Mortality and Cardiovascular Risk Across the Ankle-Arm Index Spectrum
http://www.100md.com
《循环学杂志》
the Nephrology Division (A.M.O.), Department of Medicine, VA Medical Center San Francisco and University of California, San Francisco, Calif
Collaborative Health Studies Coordinating Center (R.K.), Seattle, Wash
General Internal Medicine Section (M.G.S.), Veterans Affairs Medical Center, San Francisco, Calif, and Departments of Medicine, Epidemiology and Biostatistics, University of California, San Francisco, Calif
Departments of Medicine and Pathology (M.C.), University of Vermont, Burlington, Vt
Department of Epidemiology (A.B.N.), University of Pittsburgh Graduate School of Public Health and the Division of Geriatric Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pa.
Abstract
Background— A low ankle-arm index (AAI) is a strong predictor of mortality and cardiovascular events. A high AAI also appears to be associated with higher mortality risk in select populations. However, mortality and cardiovascular risk across the AAI spectrum have not been described in a more broadly defined population.
Methods and Results— We examined total and cardiovascular mortality and cardiovascular events across the AAI spectrum among 5748 participants in the Cardiovascular Health Study (CHS). The mean age of the sample population was 73±6 years, and the sample included 3289 women (57%) and 883 blacks (15%). The median duration of follow-up was 11.1 (0.1 to 12) years for mortality and 9.6 (0.1 to 12.1) years for cardiovascular events. There were 2311 deaths (953 of which were cardiovascular) and 1491 cardiovascular events during follow-up. After adjustment for potential confounders, AAI measurements 0.60 (hazard ratio [HR] 1.82, 95% CI 1.42 to 2.32), 0.61 to 0.7 (HR 2.08, 95% CI 1.61 to 2.69), 0.71 to 0.8 (HR 1.80, 95% CI 1.44 to 2.26), 0.81 to 0.9 (HR 1.73 95% CI 1.43 to 2.11), 0.91 to 1.0 (HR 1.40, 95% CI 1.20 to 1.63), and >1.40 (HR 1.57, 95% CI 1.07 to 2.31) were associated with higher mortality risk from all causes compared with the referent group (AAI 1.11 to 1.20). The pattern was similar for cardiovascular mortality. For cardiovascular events, risk was higher at all AAI levels <1 but not for AAI levels >1.4 (HR 1.00, 95% CI 0.57 to 1.74). The association of a high AAI with mortality was stronger in men than in women and in younger than in older cohort members.
Conclusions— In a cohort of community-dwelling elders, mortality risk was higher than the referent category of 1.11 to 1.2 among participants with AAI values above the traditional cutpoint of 0.9 (ie, 0.91 to 1.0 and >1.4), and the specific association of AAI with mortality varied by age and gender.
Key Words: epidemiology peripheral vascular disease mortality
Introduction
Many studies have now demonstrated that a low ankle-arm index (AAI) is strongly correlated with prevalent cardiovascular disease and is an independent predictor of incident cardiovascular events and mortality in a variety of different populations.1–12 Most prior studies focused on the predictive value of a low AAI and did not examine outcomes across the spectrum of AAI values. In fact, because of concerns about the reliability of the AAI as a measure of lower-extremity perfusion in the setting of arterial incompressibility, patients with a supranormal AAI were often excluded from the analysis of AAI with mortality.1,4,8 Recently, however, Resnick and colleagues13 reported that among Native Americans enrolled in the Strong Heart Study, the mortality risk associated with an AAI >1.4 (or arterial incompressibility) was comparable to that associated with an AAI 0.9. However, mortality risk across the AAI spectrum has not been examined in more broadly defined populations, and it is unknown whether a higher AAI is also associated with elevated risk of cardiovascular events. To address these questions, we analyzed the association of AAI with both mortality and cardiovascular events in a large cohort of community-dwelling elders.
