Prevalence of elevated serum creatinine level by the Sixth Report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure blood pressure category and self-reported treatment with antihypertensive medications. Error bars indicate SEs.
Estimated number of individuals with elevated serum creatinine by the Sixth Report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure blood pressure category and self-reported treatment with antihypertensive medications. Error bars indicate SEs.
Coresh J, Wei GL, McQuillan G, Brancati FL, Levey AS, Jones C, Klag MJ. Prevalence of High Blood Pressure and Elevated Serum Creatinine Level in the United StatesFindings From the Third National Health and Nutrition Examination Survey (1988-1994). Arch Intern Med. 2001;161(9):1207-1216. doi:10.1001/archinte.161.9.1207
Copyright 2001 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2001
The prevalence and incidence of end-stage renal disease in the United States are increasing, but milder renal disease is much more common and may often go undiagnosed and undertreated.
A cross-sectional study of a representative sample of the US population was conducted using 16 589 adult participants aged 17 years and older in the Third National Health and Nutrition Examination Survey (NHANES III) conducted from 1988 to 1994. An elevated serum creatinine level was defined as 141 µmol/L or higher (≥1.6 mg/dL) for men and 124 µmol/L or higher (≥1.4 mg/dL) for women (>99th percentile for healthy young adults) and was the main outcome measure.
Higher systolic and diastolic blood pressures, presence of hypertension, antihypertensive medication use, older age, and diabetes mellitus were all associated with higher serum creatinine levels. An estimated 3.0% (5.6 million) of the civilian, noninstitutionalized US population had elevated serum creatinine levels, 70% of whom were hypertensive. Among hypertensive individuals with an elevated serum creatinine level, 75% received treatment. However, only 11% of all individuals with hypertension had their blood pressure reduced to lower than 130/85 mm Hg (the Sixth Report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure recommendation for hypertensive individuals with renal disease); 27% had a blood pressure lower than 140/90 mm Hg. Treated hypertensive individuals with an elevated creatinine level had a mean blood pressure of 147/77 mm Hg, 48% of whom were prescribed one antihypertensive medication.
Elevated serum creatinine level, an indicator of chronic renal disease, is common and strongly related to inadequate treatment of high blood pressure.
ADEQUATE BLOOD pressure control is widely recognized as an essential factor in slowing the progression of chronic renal disease and preventing its main sequelae, end-stage renal disease (ESRD) and cardiovascular disease.1- 7 As a result, antihypertensive blood pressure treatment goals are lower for individuals with both hypertension and chronic renal disease (<130/85 mm Hg for individuals with <1 g/d of proteinuria and <125/75 mm Hg for individuals with >1 g/d of proteinuria) than for hypertensive individuals without target organ damage (<140/90 mm Hg).6 At the same time, individuals with chronic renal disease are usually asymptomatic and often go undiagnosed. Thus, hypertension may be undertreated among patients with chronic renal disease. Data on the adequacy of blood pressure treatment in this group are critical for preventive efforts to stem the epidemic of ESRD8 and cardiovascular disease in chronic renal disease.9
The Third National Health and Nutrition Examination Survey (NHANES III) data offer the first opportunity to study the prevalence and number of people with chronic renal disease (detected by an elevated serum creatinine level) in a nationally representative sample. An initial analysis from NHANES III showed that the prevalence of elevated serum creatinine level was higher among non-Hispanic blacks than non-Hispanic whites and higher among older than younger individuals.10 The present analysis was undertaken to measure the prevalence of elevated serum creatinine level across different categories of blood pressure and antihypertensive medication use. In addition, we examined the number and type of antihypertensive medications used and the achieved blood pressure among individuals with elevated serum creatinine levels who were prescribed medicine for hypertension. We hypothesize that many individuals with both elevated serum creatinine levels and hypertension are inadequately treated.
We estimated the burden of hypertension-related chronic renal disease using 16 589 participants aged 17 years and older in NHANES III. This survey, conducted from 1988 to 1994 by the National Center for Health Statistics (Hyattsville, Md) of the Centers for Disease Control and Prevention (Atlanta, Ga), provides cross-sectional, nationally representative data on the health and nutritional status of the civilian, noninstitutionalized US population.11,12 Non-Hispanic blacks and Mexican Americans as well as the elderly and children were deliberately oversampled in this survey. This oversampling makes it possible to obtain reliable estimates of the distribution of creatinine in the 2 largest minority groups of the civilian, noninstitutionalized US population as well as in a broad range of age groups. Standardized questionnaires were administered in the home, followed by a detailed physical examination at a mobile examination center.
