Abstract
Background
Systolic blood pressure (SBP) often varies between clinic visits within individuals, which can affect estimation of cardiovascular disease (CVD) risk.
Methods and Results
We analyzed data from participants with two clinic visits separated by a median of 17 days in the Third National Health and Nutrition Examination Survey (n=808). Ten-year CVD risk was calculated with SBP obtained at each visit using the Pooled Cohort Equations. The mean age of participants was 46.1 years and 47.3% were male. The median SBP difference between the two visits was −1 mmHg (1st – 99th percentiles: −23 to 32 mmHg). The median estimated 10-year CVD risk was 2.5% and 2.4% at the first and second visit, respectively (1st – 99th percentiles −5.2% to +7.1%). Meaningful risk reclassification (i.e., across the guideline recommended 7.5% threshold for statin initiation) occurred in 12 (11.3%) of 106 participants whose estimated CVD risk was between 5% to 10%, but only in 2 (0.3%) of 702 participants who had a 10-year estimated CVD risk of <5% or >10%.
Conclusions
SBP variability can affect CVD risk estimation, and can influence statin eligibility for individuals with an estimated 10-year CVD risk between 5% and 10%.
Keywords: systolic blood pressure, risk assessment, statins, adults
Introduction
The newly published American College of Cardiology (ACC)/ American Heart Association (AHA) Guideline on the Treatment of Blood Cholesterol to Reduce Atherosclerotic Cardiovascular Risk in Adults recommend estimating 10-year cardiovascular disease (CVD) risk using the Pooled Cohort Equations to guide initiation of statin therapy for primary prevention.(1, 2) Systolic blood pressure is one of the measures used in the calculation of estimated 10-year CVD risk.(1) Substantial change in systolic blood pressure between sequential clinic visits has been noted in research studies and in clinical practice.(3–6) For example, Rothwell et al. demonstrated that 15–35% of participants treated for hypertension or who had a prior transient ischemic attack had changes in systolic blood pressure greater than 50 mm Hg across multiple clinic visits.(3) Therefore, systolic blood pressure variability has the potential to affect estimation of an individual's 10-year CVD risk, potentially leading to uncertainty for the classification of patients into risk categories intended to guide prevention efforts. However, scarce data are available on how likely this reclassification is to occur.
With the release of the 2013 ACC / AHA guidelines that emphasize absolute CVD risk for the initiation of statin therapy,(2) there is a need to characterize the impact of changes in systolic blood pressure across clinic visits on the stability of CVD risk estimation. To address this, we used data from the subset of participants in the Third National Health and Nutrition Examination Survey (NHANES III) who completed two clinic visits over a period of several weeks, to determine the effect of changes in systolic blood pressure across visits on changes in estimated 10-year CVD risk.
Methods
NHANES III was a cross-sectional survey designed to select a representative sample of the civilian non-institutionalized US population.(7) The protocol for NHANES III was approved by the National Center for Health Statistics of the Centers for Disease Control and Prevention Institutional Review Board, and all participants provided informed consent. Over the survey period of 1988–1994, 18,825 adults ≥ 20 years of age completed an in-home interview and a visit to a mobile clinic (Visit 1) for medical evaluation including three blood pressure measurements. A subsample (n=2,174) attended a second visit at the mobile clinic (Visit 2), at which time the medical evaluation, including three blood pressure measurements, was repeated. The median duration between the two clinic visits was 17 days (range, 1 to 48 days). Although more recent NHANES are available, we chose to analyze NHANES III as it has data from a second study visit allowing for the calculation of variability of blood pressure across two study visits. For the current analyses, we included NHANES III participants who had two clinic visits with three blood pressure measurements performed in the same arm at each visit (n=956). We excluded 32 participants missing data on measures (e.g., total cholesterol) needed to calculate estimated 10-year CVD risk, and 116 participants with a history of coronary heart disease (CHD), diabetes, or stroke, as they would be classified as having high CVD risk based on medical history alone.(2) After these exclusions, 808 participants were included in this analysis.
