Abstract
Background
Many traditional cardiovascular risk factors do not predict survival to very old age. Studies have shown associations of estimated glomerular filtration rate (eGFR) and N-terminal pro-B-type natriuretic peptide (NT-pro-BNP) with cardiovascular disease and mortality in older populations. This study aimed to evaluate the associations of the level and change in eGFR and NT-pro-BNP with longevity to age 90 years.
Method
The population included participants (n = 3,645) in the Cardiovascular Health Study, aged between 67 and 75 at baseline. The main exposures were eGFR, calculated with the Berlin Initiative Study (BIS) 2 equation, and NT-pro-BNP, and the main outcome was survival to age 90. Mixed models were used to estimate level and change of the main exposures.
Results
There was an association between baseline level and change of both eGFR and NT-pro-BNP and survival to 90, and this association persisted after adjustment for covariates. Each 10 mL/min/1.73 m2 higher eGFR level was associated with an adjusted odds ratio (OR) of 1.23 (95% CI: 1.13, 1.34) of survival to 90, and a 0.5 mL/min/1.73 m2 slower decline in eGFR was associated with an OR of 1.51 (95% CI: 1.31, 1.74). A twofold higher level of NT-pro-BNP level had an adjusted OR of 0.67 (95% CI: 0.61, 0.73), and a 1.05-fold increase per year in NT-pro-BNP had an OR of 0.53 (95% CI: 0.43, 0.65) for survival to age 90.
Conclusion
eGFR and NT-pro-BNP appear to be important risk factors for longevity to age 90.
Keywords: Longevity, eGFR, NT-pro-BNP
The factors that portend longevity, defined here as survival to 90 years, remain uncertain. The associations of many traditional risk factors, such as systolic blood pressure, obesity, and lipids, with mortality are diminished in very old age (1–4). Thus, the risk factors for cardiovascular disease and survival in middle-aged and older adults cannot be equated with risk factors for longevity. A previous study in the Cardiovascular Health Study (CHS) found that conditions that are affected by traditional risk factors and reflect vascular damage, kidney function and myocardial stretch (measured by cystatin C and N-terminal pro-B-type natriuretic peptide [NT-pro-BNP]), remain strongly associated with cardiovascular outcomes among persons aged ≥ 85 years (4). We propose that primary risk factors are integrated by target organ damage, which is reflected by cystatin C and NT-pro-BNP, as the kidney and heart are such vascular organs. These target organs appear to capture the lifetime exposure to risk factors better than the risk factors themselves in older adults (4). Therefore, we hypothesize that measures that reflect cumulative vascular damage may be important predictors of longevity.
Several studies have demonstrated the associations of kidney function and NT-pro-BNP with cardiovascular disease and mortality (5–10); however, the studies that evaluate survival to very old age are scarce. Investigators using the cohort from a previous CHS study found an association between cystatin C and mortality among octogenarians (11). Another study from CHS demonstrated an association between rapid estimated glomerular filtration rate (eGFR) decline (>3 mL/min/1.73 m2/year) and increased risk of cardiovascular and all-cause mortality in older adults (12). The Leiden 85-plus study followed adults 85 years and older for 5 years and found that higher NT-pro-BNP levels were associated with accelerated cardiovascular morbidity and mortality (13).
In this study, we evaluated the associations of eGFR and NT-pro-BNP as predictors of survival to age 90. We conducted this investigation in the CHS, a cohort study of adults aged 65 and older, among whom nearly a third reached age 90 or older. We evaluated associations of both the level and change in eGFR and NT-pro-BNP with longevity. As a secondary goal, we investigated whether there was an interaction between two risk factors levels and the age at which the risk factor was measured. This evaluated whether the associations between age and survival to 90 attenuate when the risk factor is measured in older age, as this has been observed for traditional risk factors. The goal of our study is to identify factors associated with survival to 90 to enhance our understanding of the pathophysiologic mechanisms that promote or inhibit longevity.
Method
Study Population
The CHS was designed as an observational study to evaluate risk factors for cardiovascular health in adults aged 65 years or older at baseline. The study recruited persons from Medicare eligibility lists in Forsyth County, North Carolina; Sacramento County, California; Washington County, Maryland; and Pittsburgh, Pennsylvania in 1989–1990. An additional cohort of black participants was recruited during a supplemental enrollment process of CHS during 1992–1993; they constitute 15% of CHS participants. Patients underwent yearly extensive clinical examinations and were contacted biannually via telephone to ascertain health status and hospitalizations. The participants included in this study were 67–75 years at baseline (n = 3,645). We selected this age group because analyses from a previous study from CHS found that this was the widest age range one could use without birth cohort effects confounding the results (14). Participants with active treated cancer were excluded from the study at baseline.