Blood pressure measurements were obtained 3 times during the home interview and another 3 times during the examination. Each measurement was made using a mercury sphygmomanometer, with the participant seated. The arithmetic mean was then calculated using all available systolic and diastolic readings. In this analysis, individuals with a systolic blood pressure lower than 140 mm Hg and diastolic blood pressure lower than 90 mm Hg who were not receiving antihypertensive treatment were defined as normotensive. Individuals were classified as hypertensive if they had a mean blood pressure of 140 mm Hg or higher (systolic) or 90 mm Hg or higher (diastolic) or reported current use of medication for hypertension.13 Among the hypertensive participants, we further categorized individuals by treatment status. Individuals were considered not treated if they had hypertension at the time of the survey and did not report taking antihypertensive medication, regardless of whether these individuals had been previously diagnosed as hypertensive. We also analyzed the medications prescribed and classified them as angiotensin-converting enzyme inhibitors, calcium channel blockers, β-blockers, α-blockers or α-agonists, or diuretics.
Serum was collected at the mobile examination center and creatinine measurements were performed at the White Sands Research Center laboratory, Almogordo, NM, by the modified kinetic Jaffe reaction14 using a Hitachi 737 analyzer (Boehringer Mannheim Corporation, Indianapolis, Ind) and were reported using conventional units (1 mg/dL = 88.4 µmol/L). The coefficient of variation for creatinine determination ranged from 0.2% to 1.4% during the 6 years of the NHANES III study. Data on physiologic variation in creatinine level were obtained in a sample of 1921 participants who had a second creatinine measurement. The assay had very stable quality-control measures for the duration of the study. A review of the College of American Pathologists Survey data indicates that the laboratory mean value for serum creatinine levels during 1992-1994 was within acceptable limits but higher than the mean value of all laboratories surveyed.
Blood pressure was categorized according to the Sixth Joint National Committee Report on Detection, Evaluation, and Treatment of High Blood Pressure (JNC-VI).6 Persons were categorized into 6 groups based on the higher of their systolic or diastolic blood pressure measurement: (1) optimal blood pressure (<120 mm Hg systolic and <80 mm Hg diastolic); (2) normal blood pressure (120-130 mm Hg systolic and 80-85 mm Hg diastolic); (3) high-normal blood pressure (130-139 mm Hg systolic or 85-89 mm Hg diastolic); (4) stage 1 hypertension (140-159 mm Hg systolic or 90-99 mm Hg diastolic); (5) stage 2 hypertension (160-179 mm Hg systolic or 100-109 mm Hg diastolic); and (6) stage 3 hypertension (≥180 mm Hg systolic or ≥110 mm Hg diastolic).
The primary outcome measure in this analysis was elevated serum creatinine level (defined as a sex-specific cutoff point of ≥141 µmol/L [≥1.6 mg/dL] for men and ≥124 µmol/L [≥1.4 mg/dL] for women). These cutoffs, which are higher than the reported reference range for serum creatinine level using this assay, were used to provide greater specificity when defining chronic renal disease based on a single serum creatinine measurement. These criteria correspond to the 99.4th and 99.8th percentiles for men and women aged 20 to 39 years without diabetes or hypertension in this study. In this young healthy group, the mean (SD) creatinine level was 101 (13) µmol/L (1.14 [0.15] mg/dL) for men and 80 (15) µmol/L (0.91 [0.17] mg/dL) for women (29% and 35% lower than the cutoffs for elevated serum creatinine level). Because serum creatinine level is inversely proportional to glomerular filtration rate, young individuals with elevated serum creatinine levels are likely to have lost approximately a third of their renal function while older persons with elevated creatinine levels will have lost even more function because muscle mass and creatinine production decreases with age. Analyses were repeated using alternative cutoff points and yielded similar results.