Methods
Participants
The Cardiovascular Health Study (CHS) is an ongoing community-based, longitudinal study of Medicare-eligible adults aged 65 years and older at enrollment.14,15 The objective of the study was to evaluate risk factors for the development and progression of cardiovascular disease. An initial 5201 study participants were recruited between 1989 and 1990 from 4 US communities: Sacramento County, California; Forsyth County, North Carolina; Washington County, Maryland; and Allegheny County, Pennsylvania. An additional 687 blacks were recruited in 1992 and 1993. Follow-up interviews for events were conducted semiannually. The present analyses were conducted among the 5748 participants for whom AAI was recorded at the baseline examination.
Predictors
The primary predictor variable was the lowest leg AAI measured at study entry. The AAI is the ratio of ankle to arm systolic blood pressure. The protocol for AAI measurement in CHS has been described previously.8 Briefly, manual blood pressure measurements were obtained with a Doppler stethoscope (8 MHz, Huntleigh Technology, Inc) over the right brachial artery and both posterior tibial arteries. Duplicate blood pressure measurements were averaged for each location (right brachial and right and left posterior tibial arteries) to define an AAI for each leg. The reproducibility of duplicate blood pressure measurements in the baseline examination has been reported elsewhere.6 To evaluate the nature of the association of AAI with mortality and cardiovascular events, we categorized AAI into 0.1-U increments as follows: 0.6, 0.61 to 0.7, 0.71 to 0.8, 0.81 to 0.9, 0.91 to 1.0, 1.01 to 1.1, 1.11 to 1.2, 1.21 to 1.3, 1.31 to 1.4, and >1.4. The 1.11 to 1.2 category served as the referent for all analyses. In the primary analysis, we examined the association of the lowest leg AAI with each study outcome. In a secondary analysis, we also examined the association of the highest leg AAI with each outcome. AAI measurements were available for only 1 leg in 90 participants in the baseline examination. For these patients, we used the same AAI measurement for both analyses.
Baseline characteristics considered were age; gender; race; diabetes (defined as history of diabetes, use of hypoglycemic agent or insulin, or fasting glucose 126 mg/dL); systolic and diastolic blood pressure; use of antihypertensive medications; prior confirmed diagnoses of coronary heart disease, stroke, or congestive heart failure; smoking status (current versus past or never); body mass index; LDL and HDL cholesterol; triglycerides; serum creatinine; and C-reactive protein measured with stored baseline plasma in 1997, with a high-sensitivity ELISA developed at the CHS central blood laboratory.16,17 The analytical coefficient of variation was 6.2%.
Outcome
The primary outcome was time from enrollment to death. Secondary outcomes were, respectively, time to cardiovascular death (defined as death due to coronary heart disease, myocardial infarction, sudden cardiac death, or stroke) and time to first cardiovascular event (defined as myocardial infarction, stroke, angina, angioplasty, cardiac bypass surgery, or lower-extremity amputation/revascularization). CHS methods for surveillance and ascertainment of follow-up events have been described in detail elsewhere.14 Briefly, participants were contacted every 6 months and asked about hospitalizations and outpatient visits for specific cardiovascular diagnoses. Participant deaths and type of death were confirmed by examination of death certificates, inpatient records, nursing home or hospice records, physician questionnaires, and autopsy reports. The present analysis is based on follow-up data available through June 30, 2001.
Statistical Analysis
We examined the distribution of demographic characteristics, traditional cardiovascular risk factors, prevalent cardiovascular disease, and laboratory measures across AAI categories. We generated descriptive summary statistics for each AAI category using means (and SDs) for continuous variables and proportions for dichotomous variables, and we calculated probability values for linear trend across AAI categories.
We examined event rates across AAI categories in men and women separately because rates were substantially lower in women at all AAI levels. Associations of lowest leg AAI with all-cause and cardiovascular mortality and with cardiovascular events were assessed with unadjusted and adjusted Cox proportional hazard models. We tested for interactions with age, race, gender, and diabetes in all models. Analysis of incident cardiovascular events excluded participants with prevalent cardiovascular disease at baseline (defined as prior myocardial infarction, stroke, angina, angioplasty, bypass surgery, lower-extremity amputation, or revascularization). As a sensitivity analysis, we repeated the analysis of total mortality excluding those with prevalent cardiovascular disease or congestive heart failure at baseline. We also repeated all analyses using highest rather than lowest leg AAI as the primary predictor.