Outcome
The outcome of this study was survival to age 90; participants were censored at July 16, 2015. People who died after 90 and people that were alive and at least 90 years at the end of the study were defined as surviving to age 90. People that died before reaching age 90 were defined as not surviving to 90 and people that were alive but had not turned 90 at the end of the study were excluded (n = 32). Deaths were identified by a review of obituaries, medical records, death certificates, and the Centers for Medicare and Medicaid Services files (15,16).
Main exposures
We used measurements of creatinine and cystatin C at baseline, and years 3, 7, and 16 to calculate eGFR. Serum creatinine was measured immediately after samples were drawn using a colorimetric method (Ektachem 700, Eastman Kodak, Rochester, NY), calibrated to isotope dilution mass spectrometry. The intra-assay coefficient of variation was 1.9%. Cystatin C was measured from frozen serum samples at −70°C, using a BN II nephelometer (N Latex cystatin C, Dade Behring, Munich, Germany) using a particle-enhanced immunonephelometric assay. The intra-assay coefficient of variation for cystatin C ranged from 2.0% to 2.8%. We calculated eGFR using the Berlin Initiative Study (BIS) 2 equation that includes creatinine, cystatin C, sex, and age (see Supplementary Methods 1). The BIS2 equation was developed and validated with gold standard measurements of GFR in a population of adults over the age of 70. We selected this equation because it has been validated to be a more accurate estimator of GFR in older adults when compared to the Cockcroft-Gault, Modification of Diet in Renal Disease (MDRD), and CKD-EPI equations (17–19).
NT-pro-BNP was measured in serum with the Elecsys 2010 system (Roche Diagnostics, Indianapolis, IN), and we used measurements at baseline, and years 3, 5, and 7. We log-transformed NT-pro-BNP for analysis because it was right-skewed.
Other variables
We used baseline measurements of age, sex, race, smoking status, body mass index, systolic blood pressure, diastolic blood pressure, low-density lipoprotein, high-density lipoprotein, C-reactive protein, hypertension, diabetes, use of antihypertensive medication, and lipid-lowering medication. Age, sex, race, and smoking status were obtained by self-report. Race was coded by self-report as white, black, or other (<1% of the participants identified with another race). Body mass index was calculated as weight (in kilograms) divided by height (in meters) squared. Blood pressure was measured in seated participants after sitting for 5 minutes. Trained study personnel obtained three measurements and the average of the last two was recorded. High-density lipoprotein was measured in fasting blood samples and low-density lipoprotein was calculated according to the Friedewald equation (20). Diabetes was defined as a fasting glucose ≥ 126 mg/dL or use of insulin or hypoglycemic medications. CHS participants were classified at baseline as having a history of myocardial infarction, stroke, or heart failure using a combination of hospital records and physician confirmation (21).
Statistical Analysis
Descriptive statistics were tabulated by tertiles of eGFR and NT-pro-BNP. Continuous variables were presented in means and SD, and categorical variables were presented using frequency and percentages.
We used mixed models to estimate level and change of eGFR and NT-pro-BNP. Next, we added the fixed effects to each of the random effects to calculate intraindividual level and change of eGFR and NT-pro-BNP for each person. We evaluated changes in the level and change by age category, using box plots. For the box plots, we winsorized NT-pro-BNP at the 99th percentile since it is highly right-skewed. We then ran a logistic regression to evaluate the association of level and change in the primary exposures and survival to age 90. In order to provide results that are clinically interpretable, we modeled eGFR level per 10 mL/min/1.73 m2 and change per 0.5 mL/min/1.73 m2 slower decline per year. Since NT-pro-BNP increases logarithmically, we modeled level per twofold higher level (0.69 ln(pg/mL)) and change per 5%-fold (0.049 ln(pg/mL)/year) faster increase. We selected covariates that were associated with the outcome and were considered potential confounders. Both models evaluating associations of level and change of eGFR and NT-pro-BNP with survival to 90 were adjusted for age, sex, race, education level, diabetes, smoking status, body mass index, alcohol consumption, antihypertensive medication use, low-density lipoprotein, high-density lipoprotein, lipids, C-reactive protein, and history of myocardial infarction, stroke, heart failure, and nonmelanoma cancer diagnosis at baseline. A third model included the concentrations and changes in both eGFR and NT-pro-BNP and was adjusted for the same confounders mentioned above. We also tested for interactions with sex, race, and blood pressure in all models, and an interaction between level of eGFR and NT-pro-BNP in the third set of models.