Diabetes was defined by participants' medical history as well as their blood glucose value. The primary analysis stratified individuals based on a history of diagnosed diabetes mellitus because this information was available for nearly all individuals and could be used by physicians for risk stratification. Ancillary analyses examined the impact of using the American Diabetes Association (ADA) criteria15 for diabetes mellitus in the subset of individuals who fasted at least 8 hours.
The complex survey design of NHANES III incorporated differential probabilities of selection. To derive national estimates, sampling weights were used to adjust for noncoverage and nonresponse. All prevalence estimates were weighted to represent the civilian, noninstitutionalized US population and to account for oversampling and nonresponse to the household interview and physical examination.12 All data analyses were conducted using STATA svy commands (Stata Corporation, College Station, Tex; 1999) for analyzing complex survey design data with 49 strata and 98 primary sampling units. A total of 16 589 participants (82.7%) of 20 050 examined had both blood pressure and serum creatinine data available for this analysis. The rate of missing data was higher among older than younger individuals (individuals missing data were 4 years older), higher among men than among with women (17.9% vs 16.7%), and lower among Mexican Americans and other ethnicities (14%) than among non-Hispanic whites (19%) and non-Hispanic blacks (18%). These differences were primarily owing to missing phlebotomy data. To minimize bias, the combined mobile examination center and home examination weights were divided by the proportion of participants missing creatinine data in each of the design age, sex, and race ethnicity strata. This corrected differences caused by missing data across sampling strata but assumed that data are missing randomly within strata.
Table 1 gives the number of survey participants by demographic group, diabetes mellitus, and blood pressure category as well as estimates of the proportion of individuals, mean serum creatinine levels, and prevalence of elevated creatinine levels for the civilian, noninstitutionalized US population. Overall, from 1988 to 1994, the prevalence of elevated serum creatinine level was 3.0%, corresponding to 5.6 million adults (803 survey participants). Older age and diabetes mellitus were very strongly associated with a higher prevalence of elevated creatinine level. Non-Hispanic blacks had a higher prevalence and Mexican Americans had a lower prevalence of elevated creatinine level than non-Hispanic whites. Higher categories of blood pressure, hypertension status, and use of antihypertensive medications were associated with higher mean serum creatinine level and a higher prevalence of elevated serum creatinine level. Elevated serum creatinine level was 8 times more common in hypertensive (9.1%) than normotensive (1.1%) individuals and 8 times more common in people already using medication for high blood pressure compared with people not using such medication (13.0% vs 1.6%).
Figure 1 shows that the prevalence of elevated serum creatinine level was markedly higher in individuals treated for hypertension compared with untreated individuals at every category of blood pressure. Among untreated individuals, prevalence of elevated serum creatinine level increased monotonically with higher blood pressure categories from 0.8 % to 13.6%. Among treated individuals, the association of prevalence of elevated serum creatinine level with blood pressure was J-shaped. Prevalence was lower in individuals with normal blood pressure (6.3%) compared with individuals with optimal blood pressure (9.8%) or higher levels of blood pressure. The higher prevalence of elevated creatinine level among treated hypertensive individuals with an optimal blood pressure cannot be estimated reliably given the relatively small size of this group (189 participants; 95% confidence interval [CI], 3%-16%). In persons with high-normal blood pressure, the prevalence was 13.6%, followed by a steady increase from 13.3% to 22.5% in hypertension stages 1 to 3. The ratio of prevalence rates of treated to untreated individuals declined with increasing blood pressure category from 12.2 for optimal blood pressure category to 1.7 for stage 3 hypertension; hence, the relative difference in prevalence between the untreated and treated groups narrowed at categories of higher blood pressure.
