, high blood pressure research, clinical cardiology, and epidemiology and prevention. Hypertension. 2003; 42: 1050–1065.
Cirillo M, Senigalliesi L, Laurenzi M, Alfieri R, Stamler J, Stamler R, Panarelli W, De Santo NG. Microalbuminuria in non-diabetic adults: relation of blood pressure, body mass, plasma cholesterol, and smoking – The Gubbio Population Study. Arch Intern Med. 1998; 158: 1933–1939.
Jacobs DR Jr, Murtaugh MA, Steffes M, Yu X, Roseman J, Goetz FC. Gender- and race-specific determination of albumin excretion rate using albumin-to-creatinine ratio in single, untimed specimens: the Coronary Artery Risk Development in Young Adults Study. Am J Epidemiol. 2002; 155: 1114–1119.
Mattix HJ, Hsu C-Y, Shaykevich S, Curhan G. Use of the albumin/creatinine ratio to detect microalbuminuria: implications of sex and race. J Am Soc Nephrol. 2002; 13: 1034–1039.
Cirillo M, Laurenzi M, Trevisan M, Stamler J. Hematocrit, blood pressure, and hypertension - The Gubbio Population Study. Hypertension. 1992; 20: 1305–1311.
Laurenzi M, Cirillo M, Panarelli W, Trevisan M, Stamler R, Dyer AR, Stamler J. Baseline sodium-lithium countertransport and 6-year incidence of hypertension: the Gubbio Population Study. Circulation. 1997; 95: 581–587.
Rose GA, Blackburn H, Gillman RF, Prineas RJ. Cardiovascular Survey Methods. Geneva, Switzerland: World Health Organization; 1982.
Ruilope LM, Rodicio JL. Microalbuminuria in clinical practice. Kidney. 1995; 4: 211–216.
Heymsfield SB, Arteaga C, McManus C, Smith J, Moffitt S. Measurement of muscle mass in humans: validity of the 24-hour urinary creatinine method. Am J Clin Nutr. 1983; 37: 478–494.
Stamler J, Elliott P, Kesteloot H, Nichols R, Claeys G, Dyer AR, Stamler R. Inverse relation of dietary protein markers with blood pressure. Findings for 10,020 men and women in the INTERSALT Study. INTERSALT Cooperative Research Group. INTERnational study of SALT and blood pressure. Circulation. 1996; 94: 1629–1634.
Walser M. Creatinine excretion as a measure of protein nutrition in adults of varying age. J Parenter Enteral Nutr. 1987; 11: 73S–78S.
Berton G, Citro T, Palmieri R, Petucco S, De Toni R, Palatini P. Albumin excretion rate increases during acute myocardial infarction and strongly predicts early mortality. Circulation. 1997; 96: 3338–3345.
Diercks GF, van Boven AJ, Hillege HL, Janssen WM, Kors JA, de Jong PE, Grobbee DE, Crijns HJ, van Gilst WH. Microalbuminuria is independently associated with ischaemic electrocardiographic abnormalities in a large non-diabetic population. The PREVEND study. Eur Heart J. 2000; 21: 1922–1927.