Next, we used generalized estimating equation models to evaluate population-level survival until age 90. The goal of these models was to evaluate if age at the time of risk factor measurement altered the association between the risk factors and survival to 90. We used all available measures of eGFR and NT-pro-BNP as continuous predictors and included the age at the time of the measurement as an interaction term with the biomarker measurement. We adjusted for the same set of variables selected for the logistic regression.
All statistical analyses and graphical displays were performed using STATA/IC 14.0 (StataCorp, College Station, TX).
Results
Of the 5,888 CHS participants, 3,613 who were aged between 67 and 75 at baseline were included in the study, and 1,160 (32%) reached age 90 or older. Available measures enabled us to calculate eGFR for 3,293 (90%) of the included participants at baseline, and there were 9,092 available measurements across follow-up. The mean eGFRs in each tertile were 53, 67, and 81 mL/min/1.73 m2 (Table 1). Participants in the lowest eGFR tertile were older, had higher body mass index, higher systolic blood pressure, more frequent hypertension, and use of antihypertensive medications. There were 2,827 (78%) measurements of NT-pro-BNP at baseline and 7,804 measurements in total. The mean values within NT-pro-BNP tertiles at baseline were 36, 101, and 559 pg/mL (Table 2). Participants with higher NT-pro-BNP had higher systolic blood pressure, and more often had hypertension and used antihypertensive medication.
Table 1.
Descriptive Characteristics of eGFR at Baseline Divided as Tertiles
| Characteristics | BIS2 eGFR Tertiles mL/min/1.73 m2 (range), n | ||
|---|---|---|---|
| First Tertile | Second Tertile | Third Tertile | |
| (72–116.2) | (61.6–71.9) | (8.4–61.7) | |
| 1,097 | 1,098 | 1,098 | |
| Cystatin C (mg/L) | 0.8 ± 0.1 | 1.0 ± 0.1 | 1.3 ± 0.4 |
| Creatinine | 0.7 ± 0.2 | 0.9 ± 0.2 | 1.2 ± 0.6 |
| Age | 69.9 ± 2.3 | 70.5 ± 2.5 | 71.0 ± 2.5 |
| Female, n (%) | 704 (64.2) | 660 (60.1) | 652 (59.4) |
| Race | |||
| White | 934 (86.1) | 924 (84.2) | 899 (81.9) |
| Black | 155 (13.4) | 172 (15.7) | 192 (17.5) |
| Smoking status | |||
| Never | 488 (44.5) | 506 (46.1) | 445 (40.5) |
| Former | 477 (43.5) | 468 (42.6) | 472 (43.0) |
| Current | 130 (11.9) | 122 (11.1) | 181 (16.5) |
| BMI (kg/m2) | 26.2 ± 4.4 | 27.0 ± 4.8 | 27.8 ± 4.9 |
| Systolic blood pressure (mmHg) | 134± 20 | 133 ± 20 | 136 ± 22 |
| Diastolic blood pressure (mmHg) | 71 ± 11 | 72 ± 11 | 71 ± 12 |
| LDL cholesterol (mg/dL) | 130 ± 35 | 133 ± 34 | 130 ± 38 |
| HDL cholesterol (mg/dL) | 58 ± 17 | 54 ± 15 | 51 ± 14 |
| Hypertension | 385 (35.3) | 436 (39.7) | 617 (56.2) |
| Diabetes | 146 (13.3) | 105 (9.6) | 148 (13.5) |
| NT-pro-BNP (pg/mL), median ± IQR | 80 ± 101 | 92 ± 120 | 139 ± 219 |
| Antihypertensive medication | 419 (38.2) | 462 (42.2) | 660 (60.2) |
| Lipid-lowering medication | 72 (6.6) | 59 (5.4) | 82 (7.5) |
Note: BIS2 = Berlin Initiative Study 2; BMI = body mass index; CRP = C-reactive protein; eGFR = estimated glomerular filtration rate; HDL = high-density lipoprotein; IQR = interquartile range; LDL = low-density lipoprotein; NT-pro-BNP = N-terminal pro-B-type natriuretic peptide.
Table 2.