Figure 2 shows that the distribution of individuals with elevated serum creatinine levels across blood pressure categories is markedly different from the association of prevalence rates with blood pressure categories shown in Figure 1. The prevalence of elevated serum creatinine level was greatest in persons with stage 3 hypertension. In contrast, the largest number of the individuals with an elevated serum creatinine level was in the stage 1 hypertension category (Figure 2), the most common form of hypertension (56% of all hypertensive individuals). More individuals with an elevated serum creatinine level were treated than not treated for blood pressure categories of high-normal and above. The largest number of cases within a specific category was in treated persons with blood pressure in the stage 1 hypertension category (n = 1.1 million treated and 0.5 million untreated or 29% of all individuals with an elevated serum creatinine level). Among the hypertensive individuals with elevated serum creatinine levels, 75% were treated, of whom only 36% had a blood pressure lower than 140/90 mm Hg (the JNC-VI recommendation for hypertensive individuals without target organ damage) and only 14% had their blood pressure reduced to lower than 130/85 mm Hg (the JNC-VI recommendation for hypertensive individuals with renal disease).6,16 Overall, only 11% of all hypertensive individuals with an elevated serum creatinine level had their blood pressure reduced to lower than 130/85 mm Hg and 27% had a blood pressure lower than 140/90 mm Hg. Using lower or higher cutoff values to define elevated serum creatinine level influenced the prevalence of elevated creatinine level (11.7% of men had a creatinine level of ≥124 µmol/L [≥1.4 mg/dL] and 9.3% of women had a creatinine level of ≥106 µmol/L [≥1.2 mg/dL]) but not the shape of the association with blood pressure category.
Table 2 also gives the prevalence and number of cases of elevated serum creatinine level in the major ethnic subgroups as well as among older individuals. The number of individuals in subgroup analyses was smaller, which resulted in decreased precision as indicated by the higher standard errors. The prevalence rate of elevated serum creatinine level among treated hypertensive individuals was greatest in non-Hispanic blacks (19.3%), followed by non-Hispanic whites (11.9%) and Mexican Americans (8.1%). Among individuals older than 60 years, the prevalence of elevated creatinine level was higher among treated (18.0%) than untreated (6.9%) individuals. Seventy-seven percent of all individuals with elevated creatinine were at least 60 years old.
Individuals using antihypertensive medication for controlling blood pressure consistent with stage 1 hypertension made up the greatest proportion of cases in each group. The prevalence and estimated number of individuals in the US population with elevated serum creatinine levels stratified by self-reported diabetes status are given in Table 3. Diabetes was diagnosed in 1.1 million (19%) of the 5.6 million individuals with an elevated serum creatinine level. The prevalence of elevated serum creatinine level was higher at higher blood pressure categories in both individuals with diabetes mellitus and individuals without diabetes mellitus, which is similar to the total population. Individuals with diabetes mellitus treated for high blood pressure (18.4%) had a higher prevalence of elevated serum creatinine level than treated individuals without diabetes mellitus (12.0%). Untreated individuals with diabetes mellitus (7.8%) had a higher prevalence of elevated serum creatinine level than untreated individuals without diabetes mellitus (1.4%). The associations with blood pressure categories were similar to those in the total population. The inclusion of undiagnosed individuals with both diabetes mellitus and fasting hyperglycemia (2.7% of the population17) increased the prevalence of diabetes mellitus and slightly decreased the prevalence of elevated serum creatinine level among individuals with diabetes mellitus, but it altered very little the prevalence among the much larger number of individuals without diabetes mellitus.
Hypertensive individuals were more likely to be non-Hispanic black and older and have diabetes mellitus than the rest of the general population. With the use of logistic regression analysis (after adjustment for age, race, sex, diabetes, and a history of stroke, congestive heart failure, and cardiovascular disease) the prevalence of elevated serum creatinine level was still directly associated with higher blood pressure category, with the highest risk seen at stage 3 hypertension (Table 4; P = .001). Repeating the analysis by estimating an adjusted odds ratio for each of the blood pressure categories by hypertension treatment group compared with individuals with optimal blood pressure without treatment (last 2 columns of Table 4) yielded wider CIs and some attenuation of the association. However, the prevalence of elevated serum creatinine level remained associated with both elevated blood pressure and blood pressure treatment.
Table 5 gives the number of antihypertensive medications used to treat hypertensive individuals by the presence of elevated serum creatinine level and achieved mean blood pressure (this analysis is unweighted owing to small numbers in some data cells). Hypertensive individuals with an elevated serum creatinine level were more likely to be treated than hypertensive individuals with a lower serum creatinine level (75% vs 49%; P<.001). When treated, persons with an elevated serum creatinine level were prescribed a higher mean number of medications (1.7 vs 1.4; P = .001) and, consequently, were more likely to be prescribed every type of antihypertensive medication. In a logistic regression model among individuals with hypertension (which adjusted for the number of medications prescribed, age, sex, race, and diabetes) the presence of elevated creatinine level was positively associated with the prescription of diuretics (odds ratio, 1.7; 95% CI, 1.4-2.2), negatively associated with prescription of β-blockers (odds ratio, 0.7; 95% CI, 0.5-0.9), and not associated with prescription of angiotensin-converting enzyme inhibitors (odds ratio, 0.8; 95% CI, 0.6-1.0) or other agents.