The proportional hazards assumption was tested by standard residual-based techniques. All analyses were performed with S-Plus (release 6.1, Insightful Inc) and SPSS statistical software (release 12.0.2, SPSS Inc). A probability value of <0.05 was considered statistically significant.
Results
Linear associations were noted between AAI, race, systolic pressure, smoking, body mass index, and LDL (Table 1). Several covariates demonstrated a U-shaped distribution across AAI categories (sex, use of antihypertensive medications, diabetes, congestive heart failure, coronary artery disease, stroke, and serum creatinine). HDL levels were lowest at both extremes of AAI. Women and blacks were underrepresented in the high AAI categories, with only 14 women and 6 blacks having an AAI >1.4.
A total of 53 246 person-years were available for mortality analysis, with a median follow-up duration of 11.1 years (range 0.1 to 12 years). During follow-up, there were 2311 deaths (953 cardiovascular). After exclusion of those with prevalent cardiovascular disease, a total of 37 144 years and a median of 9.6 (0.1 to 12.1) years were available for analysis. During this time, there were 1491 incident cardiovascular events.
All-cause and cardiovascular mortality and events were nonlinearly distributed across the lowest leg AAI categories. Rates were consistently highest among participants with an AAI 0.6, lowest among those with an AAI between 1.11 and 1.2, and intermediate among participants with AAI levels above this (Table 2). This pattern was most pronounced for all-cause mortality. Event rates were higher for men than for women across all AAI categories, and the upward trend in event risk above AAI values of 1.2 was more evident among men.
In unadjusted analysis, risks of all-cause and cardiovascular mortality were nonlinearly associated with lowest leg AAI (Table 3). After adjustment for confounders, lowest leg AAI measurements 1.0 and >1.4 were associated with higher risks of all-cause and cardiovascular mortality compared with the referent category (Table 3). AAI measurements 1.0 were also associated with a higher risk of fatal and nonfatal cardiovascular events, but this was not observed with measurements >1.4.
The association of AAI with all-cause mortality differed in men and women (P for interaction=0.018). An AAI >1.4 was associated with a statistically significantly higher risk of all-cause mortality in men (hazard ratio [HR] 1.78, 95% CI 1.18 to 2.68) but not in women (HR 0.67, 95% CI 0.17 to 2.71), although the accuracy of the point estimate for AAI >1.4 in women was limited by the small number of women with an AAI in this range (n=14).
An age interaction was present for both all-cause mortality (P<0.001) and cardiovascular mortality (P<0.001). Both low and high AAI values were more strongly associated with all-cause and cardiovascular mortality among cohort members <75 years old than among older cohort members, and this was most pronounced at high AAI levels. For example, the risk of death associated with an AAI >1.4 was 1.82 (95% CI 1.12 to 2.97) among participants <75 and 1.37 (95% CI 0.72 to 2.60) among those 75 years of age. Hazard ratios for all-cause mortality for an AAI 0.6 were 2.22 (95% CI 1.48 to 3.33) for those <75 years of age and 1.91 (95% CI 1.40 to 2.60) for those aged 75 years and older. There were no interactions for race or diabetes in any of the models.
Although only 66 participants had a lowest leg AAI >1.4, 178 participants had a highest leg AAI >1.4. Among these, most (n=89) had a lowest leg AAI between 1.21 and 1.4, and only 2 had an AAI 0.9. When we defined AAI categories by the highest rather than lowest leg AAI, associations with all-cause and cardiovascular mortality were stronger for low and weaker for high AAI groups (Table 4). In addition, the threshold level of AAI below which there was a statistically significant increase in mortality risk was higher (1.01 to 1.1) than for the primary analysis using lowest leg AAI. When we restricted the primary analysis of total mortality to those without baseline cardiovascular disease or congestive heart failure (n=4331), a high AAI (>1.4) was still associated with higher mortality risk (adjusted HR 1.88, 95% CI 1.23 to 2.89).