Descriptive Characteristics of NT-Pro-BNP Tertiles at Baseline
| Characteristics | NT-Pro-BNP Tertiles pg/mL (range), n | ||
|---|---|---|---|
| First Tertile | Second Tertile | Third Tertile | |
| (5–66) | (67–148) | (149–23,445) | |
| 943 | 942 | 942 | |
| BIS2 eGFR (mL/min/1.73 m2) | 70.5 ± 12.4 | 67.6 ± 12.0 | 62.4 ± 14.3 |
| Age | 70.1 ± 2.4 | 70.4 ± 2.5 | 71.0 ± 2.5 |
| Female, n (%) | 504 (53.5) | 590 (62.6) | 589 (62.5) |
| Race | |||
| White | 743 (78.8) | 810 (86.0) | 810 (86.0) |
| Black | 192 (20.4) | 129 (13.7) | 128 (13.6) |
| Smoking status | |||
| Never | 405 (43.0) | 423 (44.9) | 423 (45.0) |
| Former | 418 (44.3) | 417 (44.3) | 397 (42.2) |
| Current | 119 (12.6) | 102 (10.8) | 120 (12.8) |
| BMI (kg/m2) | 27.5 ± 4.7 | 27.0 ± 4.7 | 26.7 ± 4.8 |
| Systolic blood pressure (mmHg) | 132 ± 19 | 134 ± 20 | 140 ± 23 |
| Diastolic blood pressure (mmHg) | 72 ± 11 | 71 ± 11 | 71 ± 12 |
| LDL cholesterol (mg/dL) | 134 ± 36 | 132 ± 34 | 126 ± 36 |
| HDL cholesterol (mg/dL) | 54 ± 15 | 55 ± 16 | 54 ± 16 |
| Hypertension | 354 (38.6) | 390 (41.4) | 504 (53.5) |
| Diabetes | 136 (14.5) | 114 (12.1) | 137 (14.6) |
| Antihypertensive medication | 389 (41.3) | 396 (42.2) | 548 (58.2) |
| Lipid-lowering medication | 62 (6.6) | 59 (6.3) | 62 (6.6) |
Note: BIS2 = Berlin Initiative Study 2; BMI = body mass index; CRP = C-reactive protein; eGFR = estimated glomerular filtration rate; HDL = high-density lipoprotein; LDL = low-density lipoprotein; NT-pro-BNP = N-terminal pro-B-type natriuretic peptide.
The level of eGFR decreased with age (Figure 1A), from 70 mL/min/1.73 m2 in participants <70 years to 54 mL/min/1.73 m2 in participants 80+ years. The estimated change in eGFR remained constant across age groups at a rate of approximately −1 mL/min/1.73 m2/year (Figure 1B). NT-pro-BNP levels were highly skewed and increased with age, from a median level of 86 pg/mL in participants <70 years to 208 pg/mL in participants 80+ years (Figure 2A). The estimated change in NT-pro-BNP also remained relatively constant across age groups at 0.1 ln(pg/mL)/year, which is equivalent to a 1.11-fold increase (Figure 2B).
Figure 1.
Box plots of estimated glomerular filtration rate (eGFR) level (left) and change (right) by age category. BIS2 = Berlin Initiative Study 2.
Figure 2.
Box plot N-terminal pro-B-type natriuretic peptide (NT-pro-BNP) level (left) and change (right) by age category.
There was an association between level and change of eGFR with longevity that persisted after adjustment for potential confounders. In models of level and change in eGFR, each 10 mL/min/1.73 m2 higher eGFR was associated with a 23% higher likelihood of surviving to age 90 (odds ratio [OR]: 1.23, 95% CI: 1.13, 1.34) (Table 3). An 0.5 mL/min/1.73 m2 slower decline per year was associated with a higher odds of surviving to age 90 (OR: 1.51, 95% CI: 1.31, 1.74). These associations were attenuated when NT-pro-BNP was included in the model, and the effect size for eGFR level was no longer statistically significant (OR: 1.06, 95% CI: 0.97, 1.16, per 10 mL/min/1.73 m2 higher eGFR).
Table 3.
Unadjusted and Adjusted Association of Standardized Level and Change of eGFR and Log NT-Pro-BNP and Survival to Age 90
| Standardization Units | Unadjusted | Adjusted* | Adjusted eGFR + NT-Pro-BNP† | |
|---|---|---|---|---|
| OR (CI) | OR (CI) | OR (CI) | ||
| eGFR level (mL/min/1.73 m2) | 10 higher level | 1.25 (1.16–1.35) | 1.23 (1.13–1.34) | 1.06 (0.97–1.16) |
| eGFR change (mL/min/1.73 m2/year) | 0.5 slower decline | 1.56 (1.37–1.79) | 1.51 (1.31–1.74) | 1.36 (1.18–1.58) |
| NT-pro-BNP level (pg/mL) | Doubling in level | 0.71 (0.65–0.77) | 0.67 (0.61–0.73) | 0.69 (0.62–0.76) |
| NT-pro-BNP change (pg/mL/year) | 1.05-fold increase | 0.49 (0.40–0.59) | 0.53 (0.43–0.65) | 0.55 (0.45–0.68) |
Notes: BMI = body mass index; CRP = C-reactive protein; eGFR = estimated glomerular filtration rate; HDL = high-density lipoprotein; LDL = low-density lipoprotein; NT-pro-BNP = N-terminal pro-B-type natriuretic peptide; OR = odds ratio.
