Abstract
Background
An estimated 75% of the seven million Americans with moderate-to-severe chronic kidney disease are undiagnosed. Improved prediction models to identify high-risk subgroups for chronic kidney disease enhance the ability of health care providers to prevent or delay serious sequelae, including kidney failure, cardiovascular disease, and premature death.
Methods
We identified 11,955 adults ≥18 years of age in the Third National Health and Nutrition Examination Survey. Chronic kidney disease was defined as an estimated glomerular filtration rate of 15 to 59 ml/minute/1.73 m2. High-risk subgroups for chronic kidney disease were identified by estimating the individual probability using β coefficients from the model of traditional and non-traditional risk factors. To evaluate this model, we performed standard diagnostic analyses of sensitivity, specificity, positive predictive value, and negative predictive value using 5%, 10%, 15%, and 20% probability cutoff points.
Results
The estimated probability of chronic kidney disease ranged from virtually no probability (0%) for an individual with none of the 12 risk factors to very high probability (98%) for an older, non-Hispanic white edentulous former smoker, with diabetes ≥10 years, hypertension, macroalbuminuria, high cholesterol, low high-density lipoprotein, high C-reactive protein, lower income, and who was hospitalized in the past year. Evaluation of this model using an estimated 5% probability cutoff point resulted in 86% sensitivity, 85% specificity, 18% positive predictive value, and 99% negative predictive value.
Conclusion
This United States population–based study suggested the importance of considering multiple risk factors, including periodontal status, because this improves the identification of individuals at high risk for chronic kidney disease and may ultimately reduce its burden.
Keywords: Glomerular filtration rate, hypertension, jaw, edentulous, kidney diseases, periodontal diseases, sensitivity and specificity
Chronic kidney disease is a major public health problem related to serious sequelae, including end-stage kidney failure, cardiovascular disease, premature death, and increased health care expenditures.1,2 The risk for these serious sequelae increases as the glomerular filtration rate (GFR) decreases below 60 ml/minute/1.73 m2,3–5 corresponding to the National Kidney Foundation’s Kidney Disease Outcomes Quality Initiative (K/DOQI) stage 3 and stage 4 chronic kidney disease defined as GFR 15 to 59 ml/minute/1.73 m2.1 Fewer than 25% of the estimated seven million adults with chronic kidney disease reported ever being told that they had weak or failing kidneys,6 suggesting that more than 75% of adults with chronic kidney disease are undiagnosed in the United States (U.S.).
The identification of high-risk individuals provides an opportunity for early detection and intervention to prevent or delay the onset of end-stage renal disease, cardiovascular events, premature death, and increased health care expenditures. Traditional risk factors for chronic kidney disease include age >60 years,1,7–9 hypertension,2,7,9–16 diabetes, 2,7–17 poor glycemic control,1,7,12,13 obesity,2,7,13–17 macroalbuminuria,1,6,7,9,18 smoking,1,7,10,13,15–17 C-reactive protein (CRP),7,15,19 elevated total cholesterol,1,7,13,15 low levels of high-density lipoprotein (HDL) cholesterol,1,7,11,13,16 elevated levels of low-density lipoprotein cholesterol, 7,16 race/ethnicity,1,7,12,14,15,17 gender,1,7,8,10,13,15–17 and income/poverty.1,7,14 Non-traditional risk factors that may contribute to chronic kidney disease include periodontal disease,7,20–22 education,7,14,17 and access to medical care.7,14,23
The biologic plausibility for considering periodontal disease as a risk factor is derived from the potential role of the inflammatory response to periodontal disease in the chronic systemic inflammatory burden (e.g., increased CRP levels)15,19 associated with chronic kidney disease. The local tissue destructive immunoinflammatory response to periodontal pathogens, their products (i.e., lipopolysaccharides), and inflammatory cytokines are believed to contribute to the chronic systemic inflammatory burden of periodontal disease.24,25
The objective of our study was two-fold: 1) to identify high-risk subgroups by estimating an individual’s probability of chronic kidney disease by applying a multivariable model of traditional and non-traditional risk factors in a representative sample of the U.S. population using the Third National Health and Nutrition Examination Survey (NHANES III) dataset;26 and 2) to validate the model using the standard diagnostic analyses of sensitivity, specificity, positive predictive value, and negative predictive value.