Data Collection and Blood Pressure Measurement
Demographic and health-related information including age, sex, race/ethnicity, and smoking status was collected using a standardized questionnaire during the in-home interview. The use of anti-hypertensive medications was ascertained via self-report. Blood specimens were collected during the medical evaluation to determine total cholesterol and high-density lipoprotein (HDL) cholesterol levels. Blood pressure was measured three times during each of the two clinic visits, with the average of the second and third measurements recorded as the blood pressure for that visit. Blood pressure measurements were performed by a trained clinician using the standard protocol of the American Heart Association at both clinic visits. Each participant's arm was measured to determine the appropriate sized cuff. Participants were asked to remain in a seated position for five minutes of quiet rest prior to their first blood pressure measurement being taken. Quality control for the blood pressure measurements included quarterly recertification with retraining if necessary, annual retraining of all physicians, and monitoring of equipment and equipment repair. Additional details regarding blood pressure measurement and quality control procedures are provided in the NHANES III manual of operations.(7)
Statistical Analysis
At each clinic visit, we calculated the participant's estimated 10 year CVD risk using the Pooled Cohort Equations based on age, sex, race, systolic blood pressure, total cholesterol, HDL cholesterol, smoking status, and use of antihypertensive medication.(1) In order to isolate the effect of systolic blood pressure variability on estimated 10-year CVD risk, we used the systolic blood pressure measured at each visit, while assuming other parameters (e.g., cholesterol levels, smoking status) remain unchanged from the first visit. We then computed the difference in 10-year CVD risk estimates calculated at the two visits, defined as the risk estimate from Visit 2 minus the risk estimate from Visit 1. The summary measures for number of days between clinic visits, systolic blood pressure, diastolic blood pressure, and the estimated 10-year CVD risk at each visit were expressed as medians and 25th to 75th percentiles due to non-normal distributions of these variables. Agreement between the two 10-year CVD risk estimates was assessed using a Bland-Altman plot,(8) and the proportion of individuals with ≥1%, ≥2%, ≥3%, ≥4%, and ≥5% absolute change in estimated 10-year CVD risk across two clinic visits was calculated. Next, we examined the effect of change in blood pressure on change in estimated 10-year CVD risk across two visits. First, we calculated the percentage of participants with systolic blood pressure < 130, 130 to < 150, and ≥ 150 mm Hg and diastolic blood pressure < 75, 75 to 84 and ≥ 85 mm Hg at the first visit who experienced ≥3% absolute change in estimated 10-year CVD risk between Visit 1 and Visit 2; ≥3% was chosen to represent a change large enough to have potential clinical relevance. Using logistic regression models with systolic blood pressure < 130 mm Hg as the reference group, we analyzed whether systolic blood pressure between 130 to < 150 mm Hg and ≥ 150 mm Hg at time of Visit 1 were associated with a ≥3% absolute change in estimated 10-year CVD risk at Visit 2 after multivariable adjustment. An initial model adjusted for age, sex, race-ethnicity, and a second model further included additional adjustment for diastolic blood pressure. Logistic regression models were also constructed for diastolic blood pressure (<75 [reference], 75 to 84 and ≥ 85 mm Hg) with two levels of adjustment for (Model 1) age, sex, race-ethnicity and (Model 2) age, sex, race-ethnicity and systolic blood pressure.
Using 10-year CVD risk categories of <5%, 5% to <7.5%, 7.5% to <10%, and ≥10%,(2) we calculated the proportion of participants who were reclassified from Visit 1 to Visit 2. Specifically, we calculated the percentage of participants who stayed in the same risk-category between Visits 1 and 2, were re-classified downwards (e.g., from ≥10% at Visit 1 to 7.5% to <10% at Visit 2), or were re-classified upwards (e.g., from 7.5% to <10% at Visit 1 to ≥10% at Visit 2). Examples of NHANES III participants whose risk categories changed due to change in systolic blood pressure across the two clinic visits are provided to illustrate the clinical implications of our findings. In a secondary analysis, we repeated the above analysis using the Framingham risk equation for estimated 10-year CHD risk from the Adult Treatment Panel III guidelines and the recommended risk categories of <10%, 10% to <20%, and ≥20%.(9, 10)
All analyses were conducted without sampling weights as recommended for NHANES III second examination data, and were performed using the statistical package SAS version 9.2 (SAS Institute, Cary, NC).
Results
Of the 808 participants included in the analyses, the mean (SD) age was 46.1 (17.2) years; 47.3% were men; 42.2% were non-Hispanic white; and 29.2% were current smokers (Table 1). The mean (SD) level for total cholesterol was 5.31 (1.08) mmol/L and for HDL cholesterol was 1.31 (0.39) mmol/L. Overall, 14.5% of participants were taking antihypertensive medications.