*The models were adjusted for age, sex, race, education level, systolic blood pressure, diastolic blood pressure, diabetes, smoking status, BMI, alcohol consumption, antihypertensive medication use, lipids, HDL, LDL, CRP, myocardial infarction at baseline, stroke at baseline, nonmelanoma cancer diagnosis at baseline.
†The model included level and change for both eGFR and NT-pro-BNP in the same model and was adjusted for all the covariates included in the prior model.
Similarly, both level and change in NT-pro-BNP were associated with longevity. Each doubling in the level of NT-pro-BNP was associated with a lower likelihood of surviving to age 90 (OR: 0.67, 95% CI: 0.61, 0.73) (Table 3). Additionally, a 1.05-fold increase in NT-pro-BNP per year was associated with a lower likelihood of surviving to age 90 (0.53, 95% CI: 0.43, 0.65). The association of level and change in NT-pro-BNP was not meaningfully altered by concurrent adjustment for eGFR.
There was a small, but statistically significant, multiplicative interaction between low eGFR and high NT-pro-BNP level (OR: 1.12, 95% CI: 1.03, 1.21, p = .009). The estimated effect of eGFR was greater in persons with higher NT-pro-BNP, and the estimated effect of NT-pro-BNP was greater in persons with lower eGFR (Supplementary Figure 1). The effect size of the interaction is equivalent to an OR = 1.19 per 10 mL/min/1.73 m2 higher eGFR for participants with a NT-pro-BNP level double the median, and an OR = 0.62 for a doubling of NT-pro-BNP for participants with an eGFR 10 mL/min/1.73 m2 lower than the mean. In a sensitivity analysis that included baseline and incident stroke, heart failure, and myocardial infarction, the association between eGFR level and change with survival to age 90 were unchanged: 1.07 (95% CI: 0.97, 1.17) per 10 mL/min/1.73 m2, and 1.37 (95% CI: 1.28, 1.59) per 0.5 mL/min/1.73 m2 slower decline per year. The estimates of NT-pro-BNP level and change with survival to age 90 were also unchanged: 0.71 (95% CI: 0.64, 0.78) per doubling of NT-pro-BNP and 0.59 (95% CI: 0.48, 0.73) per 1.05-fold increase, respectively. In a second sensitivity analysis which included albumin to creatinine ratio in the model, the associations between eGFR level and change with survival to age 90 were 1.08 (95% CI: 0.96, 1.22) and 1.22 (95% CI: 1.05, 1.43), respectively.
In generalized estimating equation models of repeated measures of the risk factors, there was a significant, but modest, interaction between eGFR and age such that the association of eGFR with survival to 90 was attenuated at older ages (p = .001). For example, when measured at age 70, each SD higher eGFR was associated with an OR of reaching age 90 of 1.28 (95% CI: 1.19, 1.37), whereas this effect was attenuated to 1.12 (95% CI: 0.96, 1.30) when measured at age 80. In contrast, the association of NT-pro-BNP with survival to 90 was strengthened when measured at older age (p-value for age interaction = .008). For example, when measured at age 70, each SD higher ln-NT-pro-BNP was associated with an OR of reaching age 90 of 0.78 (95% CI: 0.73, 0.83), whereas this effect was strengthened to 0.63 (95% CI: 0.50, 0.79) when measured at age 80. There was not an interaction of either eGFR or NT-pro-BNP with systolic blood pressure (p = .1, .5, respectively), sex (p = .9, .4, respectively), or race (p = .8, .05, respectively).
Discussion
The results of this study suggest that both a low mean level and a decrease in eGFR, and a high mean NT-pro-BNP level and an increase are associated with a lower likelihood of surviving to age 90. Additionally, we observed a small synergistic multiplicative interaction between levels of eGFR and NT-pro-BNP, such that persons with low eGFR and high NT-pro-BNP have a lower likelihood of reaching age 90 than persons with either risk factor alone, although a higher NT-pro-BNP baseline level is the stronger contributor to the decrease of the probability of survival to age 90. Finally, we observed a modest attenuation of the relationship between eGFR and survival to 90 when measured at older ages, and a modest increase with age in the association of NT-pro-BNP and survival to 90.