MATERIALS AND METHODS
Study Population
This cross-sectional study was deemed exempt by the institutional review board. NHANES III, conducted between 1988 and 1994, is a complex, multistage, stratified, clustered sample of the civilian, non-institutionalized U.S. population, which is representative of the U.S. population. We identified 11,955 adults ≥18 years of age in NHANES III, which represented 124.3 million Americans. The study population was randomly divided into two separate and distinct samples. Sample1(n=5,978)was used to develop the multivariable logistic regression model. To evaluate thismodel, we performed the standard diagnostic analyses of sensitivity, specificity, positive predictive value, and negative predictive value using 5%, 10%, 15%, and 20% probability cutoff points in the sample 2 dataset (n = 5,977).
Description of Main Outcome
The main outcome was moderate-to-severe chronic kidney disease defined as an estimated GFR of 15 to 59 ml/minute/1.73 m2.1 This definition was reported to more precisely estimate decreased kidney function.6 Henceforth, this moderately to severely decreased kidney function will be referred to as chronic kidney disease. The GFR was estimated using the simplified Modification of Diet in Renal Disease Study equation: GFR = 186.3 × (serum creatinine in mg/dl)−1.154 × age −0.203 × 0.742 (if female) × 1.21 (if black). The serum creatinine value was calibrated by subtracting the value of 0.23 to align the NHANES measures with creatinine assays in the aforementioned equation.27
Description of Risk Factors
The model for chronic kidney disease includes the following 12 traditional and non-traditional risk factors.
Health status
Periodontal status was based on a clinical examination and categorized as no periodontal disease, periodontal disease, or edentulous. Periodontal disease was defined as having one or more sites with loss of attachment ≥4 mm and bleeding on the same tooth, where bleeding is an indicator of active inflammation.28 Edentulous was defined as having lost all natural teeth. Periodontal disease is a major cause of edentulism in adults.29 Systemic hypertension was defined as systolic pressure >140 mm Hg, or diastolic pressure >90 mm Hg, or the individual was told on two or more visits that he/she had hypertension. Diabetes was defined as fasting plasma glucose level ≥126 mg/dl, or ≥200 mg/dl after an oral glucose tolerance test, or self-reported physician diagnosed. Diabetes duration was dichotomized as ≥10 or <10 years based on the K/DOQI report that diabetes duration of 10 to 15 years is a clinical feature of stage 1 or 2 chronic kidney disease.1
Sociodemographic status
These variables included: age, dichotomized as 18 to 59 years (referent) and ≥60 years; race/ethnicity, categorized as non-Hispanic white, non-Hispanic black, and Mexican American (referent); and low income, defined as <$20,000 annual household income.
Health behavior and biomarkers
Self-reported smoking status was current (referent), former, or never, excluding those who reported former or never smoking but were current smokers based on having serum cotinine levels ≥15 ng/ml, which is considered the gold standard to detect tobacco use.30 Macroalbuminuria was defined as urinary albumin-to-creatinine excretion ratio ≥300 mg/g. High cholesterol was defined as total serum cholesterol ≥240 mg/dl. Low HDL cholesterol was ≤35 mg/dl. CRP was specified as a continuous variable in milligrams per deciliter.
Health care use
Health care use was defined as being hospitalized in the past year.
Additional risk factors that were considered in the multivariable modeling include gender, education, obesity, and having an annual physician visit.
Statistical Analyses
Tests of the hypothesis that chronic kidney disease is associated with traditional and non-traditional risk factors used univariable (Table 1) and multivariable (Table 2) logistic regression modeling with statistical significance reported as a 95% confidence interval (CI). We speculated that certain high-risk subgroups could be identified through the application of the most parsimonious model derived from sample 1 data to estimate an individual’s probability of chronic kidney disease. Analyses were conducted on sample 2 data using statistical software packages‡,§ to account for complex survey design and sample weights and to produce national estimates. 31 The estimated probability (π[x]) that an individual with specific risk factors (covariates in multiple logistic regression model)32 will have chronic kidney disease was calculated using the formula π(x) = eβ0+β1X1+…+β12X12/(1 + eβ0+β1X1+…+β12X12), where β0 is the intercept, β1 is the regression coefficient for the first independent variable (x1), β2 is the regression coefficient for the second independent variable (x2), and so forth for each of the 12 independent variables in the final model. The β coefficients derived in the multivariable logistic regression model were used to estimate the probability that individuals with specific characteristics had chronic kidney disease. To demonstrate the application of this model, we report the two extreme scenarios representing the very lowest risk individual with no risk factors and the very highest risk individual with all risk factors (Table 2), along with the successive addition of one to 12 risk factors (Figs. 1 and 2) based on the strength of their association. When the risk factor hadmore than two levels, the risk category with the largest coefficient (i.e., highest odds ratio [OR]) was incorporated in the cumulative risk factor estimation of the proportion of individuals with chronic kidney disease.