Regardless of the number of medications they received, individuals with an elevated serum creatinine level had a slightly higher mean systolic blood pressure than individuals with a lower serum creatinine level (significant before but not after adjustment for age). In contrast, their mean diastolic blood pressure was slightly, although often not significantly, lower with 48% of individuals prescribed only 1 medication. Individuals prescribed 2 medications had a lower mean blood pressure than individuals prescribed 1 medication, despite the tendency to prescribe more medications to individuals with more severe underlying hypertension.
The analyses were repeated using different definitions of antihypertensive medication treatment and different cutoffs for elevated serum creatinine level to examine the consistency of our results. The proportion of individuals with an elevated serum creatinine level who were taking antihypertensive medications for any indication was 66%, somewhat higher than the percentage of these individuals reporting medical treatment for hypertension (52%). However, the pattern of the association between elevated serum creatinine level and blood pressure category as well as medication use was similar.
A sample of 1921 NHANES III participants had a second creatinine measurement at a mean (SD) of 17.5 (8.0) days after the initial examination. The second creatinine measurement showed good agreement with the initial creatinine measurement. The mean (SD) percent difference between the 2 measurements was 0.2% (9.7%). The percent difference was independent of the time difference between the visits and the absolute creatinine level. This amount of measurement error was added to the observed creatinine data in a computer simulation to estimate its possible impact on prevalence estimates. The prevalence of elevated serum creatinine level increased by a factor of 1.19 in women and 1.25 in men, suggesting that the effect of measurement error and short-term physiologic variation on our estimates of the prevalence of elevated serum creatinine level was moderate.
Our results show that approximately 5.6 million individuals in the civilian, noninstitutionalized US population have elevated serum creatinine levels, 25 times the number of prevalent ESRD cases, and 108 times the number of incident ESRD cases in 1991 (the midpoint of NHANES III). Elevated serum creatinine level is rare among young and middle-aged individuals with optimal blood pressure (<1% prevalence) and becomes increasingly more common among individuals in higher blood pressure categories as well as among individuals using antihypertensive medications. Most (75%) hypertensive persons with elevated serum creatinine levels in the US population are treated with antihypertensive medications, but most are treated suboptimally. These cross-sectional data are consistent with the hypothesis that poorly controlled hypertension is responsible for much of the high-burden chronic kidney disease, but the data can only demonstrate a strong association, not temporal sequence or causation. The largest number of individuals with an elevated serum creatinine level have blood pressure in the stage 1 hypertension category. Only half of the treated hypertensive individuals with an elevated serum creatinine level are receiving more than 1 antihypertensive medication, and only 36% of them meet the blood pressure guidelines for individuals without target organ damage (<140/90 mm Hg, also the target blood pressure for all hypertensive individuals since 198816,18), while only 14% of treated hypertensive individuals meet the blood pressure guidelines proposed in 199119 for persons with renal disease (<130/85 mm Hg). Given that effective antihypertensive treatment can slow the progression of renal disease, this finding implies that improved management of hypertension might have a major impact on stemming the epidemic of ESRD and cardiovascular disease in chronic renal disease.
Individuals with chronic renal disease have an increased risk of developing ESRD or death.20- 23 End-stage renal disease is a major public health problem in the United States. Data from the US Renal Data System shows a prevalence of 304 083 treated ESRD cases (110/100 000) and an incidence of 79 102 cases (28.7/100 000 person years) in 1997.8 Mortality among patients treated by dialysis is approximately 23% per year, with cardiovascular disease being the major cause of death. Annual treatment costs for patients treated with dialysis exceeded $15 billion in 1997.