Table 1.
Characteristics of NHANES III participants who had blood pressure measurements at two clinic visits.
| Characteristic | N=808 |
|---|---|
|
| |
| Age, years | 46.1 (17.2) |
|
| |
| Male sex | 47.3% |
|
| |
| Race-ethnicity | |
| Non-Hispanic white | 42.2% |
| Non-Hispanic black | 27.9% |
| Mexican American | 25.7% |
| Other | 4.2% |
|
| |
| Current smoking | 29.2% |
|
| |
| Total cholesterol, mmol/L | 5.31 (1.08) |
|
| |
| HDL-cholesterol, mmol/L | 1.31 (0.39) |
|
| |
| On antihypertensive treatment | 14.5% |
HDL, high-density lipoprotein; numbers in table are mean (standard deviation) or percentage.
Blood Pressure and Estimated CVD Risk at Two Clinic Visits
At Visit 1, the median (25th to 75th percentiles) systolic blood pressure was 118 (108 to 131) mm Hg (Table 2), and the median (25th to 75th percentiles) estimated 10-year CVD risk was 2.5% (0.5% to 10.0%). At Visit 2, the median (25th to 75th percentiles) systolic blood pressure was 117 (107 to 129) mm Hg, and the median (25th to 75th percentiles) estimated 10-year CVD risk was 2.4% (0.5% to 10.0%). The median within-person differences in systolic blood pressure between Visit 1 and Visit 2 was -1 mm Hg (1st to 99th percentiles, −23 to 32 mmHg). For individual participants, changes in systolic blood pressure from Visit 1 to Visit 2 resulted in a wide distribution of changes in estimated 10-year CVD risk (median = 0.0%; 1st to 99th percentiles –5.2% to +7.1%). Overall, 23% of participants had an absolute difference in estimated 10-year CVD risk that was ≥1% across the 2 visits; 8% of participants had an absolute difference that was ≥3%; and 4% of participants had an absolute difference that was ≥5% (Figure 1). On average, the difference in estimated 10-year CVD risk between the two visits was larger for individuals with higher estimated risk (Figure 2).
Table 2.
Systolic and diastolic blood pressure and Pooled Cohort Equations estimated 10-year cardiovascular disease risk at each NHANES III visit.
| Percentile | |||||
|---|---|---|---|---|---|
|
| |||||
| 1st | 25th | 50th | 75th | 99th | |
|
| |||||
| Clinic Visit 1 BP, mmHg | |||||
| Systolic | 91 | 108 | 118 | 131 | 180 |
| Diastolic | 47 | 65 | 73 | 80 | 100 |
|
| |||||
| Clinic Visit 2 BP, mmHg | |||||
| Systolic | 90 | 107 | 117 | 129 | 174 |
| Diastolic | 46 | 65 | 72 | 79 | 97 |
|
| |||||
| Difference in BP, mmHg (clinic visit 2 - clinic visit 1) | |||||
| SBP | −23 | −7 | −1 | 4 | 32 |
| DBP | −17 | −5 | 0 | 4 | 21 |
|
| |||||
| Estimated 10-year CVD risk*, % | |||||
| Clinic Visit 1 SBP | <0.1% | 0.5% | 2.5% | 10.0% | 49.9% |
| Clinic Visit 2 SBP | <0.1% | 0.5% | 2.4% | 10.0% | 45.8% |
|
| |||||
| Differences in estimated 10-year CVD risk, % (clinic visit 1 - clinic visit 2) | −5.2% | −0.1% | 0.0% | 0.3% | 7.1% |
BP: blood pressure; SBP, systolic blood pressure; DBP, diastolic blood pressure; CVD: cardiovascular disease.
Estimated 10-year CVD risk are calculated from the Pooled Cohort Equations recommended by the American College of Cardiology Foundation / American Heart Association 2013 Guideline on the Assessment of Cardiovascular Risk.
Figure 1. Percentage of participants with an absolute change in estimated 10-year cardiovascular disease (CVD) risk ≥ 1% to ≥ 5% resulting from change in systolic blood pressure across two clinic visits.