Table 1.
Descriptive Summary and Association Between Risk Factors and Chronic Kidney Disease
| Risk Factor (n [%]) | No Chronic Kidney Disease (96.3%) | Chronic Kidney Disease (3.7%) | ORCrude (95% CI) |
|---|---|---|---|
| Socioeconomic status | |||
| Age (years) | |||
| 18 to 59 (4,396 [80.3]) | 99.3% | 0.7% | 1.00 |
| ≥60 (1,582 [19.7]) | 84.3% | 15.7% | 26.79 (16.47 to 43.60)* |
| Race/ethnicity | |||
| Non-Hispanic white (2,446 [82.2]) | 96.0% | 4.0% | 4.60 (2.91 to 7.29)* |
| Non-Hispanic black (1,678 [11.5]) | 97.5% | 2.5% | 2.87 (1.67 to 4.93)* |
| Mexican American (1,854 [6.3]) | 99.1% | 0.9% | 1.00 |
| Gender | |||
| Female (3,123 [50.2]) | 95.5% | 4.5% | 1.64 (1.17 to 2.32)* |
| Male (2,855 [49.8]) | 97.2% | 2.8% | 1.00 |
| Lower income | |||
| Yes (2,808 [31.8]) | 93.4% | 6.6% | 3.05 (2.24 to 4.14)* |
| No (3,170 [68.2]) | 97.7% | 2.3% | 1.00 |
| High school graduate | |||
| Yes (3,649 [78.1]) | 97.3% | 2.7% | 1.00 |
| No (2,299 [21.9]) | 92.8% | 7.2% | 2.84 (2.06 to 3.91)* |
| Health status and health behavior | |||
| Periodontal status | |||
| Edentulous (787 [11.6]) | 84.3% | 15.7% | 10.87 (6.86 to 17.20)* |
| Periodontal disease (589 [6.1]) | 92.8% | 7.2% | 4.50 (3.02 to 6.71)* |
| No periodontal disease (4,602 [82.3]) | 98.3% | 1.7% | 1.00 |
| Diabetes duration | |||
| ≥10 years (167 [1.6]) | 77.0% | 23.0% | 8.65 (5.09 to 14.70)* |
| <10 years (5,811 [98.4]) | 96.7% | 3.3% | 1.00 |
| Hypertension | |||
| Yes (1,742 [25.0]) | 89.3% | 10.7% | 8.95 (6.31 to 12.70)* |
| No (4,236 [75.0]) | 98.7% | 1.3% | 1.00 |
| Macroalbuminuria | |||
| Yes (108 [1.1]) | 80.1% | 19.9% | 6.91 (3.48 to 13.72)* |
| No (5,870 [98.9]) | 96.5% | 3.5% | 1.00 |
| Obesity | |||
| Yes (2,557 [36.9]) | 95.4% | 4.6% | 1.50 (1.09 to 2.08)* |
| No (3,412 [63.1]) | 96.9% | 3.1% | 1.00 |
| High cholesterol | |||
| Yes (1,098 [17.9]) | 91.2% | 8.8% | 3.74 (2.67 to 5.24)* |
| No (4,880 [82.1]) | 97.5% | 2.5% | 1.00 |
| Low HDL | |||
| Yes (709 [12.8]) | 94.5% | 5.5% | 1.68 (1.13 to 2.49)* |
| No (5,269 [87.2]) | 96.6% | 3.4% | 1.00 |
| Smoking status | |||
| Never (3,047 [45.8]) | 96.4% | 3.6% | 2.86 (1.69 to 4.83)* |
| Former (1,300 [23.3]) | 93.1% | 6.9% | 5.68 (3.55 to 9.09)* |
| Current (1,631 [30.9]) | 98.7% | 1.3% | 1.00 |
| Hospitalized in past year | |||
| Yes (773 [10.7]) | 91.8% | 8.2% | 2.79 (1.92 to 4.05)* |
| No (5,205 [89.3]) | 96.9% | 3.1% | 1.00 |
| Annual physician visit | |||
| Yes (4,736 [80.1]) | 95.8% | 4.2% | 2.53 (1.40 to 4.56)* |
| No (1,242 [19.9]) | 98.3% | 1.7% | 1.00 |
| CRP (mg/dl) | NA | NA | 1.40 (1.23 to 1.59)* |
Unweighted number with weighted percent. Excluded those who reported never or former smoking with serum cotinine level indicating current smoker. Inclusion criterion: periodontal examination or edentulous.