Our study is consistent with previous literature demonstrating the importance of kidney function on survival at older ages. Previous research in the CHS has demonstrated the importance of level and change in cystatin C with mortality, but did not evaluate survival to the very old age (11,12). A study using Health Aging, and Body Composition study data evaluated patients aged 70–79 described a significant association between renal function, measured with cystatin C, and mortality (22). Our study is the first investigation to extend this to evaluate survival to very old age. Hallan and colleagues using data from Chronic Kidney Disease Prognosis Consortium, that consists in different worldwide cohorts, showed an association between eGFR and mortality and showed a reduced mortality risk with increasing age (23).
Our results are also consistent with investigations that have demonstrated the risks of high NT-pro-BNP in older age. A study by Kragelund and colleagues evaluated patients with symptoms or signs of coronary artery disease and reported a 2.4-fold increased risk of all-cause death among patients with the highest levels of NT-pro-BNP compared with those in the lowest level (24). Another study using data from the Prospective Investigation of the Vasculature in Uppsala Seniors evaluated the effect of baseline level and change of NT-pro-BNP on all-cause mortality (25). The researchers reported a 4.3-fold increase in the risk of all-cause mortality among patients with increasing levels of NT-pro-BNP compared with patients with unchanged levels. Moreover, they reported greater mortality among participants who had a higher baseline level and an increasing change. Our findings agree with what has been reported in these studies and adds valuable information by including NT-pro-BNP level and change in the same model and by evaluating survival to very old age.
A better understanding of the factors associated with survival to 90 could improve our understanding of the physiologic pathways that promote longevity. The mechanisms by which eGFR and NT-pro-BNP are associated with survival to age 90 are not entirely known. Probable explanations of the association between eGFR and survival to age 90 are that poor renal function is a marker of poor health and may be a marker of prolonged cardiovascular disease (26). Kidney function decline worsens cardiovascular risk factors like hypertension and dyslipidemia and causes retention of inflammatory solutes that lead to oxidative stress and cause vascular damage (12). Moreover, renal function might be associated with the increase of other risk factors for decreased survival, such as decreased appetite, loss of lean body mass, decreased physical function, decreased cognitive function, and frailty (27–29). A study by Montesanto and colleagues using BIS1 in two cohorts of older adults (64–89 years, 90 or more) found a significant association between eGFR and frailty status. Moreover, they found that the addition of eGFR to the frailty classification improved the prediction of mortality in the group aged 64–89 (10). The dilution of the effect of eGFR on survival with age might be explained by depletion of susceptible individuals; people that are susceptible to the risk factor will experience complications earlier in life, and people that are older may be less susceptible to that risk factor. An explanation for the effect of NT-pro-BNP and the association to survival to age 90 may be that BNP is a marker for cardiovascular disease (30). This association has been found to be significant in the oldest old and independent of decreasing renal function (30). Furthermore, high levels of NT-pro-BNP are associated with faster cognitive and functional decline (13). The strengthening of the association between NT-pro-BNP and survival to age 90 with age can probably be explained by the fact that sustained high levels of NT-pro-BNP reflect sustained cardiac dysfunction.
The strengths of the study include a large cohort of older adults with an extended follow-up, and statistical models include level and change. To our knowledge, this is the first study that models the combined effect of eGFR and NT-pro-BNP on survival to age 90. Additionally, we present a categorized effect by eGFR and NT-pro-BNP levels to facilitate the interpretation of results. We also need to take into consideration some limitations. First, we did not evaluate the mediator effect of cardiovascular diseases. Second, because of the observational nature of the study there might be unmeasured confounders.
In summary, our work extends previous investigations into level and change in eGFR and NT-pro-BNP as risk factors for survival and demonstrates that these are important predictors of survival to age 90. Unlike many traditional risk factors, the associations of these risk factors with mortality appear to extend to very old age. Future studies may investigate whether interventions that target these risk factors can help improve healthy longevity.
Supplementary Material
Acknowledgments
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Authors contributed to the following roles: study concept and design (A.D.H., M.C.O.), acquisition of participants and/or data (M.L.B., M.C., B.M.P., A.B.N., M.G.S., J.G., J.M.G.), analysis of data (A.D.H., M.C.O.), interpretation of data and results (A.D.H., M.L.B., M.C., B.M.P., A.B.N., M.G.S., C.W., N.B., M.C.O.), preparation of manuscript (A.D.H., M.C.O.), and revision of manuscript (A.D.H., M.L.B., M.C., B.M.P., A.B.N., M.G.S., J.G., C.W., J.M.G., N.B., M.C.O.).