High blood pressure is an important independent predictor of the development and progression of chronic renal disease as well as morbidity and mortality in patients with chronic renal disease.2,24,25 Prospective studies have shown the relationship between blood pressure and incidence of renal disease to be positive and continuous throughout the entire spectrum of blood pressure categories.2,24,26 As a result, the JNC-VI guidelines suggest lower target blood pressure goals for individuals with renal disease (<130/85 mm Hg for individuals without proteinuria and <125/75 mm Hg for individuals with proteinuria).6 Experience from recent clinical trials (Hypertension Optimal Treatment [HOT], Antihypertensive and Lipid Lowering to Prevent Heart Attack [ALLHAT], United Kingdom Prospective Diabetes Study [UKPDS], and Controlled Onset Verapamil Investigation of Cardiovascular Endpoints [CONVINCE]) has shown that low blood pressure goals can be achieved with high rates of success.27 More than 90% of the patients in the HOT trial achieved the diastolic blood pressure goal of 90 mm Hg or lower. The 470 HOT trial participants with a serum creatinine level of 132.6 µmol/L (1.5 mg/dL) or higher required more antihypertensive medications than patients with a lower serum creatinine level (2.8 vs 2.4 medication escalation steps) but achieved their blood pressure target just as often (71% vs 70%).28 Participants in the African-American Study of Kidney Disease and Hypertension (AASK) had a glomerular filtration rate of 70 mL/min per 1.73 m2 or lower at baseline and were randomized to a mean arterial pressure lower than 92 mm Hg or 102 to 107 mm Hg. At entry, 66% had a mean arterial pressure higher than 107 mm Hg. After 4 months on the assigned treatment, this was reduced to 27% of the participants assigned to the blood pressure goal of 102 to 107 mm Hg and 11% of the participants assigned to the goal of lower than 92 mm Hg. These marked improvements in blood pressure control were achieved in a relatively uniform manner across education and income subgroups.29
The prevalence of elevated serum creatinine level was much higher among individuals treated for hypertension compared with untreated hypertensive individuals across all categories of blood pressure. These cross-sectional data should not be used to argue that antihypertensive treatment does not slow the progression of renal disease. First, despite similar current blood pressure levels in treated and untreated individuals, treated individuals are likely to have a longer duration of high blood pressure as well as a greater past severity of hypertension, both of which are risk factors for nephrosclerosis.30 Second, antihypertensive medications may decrease glomerular filtration rate by decreasing systemic and glomerular capillary pressure and thus increasing serum creatinine levels. Angiotensin-converting enzyme inhibitors, for example, often increase serum creatinine levels by up to 20% at the initiation of therapy with greater increases seen among patients with bilateral renal vascular disease31 and smaller increases in unselected patients with chronic renal disease.32- 34 Third, and most importantly, a wealth of clinical trial data supports the benefit of effective antihypertensive therapy in slowing the progression of renal disease.
Blacks have a disproportionately higher incidence of ESRD than whites overall, especially ESRD due to hypertension.35- 38 Our data show that although there are more whites (4.2 million) than blacks (1.1 million) with elevated serum creatinine levels, blacks have a greater prevalence of elevated serum creatinine level than whites (5.3% vs 2.9%). The number of cases with an elevated serum creatinine level in each group can be compared with the number of incident ESRD cases in 1991 based on the US Renal Data System.39 This ratio is 1.3 ESRD cases per 100 individuals with an elevated serum creatinine level for blacks compared with 0.8 ESRD cases per 100 individuals with an elevated serum creatinine level for whites. These data suggest that blacks are at an approximately 1.7 times higher risk of an elevated serum creatinine level progressing into ESRD compared with whites (the ratio is higher if a definition of elevated serum creatinine level with higher cutoffs for blacks than whites is used). This observation also raises the question of whether competing causes of death differ between blacks and whites among patients with elevated serum creatinine levels.
Most individuals with elevated serum creatinine levels are in older age groups. While the prevalence of elevated serum creatinine level is associated with higher blood pressure in this group, 6.1% of older individuals without hypertension have an elevated serum creatinine level. This represents 18% of all individuals with elevated serum creatinine levels and emphasizes the need for interventions other than blood pressure control to slow the progression of renal disease. It also emphasizes the limitation of screening for renal disease only among individuals with hypertension.