ORCrude = Unadjusted odds ratio for the association between chronic kidney disease and the suspected/recognized risk factors in sample 1; NA= not applicable.
P <0.05.
Table 2.
Multivariable Logistic Regression Model of Independent Risk Factors for Chronic Kidney Disease in Descending Order of β Coefficients, Low-Risk and High-Risk Subgroups
| Risk Factor | Final Model ORAdj (95% CI) | Final Model β Coefficient (intercept = −7.39) | Low Risk (0.06%) | High Risk (98.3%) |
|---|---|---|---|---|
| Age ≥60 years | 9.03 (5.52 to 14.77)* | 2.46 | Age <60 years | Age ≥60 years |
| Macroalbuminuria | 3.41 (1.82 to 6.41)* | 1.93 | No macroalbuminuria | Macroalbuminuria |
| Race/ethnicity | ||||
| Non-Hispanic white | 3.08 (2.00 to 4.75)* | 1.63 | Mexican American | Non-Hispanic white |
| Non-Hispanic black | 2.36 (1.33 to 4.21)* | 0.99 | ||
| CRP (mg/dl) | 1.19 (1.01 to 1.41)* | 0.12 | 0.21 | 12.80 |
| Hypertension | 2.21 (1.61 to 3.03)* | 0.76 | No hypertension | Hypertension |
| Diabetes duration | 1.52 (0.93 to 2.49) | 0.69 | Diabetes <10 years | Diabetes ≥10 years |
| Low income | 1.57 (1.09 to 2.25)* | 0.54 | Not low income | Low income |
| Hospitalized in past year | 1.90 (1.26 to 2.87)* | 0.46 | Not hospitalized in past year | Hospitalized in past year |
| Periodontal status | ||||
| Edentulous | 2.03 (1.31 to 3.14)* | 0.44 | No periodontal disease | Edentulous |
| Periodontal disease | 1.60 (1.07 to 2.39)* | 0.38 | ||
| Low HDL cholesterol | 1.69 (1.07 to 2.65)* | 0.43 | Not low HDL cholesterol | Low HDL cholesterol |
| Smoking status | ||||
| Former smoker | 2.47 (1.52 to 4.02)* | 0.40 | Current smoker | Former smoker |
| Never smoker | 2.16 (1.27 to 3.66)* | 0.32 | ||
| High cholesterol | 1.93 (1.33 to 2.79)* | 0.16 | Not high cholesterol | High cholesterol |
Excluded those who reported never or former smoking with serum cotinine level indicating a current smoker.
ORAdj = OR for the association between chronic kidney disease, simultaneously taking into account all the listed potential or recognized risk factors.
P <0.05.
Figure 1.
Proportion of individuals with chronic kidney disease. Cumulative risk factors lowest to highest.
Figure 2.
Proportion of individuals with chronic kidney disease. Cumulative risk factors highest to lowest.
Furthermore, to evaluate this prediction model, we performed the standard diagnostic analyses of sensitivity, specificity, positive predictive value, and negative predictive value using 5%, 10%, 15%, and 20% probability cutoff points. Sensitivity and positive predictive values were calculated for those with a probability of having chronic kidney disease greater than or equal to the cutoff point. Specificity and negative predictive values were calculated for those with a probability of having chronic kidney disease less than the cutoff point.
RESULTS
Table 1 presents important characteristics of the study population and the univariable logistic regression models (ORCrude) based on sample 1 data. Overall, 3.7% of adults had chronic kidney disease. The proportion of adults with chronic kidney disease varied according to risk factors. For example, adults ≥60 years of age (15.7%) were 27 times (ORCrude = 26.79; 95% CI: 16.47 to 43.60) more likely to have chronic kidney disease than younger adults; edentulous adults (15.7%) were 11 times (ORCrude = 10.87; 95% CI: 6.86 to 17.20) more likely to have chronic kidney disease, and adults with periodontal disease (7.2%) were 4.5 times (ORCrude = 4.50; 95% CI: 3.02 to 6.71) more likely to have chronic kidney disease than adults who did not have periodontal disease.