Absolute change in estimated 10-year CVD risk is calculated as the absolute difference between Pooled Cohort Equations derived risk estimates determined at clinic visit 1 and clinic visit 2, using systolic blood pressure measured at each visit while holding all other variables constant.
Figure 2.
Bland-Altman plot for comparison of estimated 10-year cardiovascular disease (CVD) risk calculated using the Pooled Cohort Equations at two clinic visits.
Overall, 2.9% of participants with systolic blood pressure < 130 mm Hg at Visit 1 had a ≥3% absolute change in estimated 10-year CVD risk between Visits 1 and 2, compared to 13.6% and 36.0% of participants with systolic blood pressure 130–149 and ≥150 mm Hg, respectively, at Visit 1 (Table 3, top panel). After adjustment for age, gender, race-ethnicity, and diastolic blood pressure and compared to participants with systolic blood pressure < 130 mm Hg at Visit 1, the odds ratio for having a ≥3% absolute change in estimated 10-year CVD risk between Visits 1 and 2 associated with systolic blood pressure of 130–149 mm Hg and ≥150 mm Hg were 1.86 (95% CI 0.84–4.12) and 5.60 (95% CI 2.23–14.1), respectively. In contrast, no association was present between diastolic blood pressure categories and having a ≥3% between-visit absolute change in estimated 10-year CVD risk (Table 3, bottom panel).
Table 3.
Percent of participants whose estimated 10-year cardiovascular disease (CVD) risk was ≥3% different due to systolic blood pressure variability between two clinic visits and odds ratios for ≥3% difference by categories of systolic blood pressure (top panel) and diastolic blood pressure (bottom panel).
| Systolic blood pressure from Visit 1, mm Hg | |||
| < 130 (n=579) | 130 – 149 (n=154) | >≥ 150 (n=75) | |
| % with ≥3% change in 10-year CVD risk | 2.9% | 13.6% | 36.0% |
| Odds ratio (95% Cl) for reclassification | |||
| Model 1 | 1 (ref) | 1.76(0.80 – 3.84) | 4.72(2.06 – 10.8) |
| Model 2† | 1 (ref) | 1.86(0.84 – 4.12) | 5.60(2.23 – 14.1) |
| Diastolic blood pressure from Visit 1, mm Hg | |||
| < 75 (n=457) | 75 – 84 (n=222) | ≥ 85 (n=129) | |
| % with ≥3% change in 10-year CVD risk | 7.4% | 9.9% | 7.0% |
| Odds ratio (95% Cl) for reclassification | |||
| Model 1 | 1 (ref) | 1.35(0.68 – 2.69) | 1.41(0.55 – 3.66) |
| Model 2†† | 1 (ref) | 1.00(0.48 – 2.06) | 0.60(0.21 – 1.78) |
Model 1 includes adjustment for age, gender, and race-ethnicity
Model 2 includes adjustment for age, gender, race-ethnicity, and diastolic blood pressure
Model 2 includes adjustment for age, gender, race-ethnicity, and systolic blood pressure
Reclassification Due to Changes in Blood Pressure
For participants with an estimated 10-year CVD risk of < 5% at Visit 1, 1.2% were reclassified upwards at Visit 2, including 1 participant who was reclassified into the 7.5% to <10% category (Table 4). In contrast, 12.1% of participants who had an estimated 10-year CVD risk in the 5% to <7.5% category at Visit 1 were reclassified downwards to the < 5% category, while 10.3% were re-classified upwards to the 7.5% to <10% category. Similarly, 12.5% of participants with an estimated 10-year CVD risk in the 7.5% to <10% category at Visit 1 were reclassified downwards to the 5% to <7.5% category and 10 (20.8%) were reclassified upwards into the ≥10% category at Visit 2. Finally, 4.5% of participants with estimated 10-year CVD risk of ≥10% at Visit 1 were reclassified downwards at Visit 2, including 1 participant who was reclassified into the 5% to <7.5% category. Using the Framingham risk equation recommended in the ATP III guidelines and categories of <10%, 10% to 20%, and >20% of estimated 10-year CHD risk, reclassification from Visit 1 to Visit 2 due to change in systolic blood pressure primarily occurred in those in the highest estimated risk category (Table 5).
Table 4.