Funding
This work was supported by contracts HHSN268201200036C, HHSN268200800007C, HHSN268201800001C, N01HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083, N01HC85086, and grants U01HL080295 and U01HL130114 from the National Heart, Lung, and Blood Institute (NHLBI), with additional contribution from the National Institute of Neurological Disorders and Stroke (NINDS). Additional support was provided by R01AG023629 and R21HL135869 from the National Institute on Aging (NIA). A full list of principal CHS investigators and institutions can be found at CHS-NHLBI.org.
Conflict of Interest
M.C.O. serves as a paid consultant for Cricket Health, Inc., a kidney health care company.
References
- 1. Tate RB, Manfreda J, Cuddy TE. The effect of age on risk factors for ischemic heart disease. Ann Epidemiol. 1998;8(7):415–421. doi: 10.1016/S1047-2797(98)00011-8 [DOI] [PubMed] [Google Scholar]
- 2. Kronmal RA, Cain KC, Ye Z, Omenn GS. Total serum cholesterol levels and mortality risk as a function of age. A report based on the Framingham data. Arch Intern Med. 1993;153(9):1065–1073. doi: 10.1001/archinte.1993.00410090025004 [DOI] [PubMed] [Google Scholar]
- 3. Psaty BM, Anderson M, Kronmal RA, et al. The association between lipid levels and the risks of incident myocardial infarction, stroke, and total mortality: the Cardiovascular Health Study. J Am Geriatr Soc. 2004;52(10):1639–1647. doi: 10.1111/j.1532-5415.2004.52455.x [DOI] [PubMed] [Google Scholar]
- 4. Odden MC, Shlipak MG, Whitson HE, et al. Risk factors for cardiovascular disease across the spectrum of older age: the Cardiovascular Health Study. Atherosclerosis. 2014;237(1):336–342. doi: 10.1016/j.atherosclerosis.2014.09.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Wang TJ, Gona P, Larson MG, et al. Multiple biomarkers for the prediction of first major cardiovascular events and death. N Engl J Med. 2006;355:2631–2639. doi: 10.1056/NEJMoa055373 [DOI] [PubMed] [Google Scholar]
- 6. Kistorp C, Raymond I, Pedersen F, Gustafsson F, Faber J, Hildebrandt P. Levels as predictors of mortality and cardiovascular events in older adults. JAMA. 2014;154(13):1609–1616. doi: 10.1001/jama.293.13.1609 [DOI] [PubMed] [Google Scholar]
- 7. Kavousi M, Elias-Smale SE, Rutten JHW, Leening MJG. Evaluation of newer risk markers for coronary heart disease risk classification: a cohort study. Ann Intern Med. 2012;156(6):438–444. doi: 10.7326/0003-4819-156-6-201203200-00006. [DOI] [PubMed] [Google Scholar]
- 8. Zethelius B, Berglund L, Sundström J, et al. Use of multiple biomarkers to improve the prediction of death from cardiovascular causes. N Engl J Med. 2008;358(20):2107–2116. doi: 10.1056/NEJMoa0707064 [DOI] [PubMed] [Google Scholar]
- 9. Luo J, Wang LP, Hu HF, et al. Cystatin C and cardiovascular or all-cause mortality risk in the general population: a meta-analysis. Clin Chim Acta. 2015;450:39–45. doi: 10.1016/j.cca.2015.07.016 [DOI] [PubMed] [Google Scholar]
- 10. Montesanto A, De Rango F, Berardelli M, et al. Glomerular filtration rate in the elderly and in the oldest old: correlation with frailty and mortality. Age (Dordr). 2014;36(3):9641. doi: 10.1007/s11357-014-9641-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Shastri S, Katz R, Rifkin DE, et al. Kidney function and mortality in octogenarians: Cardiovascular Health Study All Stars. J Am Geriatr Soc. 2012;60(7):1201–1207. doi: 10.1111/j.1532-5415.2012.04046.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Rifkin DE, Shlipak MG, Katz R, et al. Rapid kidney function decline and mortality risk in older adults. Arch Intern Med. 2008;168(20):2212–2218. doi: 10.1001/archinte.168.20.2212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. van Peet PG, de Craen AJ, Gussekloo J, de Ruijter W. Plasma NT-proBNP as predictor of change in functional status, cardiovascular morbidity and mortality in the oldest old: the Leiden 85-plus study. Age (Dordr). 2014;36(3):9660. doi: 10.1007/s11357-014-9660-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Odden MC, Koh WJH, Arnold AM, Rawlings AM, Psaty BM, Newman AB. Trajectories of Nonagenarian Health: sex, age, and period effects. Am J Epidemiol. 2019;188(2):382–388. doi: 10.1093/aje/kwy241 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Ives DG, Fitzpatrick AL, Bild DE, et al. Surveillance and ascertainment of cardiovascular events. The Cardiovascular Health Study. Ann Epidemiol. 1995;5(4):278–285. doi: 10.1016/1047-2797(94)00093-9 [DOI] [PubMed] [Google Scholar]