Table 2 (column 2) presents the most parsimonious final model of traditional and non-traditional risk factors for chronic kidney disease based on sample 1 data. Next, this most parsimonious final model was applied to sample 2 data using the resulting β coefficients reported in column 3 of Table 2. Thus, we report the model’s estimate of the probability of chronic kidney disease for the lowest risk group (column 4: individuals with none of the model’s risk factors [probability = 0.06%]) and for the highest risk group(column 5: individuals with all of the model’s risk factors [probability= 98.3%]). This highest risk group was older (≥60 years), non-Hispanic white edentulous former smokers, with diabetes ≥10 years, hypertension, macroalbuminuria, high cholesterol, low HDL cholesterol, CRP= 12.80mg/dl, lower income, and hospitalization in the past year.
Twenty-two subgroups are displayed in Figures 1 and 2, ranging from having no risk factors to having all 12 risk factors based on the β coefficients listed in Table 2, along with the intercept (−7.39). Figure 1 depicts the addition of one risk factor at a time as the cumulative risk factors, lowest to highest, beginning with the risk factor with the lowest β coefficient, i.e., high cholesterol (the last risk factor listed in Table 2), then the next lowest β coefficient, i.e., former smoker, and so forth. Figure 2 depicts the addition of one risk factor at a time as the cumulative risk factors, highest to lowest, beginning with the risk factor with the highest β coefficient, namely age ≥60 years, then the risk factor with the next highest β coefficient, i.e., macroalbuminuria, and so forth in the order listed in Table 2.
In Figure 1, the high-risk subgroup includes those with risk factors having the highest β coefficients, as indicated by the steep slope when an individual had one of the four strongest risk factors in addition to the other risk factors in the lowest-to-highest scenario. In the lowest-to-highest scenario, an estimated 3% of individuals with the eight weakest risk factors, but without the four strongest risk factors, have chronic kidney disease, i.e., edentulous, former smokers with diabetes ≥10 years, hypertension, low HDL cholesterol, high total cholesterol, low income, and hospitalization in the past year. In Figure 2, the highest-to-lowest scenario depicts that an estimated 21% of individuals with the three strongest risk factors, but not any of the other nine risk factors, have chronic kidney disease, i.e., an older (≥60 years) non-Hispanic white adult with macroalbuminuria. The estimated probability of chronic kidney disease increased 33% for high CRP, 18% for hypertension, 12% for having diabetes ≥10 years, 6% for low income, 4% for being hospitalized in the past year, 2% for edentulism or periodontal disease, 1% for low HDL cholesterol, 1% for former smoking, and 0.3% for high cholesterol (Fig. 2).
Evaluation of this prediction model using the standard diagnostic analysis 33 of sample 2 is shown in Table 3, depicting the changing sensitivity, specificity, positive predictive value, and negative predictive value with increasing probability of having chronic kidney disease. As the cutoff point was changed from 5% to 10% to 15% to 20%, the sensitivity decreased, the specificity and positive predictive value increased, and there was little change in the high negative predictive value. Sensitivity ranged from 46% to 86%, specificity was 85% to 96%, positive predictive value was 18% to 32%, and negative predictive value was 98% to 99%. Evaluation of this model using a 5% probability cutoff point resulted in 86% sensitivity, 85% specificity, 18% positive predictive value, and 99% negative predictive value.
Table 3.
Evaluation of Prediction Model for Chronic Kidney Disease, U.S. Adults
| Cutoff Point (probability) | Sensitivity (%)* | Specificity (%)† | Positive Predictive Value (%)* | Negative Predictive Value (%)† |
|---|---|---|---|---|
| 5% | 86 | 85 | 18 | 99 |
| 10% | 68 | 90 | 20 | 99 |
| 15% | 58 | 94 | 26 | 98 |
| 20% | 46 | 96 | 32 | 98 |
Sensitivity and positive predictive value were calculated for those with probability greater than or equal to the cutoff point.
Specificity and negative predictive value were calculated for those with probability less than the cutoff point.
DISCUSSION
Our population-based study estimated the probability of chronic kidney disease using the β coefficients in the most parsimonious final model with 12 recognized and suspected risk factors for chronic kidney disease. Subgroups were identified ranging from virtually no probability (0.06%) to very high probability (98%). Any individual’s risk can be estimated as the probability that individuals with the same 12 specific characteristics have chronic kidney disease using the formula in the Statistical Analyses section and the β coefficients reported in Table 2.