Reclassification of estimated 10-year cardiovascular disease (CVD) risk of NHANES III participants due to change in systolic blood pressure across two clinic visits.
| Estimated 10-year CVD risk at Visit 1 | ||||
|---|---|---|---|---|
| Estimated 10-year CVD risk at Visit 2 | <5% | 5 to <7.5% | 7.5% to <10% | ≥10% |
| <5% | 494 (98.8%) | 7 (12.1%) | 0 (0%) | 0 (0%) |
| 5 to <7.5% | 5 (1.0%) | 45 (77.6%) | 6 (12.5%) | 1 (0.5%) |
| 7.5% to <10% | 1 (0.2%) | 6 (10.3%) | 32 (66.7%) | 8 (4.0%) |
| ≥ 10% | 0 (0%) | 0 (0%) | 10 (20.8%) | 193 (95.5%) |
| Total | 500 (61.9%) | 58 (7.2%) | 48 (5.9%) | 202 (25.0%) |
CVD, cardiovascular disease; numbers in table are n (%).
*Percentages are calculated per column using the total number of participants in each Clinic Visit 1 estimated 10-year CVD risk category as denominator; may not add up to 100% due to rounding
Table 5.
Reclassification of Framingham equation estimated 10-year coronary heart disease (CHD) risk in NHANES III participants due to change in systolic blood pressure across two clinic visits.
| Estimated 10-year CHD risk at Visit 1 | |||
|---|---|---|---|
| Estimated 10-year CHD risk at Visit 2 | <10% | 10–20% | >20% |
| <10% | 633 (98.9%) | 14 (10.9%) | 0 (0%) |
| 10–20% | 7 (1.1%) | 110 (85.9%) | 10 (25.0%) |
| >20% | 0 (0%) | 4 (3.1%) | 30 (75.0%) |
| Total | 640 (100%) | 128 (100%) | 40 (100%) |
CHD, coronary heart disease; numbers in table are n (%).
*Percentages are calculated per column using the total number of participants in each Clinic Visit 1 estimated 10-year CHD risk category as denominator; may not add up to 100% due to rounding.
We provide examples of two NHANES participants who were reclassified into different 10-year CVD risk categories due to change in systolic blood pressure across two clinic visits in Table 6.
Table 6.
Examples of NHANES III participants whose Pooled Cohort Equations estimated 10-year CVD risk were reclassified due to change in systolic blood pressure across two clinic visits.
| Participant A |
| A 41 year old African-American woman with total cholesterol of 4.27 mmol/L, HDL of 1.03 mmol/L, and SBP of 145 mm Hg at clinic visit 1. She is a smoker and is not on medication treatment for hypertension. Her Pooled Cohort Equations estimated 10-year CVD risk is 4.2%. At clinic visit 2, her SBP is 166 mm Hg, which changes her estimated 10-year CVD risk to 8.4%. |
| Participant B |
| A 46 year old nonHispanic white man with total cholesterol of 6.52 mmol/L, HDL of 0.78 mmol/L, and SBP of 149 mm Hg at clinic visit 1. He does not smoke and does not take any antihypertensive medications. His Pooled Cohort Equations estimated 10-year CVD risk is 7.9%. At clinic visit 2, his SBP is 140 mm Hg, which changes his estimated 10-year CVD risk to 7.1%. |
CVD: cardiovascular disease; HDL, high-density lipoprotein; SBP, systolic blood pressure.
Discussion
In the current study, we demonstrated that change in systolic blood pressure between clinic visits could affect estimation of 10-year CVD risk using the recently published Pooled Cohort Equations. This effect was especially prominent in individuals with an estimated 10-year CVD risk between 5% and 10%. However, reclassification in estimated 10-year CVD risk categories occurred in a very small percentage of individuals at low (i.e., <5%) or high (i.e., ≥10%) estimated risk. The reclassification of risk associated with changes in systolic blood pressure between clinic visits is not unique to the Pooled Cohort Equations, as similar reclassification also occurred using the Framingham 10-year CHD risk equation.