- 16. Psaty BM, Delaney JA, Arnold AM, et al. Study of cardiovascular health outcomes in the era of claims data. Circulation. 2016;133(2):156–164. doi: 10.1161/circulationaha.115.018610 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Schaeffner ES, Ebert N, Delanaye P, et al. Two novel equations to estimate kidney function in persons aged 70 years or older. Ann Intern Med. 2012;157(7):471–481. doi: 10.7326/0003-4819-157-7-201210020-00003 [DOI] [PubMed] [Google Scholar]
- 18. Alshaer IM, Kilbride HS, Stevens PE, et al. External validation of the Berlin equations for estimation of GFR in the elderly. Am J Kidney Dis. 2014;63(5):862–865. doi: 10.1053/j.ajkd.2014.01.013 [DOI] [PubMed] [Google Scholar]
- 19. Lopes MB, Araújo LQ, Passos MT, et al. Estimation of glomerular filtration rate from serum creatinine and cystatin C in octogenarians and nonagenarians. BMC Nephrol. 2013;14(1):265. doi: 10.1186/1471-2369-14-265 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18(6):499–502. https://www.scopus.com/record/display.uri?eid=2-s2.0-0015348189&origin=inward. Accessed May 28, 2019. [PubMed] [Google Scholar]
- 21. Psaty BM, Kuller LH, Bild D, et al. Methods of assessing prevalent cardiovascular disease in the Cardiovascular Health Study. Ann Epidemiol. 1995;5(4):270–277. doi: 10.1016/1047-2797(94)00092-8 [DOI] [PubMed] [Google Scholar]
- 22. Shlipak MG, Wassel Fyr CL, Chertow GM, et al. Cystatin C and mortality risk in the elderly: the health, aging, and body composition study. J Am Soc Nephrol. 2006;17(1):254–261. doi: 10.1681/ASN.2005050545 [DOI] [PubMed] [Google Scholar]
- 23. Hallan SI, Matsushita K, Sang Y, et al. ; Chronic Kidney Disease Prognosis Consortium . Age and association of kidney measures with mortality and end-stage renal disease. JAMA. 2012;308(22):2349–2360. doi: 10.1001/jama.2012.16817 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Kragelund C, Grønning B, Køber L, Hildebrandt P, Steffensen R. N-terminal pro-B-type natriuretic peptide and long-term mortality in stable coronary heart disease. N Engl J Med. 2005;352(7):666–675. doi: 10.1056/NEJMoa042330 [DOI] [PubMed] [Google Scholar]
- 25. Eggers KM, Venge P, Lind L. Prognostic usefulness of the change in N-terminal pro B-type natriuretic peptide levels to predict mortality in a single community cohort aged ≥ 70 years. Am J Cardiol. 2013;111(1):131–136. doi: 10.1016/j.amjcard.2012.08.058 [DOI] [PubMed] [Google Scholar]
- 26. Fried LF, Shlipak MG, Crump C, et al. Renal insufficiency as a predictor of cardiovascular outcomes and mortality in elderly individuals. J Am Coll Cardiol. 2003;41(8):1364–1372. doi: 10.1016/s0735-1097(03)00163-3 [DOI] [PubMed] [Google Scholar]
- 27. Shlipak MG, Stehman-Breen C, Fried LF, et al. The presence of frailty in elderly persons with chronic renal insufficiency. Am J Kidney Dis. 2004;43(5):861–867. doi: 10.1053/j.ajkd.2003.12.049 [DOI] [PubMed] [Google Scholar]
- 28. Yaffe K, Lindquist K, Shlipak MG, et al. Cystatin C as a marker of cognitive function in elders: findings from the health ABC study. Ann Neurol. 2008;63(6):798–802. doi: 10.1002/ana.21383 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Odden MC, Chertow GM, Fried LF, et al. ; HABC Study . Cystatin C and measures of physical function in elderly adults: the Health, Aging, and Body Composition (HABC) Study. Am J Epidemiol. 2006;164(12):1180–1189. doi: 10.1093/aje/kwj333 [DOI] [PubMed] [Google Scholar]
- 30. Poortvliet R, de Craen A, Gussekloo J, de Ruijter W. Increase in N-terminal pro-brain natriuretic peptide levels, renal function and cardiac disease in the oldest old. Age Ageing. 2015;44(5):841–847. doi: 10.1093/ageing/afv091 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.