The impact of the strongest risk factors on the estimated probability of chronic kidney disease is evident by the steep slope in Figure 1 when the four strongest risk factors were added in the lowest-to-highest scenario. For those with the eight lowest risk factors, the estimated probability of chronic kidney disease was 3%, the addition of high CRP quadrupled the estimated probability to 12%. The addition of the next risk factor, race/ethnicity, more than tripled the estimated probability of chronic kidney disease to 41% for non-Hispanic white adults with the same nine risk factors. Next, having macroalbuminuria doubled the estimated probability of chronic kidney disease to 83%. Finally, when adding the risk factor age ≥60 years, the estimated probability of chronic kidney disease increased to 98% (Table 2, column 5; Fig. 1).
Conversely, if an individual’s only risk factor was the strongest risk factor, age ≥60 years, the estimated probability of chronic kidney disease was only 0.7% (Fig. 2). The estimated probability of chronic kidney disease was 5% for adults with the two strongest risk factors, age ≥60 years and macroalbuminuria. The addition of the next strongest risk factor, non-Hispanic white race/ethnicity, quadrupled the estimated probability to 21%. The addition of high CRP more than doubled the estimated probability of chronic kidney disease to 54%. The addition of having hypertension increased the probability of chronic kidney disease by 18% to 72%. The addition of each of the remaining seven risk factors resulted in a smaller increase in the probability than the previous risk factor (except for being a former smoker), such that the addition of having diabetes ≥10 years, having low income, being hospitalized in the past year, being edentulous or having periodontal disease, having low HDL cholesterol, being a former smoker, and having high cholesterol increased the probability of having chronic kidney disease by 12%, 6%, 4%, 2%, 1%, 1%, and 0.3%, respectively, to 98.3% (Fig. 2).
Evaluation of our final model using a cutoff point of 5% probability for chronic kidney disease resulted in 86% sensitivity, 85% specificity, 18% positive predictive value, and 99% negative predictive value. The diagnostic characteristics of our model (age, macroalbuminuria, race/ethnicity, CRP, hypertension, diabetes duration, low income, being hospitalized in the past year, periodontal status, HDL cholesterol, smoking status, and cholesterol) are similar to the 92% sensitivity, 68% specificity, 18% positive predictive value, and 99% negative predictive value reported for a model with age, gender, hypertension, proteinuria, anemia, diabetes, peripheral vascular disease, history of cardiovascular disease, and congestive heart failure.8 Including periodontal status improved the fit of our model compared to the model without periodontal status, based on the Satterthwaite-adjusted F statistic P value = 0.0031.
A limitation of this study is that it was cross-sectional; thus, the temporal association is unknown. A major strength of our study is that the most parsimonious multivariable model was developed using a random half-sample (sample 1) of the U.S. population–based dataset, which was separate and distinct from the other random half-sample (sample 2) that was used for validation. The screening cutoff point of 5% probability had very good sensitivity, specificity, and negative predictive value. The sensitivity indicates an estimated 86% of adults with chronic kidney disease would be identified using this model, and the specificity indicates that 85% of adults without chronic kidney disease would be identified as not having chronic kidney disease. However, the positive predictive value indicates that only 18% of those identified by the model are estimated to have chronic kidney disease. Our findings exemplify the general epidemiologic premise that the rarer or lower the disease prevalence, the better the negative predictive value, whereas better positive predictive values are found for more prevalent conditions.33 This is evident in Table 3; as the cutoff point increased (higher probability of having chronic kidney disease) the positive predictive value increased. An additional strength of this study is that these findings can be generalized to the U.S. population because the NHANES III sampling methodology was designed to represent the U.S. population.26
CONCLUSIONS
This U.S. population–based study suggested the importance of considering multiple risk factors, including periodontal status, because this improves the identification of individuals at high risk for chronic kidney disease and may ultimately reduce its burden. Further research is needed to simultaneously assess the role of multiple risk factors and to validate this model in other populations.
Acknowledgments
This research was supported by a grant from the National Institutes of Health (NIH)/National Institute of Dental and Craniofacial Research, Bethesda, Maryland (NIH DE016031). This research was published in abstract form as part of the 37th Annual Meeting of the American Association for Dental Research, April 3, 2008, in Dallas, Texas. The authors report no conflicts of interest related to this study.
Footnotes
SUDAAN release 9.0.1, Software for the Statistical Analysis of Correlated Data, Research Triangle Institute, Research Triangle Park, NC.
SAS Systems for Windows, version 9.1, SAS Institute, Cary, NC.