Substantial blood pressure variability has been reported in previous studies. We have described a mean standard deviation of 7.7 mm Hg for systolic blood pressure across three visits (two clinic visits and one home visit) in a subsample of NHANES III participants, as well as a mean standard deviation of 10.9 mm Hg for systolic blood pressure measured across eleven visits in the Women's Health Initiative.(4, 11) Systolic blood pressure variability of a similar magnitude or greater has also been described in patients treated for hypertension or prior transient ischemic attack,(3) and in patients with hypertension followed in primary care clinics.(5) Our current analysis suggests that these levels of blood pressure variability (3–5, 11) can affect estimation of cardiovascular risk using the Pooled Cohort Equations recommended by current guidelines.(1) Especially for individuals whose estimated 10-year CVD risk is near the 7.5% cut-point recommended for treatment decisions in the newly published ACC/AHA Guideline on the Treatment of Blood Cholesterol,(2) systolic blood pressure variability may result in uncertainty for clinical decisions such as statin initiation.
Although the current analysis focused on 10-year CVD risk estimated using the Pooled Cohort Equation, the issue of blood pressure variability is likely relevant to other methods of CVD and CHD risk prediction. As we have demonstrated, change in systolic blood pressure across clinic visits also affected the Framingham equation estimated 10-year CHD risk. Systolic blood pressure is also part of other commonly used risk prediction tools such as the Reynolds, the QRISK, and the SCORE risk scores,(12–14) and the stability of risk classification with these other methods may be affected by blood pressure variability. The increased emphasis on absolute risk in current national and international cardiovascular disease prevention guidelines (2, 15) suggest that that the stability of risk estimation is an issue that warrants further investigation.(16) Given the known challenges of applying risk prediction methods derived from populations to individual clinical encounters,(17) more study is also needed to determine whether and how to discuss the stability of risk classification during shared decision making processes.(18)
Our results also provide guidance for the appropriate use of Pooled Cohort Equations to estimate 10-year CVD risk in real-world settings. It is reassuring that we found change in systolic blood pressure across clinic visits meaningfully impacted estimation of 10-year CVD risk in only a small percentage of participants with 10-year CVD risk estimated to be <5% or ≥10%. These results suggest that systolic blood pressure measured at a single visit is sufficient when the estimated CVD risk is in these ranges. However, for individuals whose initial estimated risk is between 5% to 10%, we found that reclassification may occur for a substantial percentage of individuals due to change in systolic blood pressure. For these individuals, there is potential for over- or under-treatment with statin therapy if the clinical decision-making is guided solely by risk estimation at the initial visit. Further research is needed to clarify the extent to which repeat blood pressure measurements are needed to reliably estimate cardiovascular risk and to avoid under or over-treatment with statins.
Our study has several strengths. First, the NHANES data we used includes a broad cross-section of the US population, making our results highly generalizable. Second, blood pressure measurements in NHANES were performed by trained physicians following a standardized protocol, suggesting that the reclassification observed was due to within-person variability in blood pressure rather than measurement error. In addition, the two blood pressure measurements were taken only several weeks apart, making it less likely that our results are due to natural progression of hypertension or to changes in medication regimens. Taken together, our findings add to the emerging literature on the inherent challenges for applying population-derived risk prediction tools such as the Pooled Cohort Equations to individual clinical encounters.(17, 19)
There are also several limitations to our study. Our sample size was limited by the number of NHANES III participants with two clinic visits, which represents only a subset of NHANES III cohort. However, the distribution of estimated cardiovascular risk in our sample is similar to what has been previously reported for the entire NHANES III cohort,(20) suggesting that our findings are not due to sampling or selection bias. As we calculated systolic blood pressure using the average of the second and third measurements at each visit, caution is also needed to generalize our results to clinical settings where blood pressure reading is often obtained through a single measurement without necessarily following a standardized protocol. However, systolic blood pressure measured in such settings will likely have higher variability across visits.(21) Finally, our analysis focused on risk estimation rather than hard clinical endpoints. However, our previous analysis of this study population also demonstrated that visit-to-visit variability of systolic blood pressure was associated with all-cause mortality, and was correlated with factors such as higher mean systolic blood pressure and the use of angiotensin converting enzyme-inhibitors.(4) Our results will need to be validated in future studies, and the clinical impact of uncertainty in risk estimation caused by systolic blood pressure variability will require further clarification.
In conclusion, we found that change in systolic blood pressure across clinic visits has a clinically meaningful effect on estimation of 10-year CVD risk using the Pooled Cohort Equations, especially for individuals whose estimated risk is near the 7.5% cut-point recommended for consideration of statin therapy. These results may be useful to guide interpretation of estimated 10-year CVD risk during individual clinical encounters where statin initiation is considered, and suggest that there is a need to better understand the effect of blood pressure variability on the stability of risk classification for risk prediction methods.