Use of a single cutoff to define an elevated serum creatinine level may have limited accuracy in light of age and ethnic differences in serum creatinine levels. Younger individuals, especially among men and blacks, have slightly higher serum creatinine levels owing to greater muscle mass. Adjusting for muscle mass was not possible because of limited availability of data in NHANES III. Although there is no gold standard to measure the actual specificity of our definition, it is possible to obtain a lower-limit estimate of specificity by assuming that none of the individuals with optimal blood pressure without antihypertensive treatment have renal disease. Only 0.8% of these individuals have elevated serum creatinine levels, indicating an approximate specificity of at least 99.2%. The specificity is likely to vary by race and age because the definition was sex specific, but not race or age specific. The percentages of untreated individuals with an optimal blood pressure who did not have elevated serum creatinine levels were 99.1%, 98.8%, and 99.8% among non-Hispanic whites, non-Hispanic blacks, and Mexican Americans, respectively, and 99.7%, 99.4%, and 93.7% among patients aged 17 to 39, 40 to 59, and over 60 years, respectively. Different definitions of elevated serum creatinine level would yield somewhat different estimates of the number of individuals with an elevated serum creatinine level in the population. However, the pattern of the association with blood pressure category and antihypertensive treatment was largely unaffected.
Although a stringent criterion for elevated serum creatinine level was used overall, a single measurement of serum creatinine level has limitations in measuring true renal function. First, although serum creatinine level varies inversely with the level of glomerular filtration rate, it is also affected by factors other than glomerular filtration rate. In addition to age, sex, and race, dietary intake of protein (in particular cooked meat) is an independent determinant of creatinine production. Second, increased tubular secretion of creatinine causes the rise in serum creatinine level to lag behind the decline in glomerular filtration rate. Third, numerous studies have shown that a reduced glomerular filtration rate is not a sensitive indicator of chronic renal disease. Glomerular adaptations such as increased glomerular capillary pressure and hypertrophy can increase single nephron glomerular filtration rate, thereby maintaining a normal glomerular filtration rate despite a reduction in nephron number.40 Misclassification is especially likely in the elderly because of the simultaneous age-related decline in glomerular filtration rate and muscle mass, leading to only minimal elevation in serum creatinine level despite reduced renal function. Therefore, misclassification of older individuals is especially likely.
Equations exist to estimate creatinine clearance41 and glomerular filtration rate42 based on serum creatinine level, but their application requires a number of assumptions, which are avoided by use of an empirically observed prevalence of elevated serum creatinine level. Despite the above limitations, studies have shown that serum creatinine level has a high specificity but low sensitivity for detecting chronic renal disease43 (thus indicating the high probability of correctly identifying the presence of disease), but many cases of chronic renal disease will be missed by applying this and any other diagnostic criterion based on serum creatinine level alone. Our results, therefore, should be considered an underestimate of the number of individuals with impaired renal function, especially among the elderly.
The burden of chronic renal disease extends far beyond ESRD. There are 25 cases of elevated serum creatinine level for every prevalent ESRD case and 108 cases of elevated serum creatinine level for every incident ESRD case. Our results show that chronic renal disease is strongly associated with inadequate blood pressure control. Most individuals with chronic renal disease are hypertensive and receive medications for controlling blood pressure. However, most of these patients seem to be undertreated. The largest number of individuals with an elevated serum creatinine level had stage 1 blood pressure (140-159 mm Hg systolic or 90-99 mm Hg diastolic). These national estimates suggest that there is a great potential for decreasing renal disease incidence and cardiovascular mortality by optimizing blood pressure treatment in this high-risk population.
Accepted for publication January 1, 2001.
This study was supported by grant R29-DK48362 (Dr Coresh) from the National Institutes of Health, Bethesda, Md, and was partially supported by grants 5M01RR00722 (Dr Coresh) and RR00035 from the National Center for Research Resources, Bethesda.
This work was started by G. Laura Wei as a Master of Health Science student advised by Josef Coresh, MD, PhD, at the School of Hygiene and Public Health, John Hopkins University, Baltimore, Md.
Corresponding author: Josef Coresh, MD, PhD, 2024 E Monument St, Suite 2-600, Baltimore, MD 21205 (e-mail: firstname.lastname@example.org).