References
- 1.National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Am J Kidney Dis. 2002;39:S1–S266. [PubMed] [Google Scholar]
- 2.Centers for Disease Control and Prevention. Prevalence of chronic kidney disease and associated risk factors – United States, 1999–2004. Morb Mortal Wkly Rep. 2007;56:161–165. [PubMed] [Google Scholar]
- 3.Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic kidney disease and the risks of death, cardiovascular events and hospitalization. N Engl J Med. 2004;351:1296–1305. doi: 10.1056/NEJMoa041031. [DOI] [PubMed] [Google Scholar]
- 4.Keith DS, Nichols GA, Gullion CM, Brown JB, Smith DH. Longitudinal follow-up and outcomes among a population with chronic kidney disease in a large managed care organization. Arch Intern Med. 2004;164:659–663. doi: 10.1001/archinte.164.6.659. [DOI] [PubMed] [Google Scholar]
- 5.Rahman M, Pressel S, Davis BR, et al. Cardiovascular outcomes in high-risk hypertensive patients stratified by baseline glomerular filtration rate. Ann Intern Med. 2006;144:172–180. doi: 10.7326/0003-4819-144-3-200602070-00005. [DOI] [PubMed] [Google Scholar]
- 6.Coresh J, Byrd-Hold D, Astor BC, et al. Chronic kidney disease awareness, prevalence, and trends among U.S. adults, 1999 to 2000. J Am Soc Nephrol. 2005;16:180–188. doi: 10.1681/ASN.2004070539. [DOI] [PubMed] [Google Scholar]
- 7.Fisher MA, Taylor GW, Shelton BJ, et al. Periodontal disease and other nontraditional risk factors for CKD. Am J Kidney Dis. 2008;51:45–52. doi: 10.1053/j.ajkd.2007.09.018. [DOI] [PubMed] [Google Scholar]
- 8.Bang H, Vupputuri S, Shoham DA, et al. SCreening for Occult REnal Disease (SCORED): A simple prediction model for chronic kidney disease. Arch Intern Med. 2007;167:374–381. doi: 10.1001/archinte.167.4.374. [DOI] [PubMed] [Google Scholar]
- 9.Stevens LA, Coresh J, Green T, Levey AS. Assessing kidney function – Measured and estimated glomerular filtration rate. N Engl J Med. 2006;354:2473–2483. doi: 10.1056/NEJMra054415. [DOI] [PubMed] [Google Scholar]
- 10.Haroun MK, Jaar BG, Hoffman SC, Comstock GW, Klag MJ, Coresh J. Risk factors for chronic kidney disease: A prospective study of 23,534 men and women in Washington County, Maryland. J Am Soc Nephrol. 2003;14:2934–2941. doi: 10.1097/01.asn.0000095249.99803.85. [DOI] [PubMed] [Google Scholar]
- 11.Sarnak MJ, Levy AS, Schoolwerth AC, et al. 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. doi: 10.1161/01.HYP.0000102971.85504.7c. [DOI] [PubMed] [Google Scholar]
- 12.Jamerson KA. Preventing chronic kidney disease in special populations. Am J Hypertens. 2005;18:106S–111S. doi: 10.1016/j.amjhyper.2004.11.018. [DOI] [PubMed] [Google Scholar]
- 13.Fox CS, Larson MG, Leip EP, Meigs JB, Wilson PWF, Levy D. Glycemic status and development of kidney disease: The Framingham Heart Study. Diabetes Care. 2005;28:2436–2440. doi: 10.2337/diacare.28.10.2436. [DOI] [PubMed] [Google Scholar]
- 14.Norris KC, Agodoa LY. Unraveling the racial disparities associated with kidney disease. Kidney Int. 2005;68:914–924. doi: 10.1111/j.1523-1755.2005.00485.x. [DOI] [PubMed] [Google Scholar]
- 15.Rule AD, Jacobsen SJ, Schwartz GL, et al. A comparison of serum creatinine-based methods for identifying chronic kidney disease in hypertensive individuals and their siblings. Am J Hypertens. 2006;19:608–614. doi: 10.1016/j.amjhyper.2005.10.025. [DOI] [PubMed] [Google Scholar]
- 16.Parikh NI, Hwang SJ, Larson MG, Meigs JB, Levy D, Fox CS. Cardiovascular disease risk factors in chronic kidney disease: Overall burden and rates of treatment and control. Arch Intern Med. 2006;166:1884–1891. doi: 10.1001/archinte.166.17.1884. [DOI] [PubMed] [Google Scholar]