Acknowledgments
Funding and Support: Dr. Ye is supported by National Institutes of Health Grant T32 HL007854-16, and by National Center for Advancing Translational Sciences, National Institutes of Health, through Grant Number UL1 TR000040, formerly the National Center for Research Resources, Grant Number UL1 RR024156. Drs. Wang and Shimbo are supported by National Institute of Health Grant P01 HL47540. The content of this publication is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Dr. Muntner has received grant funding and served as a consultant to Amgen Corporation. Dr. Levitan has received research support from Amgen Corporation.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflicts of interest: Other authors report no additional disclosures or conflicts of interest.
References
- 1.Goff DC, Lloyd-Jones DM, Bennett G, Coady S, D'Agostino RB, Gibbons R, et al. 2013 ACC/AHA Guideline on the Assessment of Cardiovascular Risk: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2013 Nov 12; 2013. [Google Scholar]
- 2.Stone NJ, Robinson J, Lichtenstein AH, Merz CNB, Blum CB, Eckel RH, et al. 2013 ACC/AHA Guideline on the Treatment of Blood Cholesterol to Reduce Atherosclerotic Cardiovascular Risk in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2013 Nov 12; 2013. [Google Scholar]
- 3.Rothwell PM, Howard SC, Dolan E, O'Brien E, Dobson JE, Dahlof B, et al. Prognostic significance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension. Lancet. 2010 Mar 13;375(9718):895–905. doi: 10.1016/S0140-6736(10)60308-X. PubMed PMID: 20226988. Epub 2010/03/17. eng. [DOI] [PubMed] [Google Scholar]
- 4.Muntner P, Shimbo D, Tonelli M, Reynolds K, Arnett DK, Oparil S. The relationship between visit-to-visit variability in systolic blood pressure and all-cause mortality in the general population: findings from NHANES III, 1988 to 1994. Hypertension. 2011 Feb;57(2):160–6. doi: 10.1161/HYPERTENSIONAHA.110.162255. PubMed PMID: 21200000. Epub 2011/01/05. eng. [DOI] [PubMed] [Google Scholar]
- 5.Powers BJ, Olsen MK, Smith VA, Woolson RF, Bosworth HB, Oddone EZ. Measuring blood pressure for decision making and quality reporting: where and how many measures? Ann Intern Med. 2011 Jun 21;154(12):781–8. doi: 10.7326/0003-4819-154-12-201106210-00005. W-289-90. PubMed PMID: 21690592. Epub 2011/06/22. eng. [DOI] [PubMed] [Google Scholar]
- 6.Cuffe RL, Howard SC, Algra A, Warlow CP, Rothwell PM. Medium-term variability of blood pressure and potential underdiagnosis of hypertension in patients with previous transient ischemic attack or minor stroke. Stroke. 2006 Nov;37(11):2776–83. doi: 10.1161/01.STR.0000244761.62073.05. PubMed PMID: 17008634. Epub 2006/09/30. eng. [DOI] [PubMed] [Google Scholar]
- 7.National Center for Health Statistics Plan and operation of the third National Health and Nutrition Examination survey, 1988–1994. Series 1: programs and collection procedures. Vital Health Stat 1. 1994:94–1308. [PubMed] [Google Scholar]
- 8.Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986 Feb 8;1(8476):307–10. PubMed PMID: 2868172. Epub 1986/02/08. eng. [PubMed] [Google Scholar]
- 9.Framingham Heart Study Hard Coronary Heart Disease (10-year risk) [Accessed September 4, 2012];Framingham Heart Study Web site. 2012 Available at: http://www.framinghamheartstudy.org/risk/hrdcoronary.html.
- 10.Grundy SM, Cleeman JI, Merz CN, Brewer HB, Jr., Clark LT, Hunninghake DB, et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation. 2004 Jul 13;110(2):227–39. doi: 10.1161/01.CIR.0000133317.49796.0E. PubMed PMID: 15249516. Epub 2004/07/14. eng. [DOI] [PubMed] [Google Scholar]
- 11.Shimbo D, Newman JD, Aragaki AK, Lamonte MJ, Bavry AA, Allison M, et al. Association Between Annual Visit-to-Visit Blood Pressure Variability and Stroke in Postmenopausal Women: Data From the Women's Health Initiative. Hypertension. 2012 Sep;60(3):625–30. doi: 10.1161/HYPERTENSIONAHA.112.193094. PubMed PMID: 22753206. Pubmed Central PMCID: 3427141. Epub 2012/07/04. eng. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Berger JS, Jordan CO, Lloyd-Jones D, Blumenthal RS. Screening for cardiovascular risk in asymptomatic patients. J Am Coll Cardiol. 2010 Mar 23;55(12):1169–77. doi: 10.1016/j.jacc.2009.09.066. PubMed PMID: 20298922. Epub 2010/03/20. eng. [DOI] [PubMed] [Google Scholar]
- 13.Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, Minhas R, Sheikh A, et al. Predicting cardiovascular risk in England and Wales: prospective derivation and validation of QRISK2. BMJ (Clinical research ed) 2008 Jun 28;336(7659):1475–82. doi: 10.1136/bmj.39609.449676.25. PubMed PMID: 18573856. Pubmed Central PMCID: PMC2440904. Epub 2008/06/25. eng. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ridker P, Buring JE, Rifai N, Cook NR. Development and validation of improved algorithms for the assessment of global cardiovascular risk in women: The reynolds risk score. JAMA. 2007;297(6):611–9. doi: 10.1001/jama.297.6.611. [DOI] [PubMed] [Google Scholar]
- 15.Perk J, De Backer G, Gohlke H, Graham I, Reiner Z, Verschuren M, et al. European Guidelines on cardiovascular disease prevention in clinical practice (version 2012). The Fifth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of nine societies and by invited experts). Developed with the special contribution of the European Association for Cardiovascular Prevention & Rehabilitation (EACPR) Eur Heart J. 2012 Jul;33(13):1635–701. doi: 10.1093/eurheartj/ehs092. PubMed PMID: 22555213. Epub 2012/05/05. eng. [DOI] [PubMed] [Google Scholar]
- 16.Bower JK, Lazo M, Juraschek SP, Selvin E. WIthin-person variability in high-sensitivity c-reactive protein. Archives of Internal Medicine. 2012;172(19):1519–21. doi: 10.1001/archinternmed.2012.3712. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lloyd-Jones DM. Cardiovascular Risk Prediction. Circulation. 2010 Apr 20;121(15):1768–77. doi: 10.1161/CIRCULATIONAHA.109.849166. 2010. [DOI] [PubMed] [Google Scholar]
- 18.Walsh MN, Bove AA, Cross RR, Ferdinand KC, Forman DE, Freeman AM, et al. ACCF 2012 health policy statement on patient-centered care in cardiovascular medicine: a report of the American College of Cardiology Foundation Clinical Quality Committee. J Am Coll Cardiol. 2012 Jun 5;59(23):2125–43. doi: 10.1016/j.jacc.2012.03.016. PubMed PMID: 22591882. Epub 2012/05/18. eng. [DOI] [PubMed] [Google Scholar]
- 19.Kent DM, Shah ND. Risk models and patient-centered evidence: should physicians expect one right answer? JAMA. 2012 Apr 18;307(15):1585–6. doi: 10.1001/jama.2012.469. PubMed PMID: 22511683. Epub 2012/04/19. eng. [DOI] [PubMed] [Google Scholar]
- 20.Ford ES, Giles WH, Mokdad AH. The distribution of 10-Year risk for coronary heart disease among US adults: findings from the National Health and Nutrition Examination Survey III. J Am Coll Cardiol. 2004 May 19;43(10):1791–6. doi: 10.1016/j.jacc.2003.11.061. PubMed PMID: 15145101. Epub 2004/05/18. eng. [DOI] [PubMed] [Google Scholar]
- 21.Levitan EB, Kaciroti N, Oparil S, Julius S, Muntner P. Blood Pressure Measurement Device, Number and Timing of Visits, and Intra-Individual Visit-to-Visit Variability of Blood Pressure. J Clin Hypertens (Greenwich) 2012 Nov;14(11):744–50. doi: 10.1111/jch.12005. PubMed PMID: 23126345. Epub 2012/11/07. Eng. [DOI] [PMC free article] [PubMed] [Google Scholar]