- 17.Vupputuri S, Sandler DP. Lifestyle risk factors and chronic kidney disease. Ann Epidemiol. 2003;13:712–720. doi: 10.1016/s1047-2797(03)00066-8. [DOI] [PubMed] [Google Scholar]
- 18.Cohn JN, Quyyumi AA, Hollenberg NK, Jamerson KA. Surrogate markers for cardiovascular disease: Functional markers. Circulation. 2004;109(25 Suppl 1):IV31–IV46. doi: 10.1161/01.CIR.0000133442.99186.39. [DOI] [PubMed] [Google Scholar]
- 19.Tonelli M, Sacks F, Pfeffer M, Jhangri GS, Curhan G. Biomarkers of inflammation and progression of chronic kidney disease. Kidney Int. 2005;68:237–245. doi: 10.1111/j.1523-1755.2005.00398.x. [DOI] [PubMed] [Google Scholar]
- 20.Shultis WA, Weil EJ, Looker HC, et al. Effect of periodontitis on overt nephropathy and end-stage renal disease in type 2 diabetes. Diabetes Care. 2007;30:306–311. doi: 10.2337/dc06-1184. [DOI] [PubMed] [Google Scholar]
- 21.Saremi A, Nelson RG, Tulloch-Reid M, et al. Periodontal disease and mortality in type 2 diabetes. Diabetes Care. 2005;28:27–32. doi: 10.2337/diacare.28.1.27. [DOI] [PubMed] [Google Scholar]
- 22.Kshirsagar AV, Moss KL, Elter JR, Beck JD, Offenbacher S, Falk RJ. Periodontal disease is associated with renal insufficiency in the Atherosclerosis Risk in Communities (ARIC) study. Am J Kidney Dis. 2005;45:650–657. doi: 10.1053/j.ajkd.2004.12.009. [DOI] [PubMed] [Google Scholar]
- 23.Feldman HI, Appel LJ, Chertow GM, et al. The chronic renal insufficiency cohort (CRIC) study: Design and method. J Am Soc Nephrol. 2003;14:S148–S153. doi: 10.1097/01.asn.0000070149.78399.ce. [DOI] [PubMed] [Google Scholar]
- 24.Beck JD, Eke P, Heiss G, et al. Periodontal disease and coronary heart disease – A reappraisal of the exposure. Circulation. 2005;112:19–24. doi: 10.1161/CIRCULATIONAHA.104.511998. [DOI] [PubMed] [Google Scholar]
- 25.Beck JD, Eke P, Lin D, et al. Association between IgG antibody to oral organisms and carotid intima-medial thickness in community dwelling adults. Atherosclerosis. 2005;183:342–348. doi: 10.1016/j.atherosclerosis.2005.03.017. [DOI] [PubMed] [Google Scholar]
- 26.National Center for Health Statistics. [Accessed April 25, 2008];National Health and Nutrition Examination Survey III data files, documentation, and SAS code. Available at http://www.cdc.gov/nchs/about/major/nhanes/nh3data.htm.
- 27.Coresh J, Astor BC, McQuillan G, et al. Calibration and random variation of the serum creatinine assay as critical elements of using equations to estimate glomerular filtration rate. Am J Kidney Dis. 2002;39:920–929. doi: 10.1053/ajkd.2002.32765. [DOI] [PubMed] [Google Scholar]
- 28.Fisher MA, Taylor GW, Tilashalski KR. Smokeless tobacco and severe active periodontal disease, NHANES III. J Dent Res. 2005;84:705–710. doi: 10.1177/154405910508400804. [DOI] [PubMed] [Google Scholar]
- 29.Phipps KR, Stevens VJ. Relative contribution of caries and periodontal disease in adult tooth loss for an HMO dental population. J Public Health Dent. 1995;55:250–252. doi: 10.1111/j.1752-7325.1995.tb02377.x. [DOI] [PubMed] [Google Scholar]
- 30.Pirkle JL, Flegal KM, Bernert JT, Brody DJ, Etzel RA, Maurer KR. Exposure of the U.S. population to environmental tobacco smoke: The Third National Health and Nutrition Examination Survey. 1988 to 1991. JAMA. 1996;275:1233–1240. [PubMed] [Google Scholar]
- 31.Research Triangle Institute. SUDAAN Language Manual, release 9.0. Research Triangle Park, NC: Research Triangle Institute; 2004. [Google Scholar]
- 32.Hosmer DW, Lemeshow S. Applied Logistic Regression. 2. New York: John Wiley and Sons; 2000. pp. 85–88. [Google Scholar]
- 33.Hennekens CH, Buring JE. Epidemiology in Medicine. Philadelphia: Lippincott Williams & Wilkins; 1987. p. 337. [Google Scholar]