Discussion
We describe a nonlinear association between AAI and mortality in a large cohort of community-dwelling elderly men and women. When we categorized patients according to their lowest leg AAI, mortality and cardiovascular risk were higher among participants with AAI measurements 1 and >1.4 than among those in the 1.11 to 1.2 referent category (for all-cause and cardiovascular mortality). The specific association of AAI with mortality differed significantly between men and women, between older and younger patients, and depending on whether the highest or lowest leg AAI measurement was used.
These findings confirm previous reports of a nonlinear association of AAI with mortality in select populations. In the Strong Heart Study, Native Americans with an AAI >1.4 or incompressible arteries had a higher risk of all-cause and cardiovascular mortality than those with an AAI between 0.9 and 1.4.13 Similarly, among a cohort of Japanese hemodialysis patients, Ono and colleagues9 noted a higher risk of cardiovascular mortality at AAI levels below 1.1 and above 1.3 compared with the referent category of 1.1 to 1.3. Collectively, these findings suggest that because the association of AAI with mortality is nonlinear, the use of the traditional AAI cutpoint of 0.9 to define high- and low-risk groups may not allow for optimal mortality and cardiovascular risk stratification.
A nonlinear association between AAI and mortality probably exists because AAI measurements are the result of at least 2 independent lower-extremity large-vessel characteristics: perfusion and compressibility. A high AAI has traditionally been thought to reflect decreased arterial compliance associated with medial arterial calcification.18 In support of this possibility, high AAI measurements are more common among those with diabetes, in whom the prevalence of medial arterial calcification is known to be increased.19 Furthermore, lower-extremity medial arterial calcification, coronary artery calcification, and cardiac valvular calcification have all been shown to be independent predictors of mortality.10,12,19–21
Consistent with recent data from the Multi-Ethnic Study of Atherosclerosis, the distribution of baseline AAI measurements differed between men and women.22 Women were overrepresented in the 0.91 to 1.2 AAI group and underrepresented at AAI measurements above this. Differences in the association between AAI and mortality in men and women reported here may relate to gender differences in the association of AAI with calcification and with subclinical atherosclerosis in other locations.22
Differences in the association of AAI with outcomes in older and younger CHS participants were unexpected. However, consistent with this, differences across age groups in the prognostic importance of blood pressure measurements have been documented previously.23–28 Possible mechanisms for alterations in the prognostic importance of AAI with advancing age include a higher prevalence of other comorbidities at older ages (lessening the potential impact of a single condition on mortality) and the presence of a survival advantage among elderly patients with AAI measurements ordinarily associated with higher mortality risk.
Variation in the specific association of high AAI with mortality between different studies may be explained, at least in part, by differences in the demographic composition of study populations. For example, in contrast to the present analysis, in which mortality risk was clearly highest among participants with a low AAI, in the Strong Heart Study, the risk of death in the high AAI group rivaled that in the low AAI group.13 Given our observation that a high AAI was a stronger predictor of mortality in men than in women and in younger than in older CHS participants, the greater prognostic significance of a high AAI in the Strong Heart Study may be explained, at least in part, by the younger age and smaller percentage of female participants in that cohort.
Differences between study results may also reflect differences in the AAI classification method selected. In the Strong Heart Study, the high AAI group included participants with incompressible arteries and those with a high AAI in either leg. All other participants were classified on the basis of the lowest leg AAI. In the present analysis, we did not have information on arterial incompressibility, and patients were assigned to each AAI category on the basis of their lowest leg AAI (in the primary analysis) or highest leg AAI (in the secondary analysis). Thus, risk stratification based on AAI measurement should take into account the specific method of AAI classification selected.
The overall clinical importance of a high AAI is likely to vary depending not only on the prognostic importance of a given AAI measurement within the population but also with the distribution of AAI measurements in that population. The prevalence of high AAI measurements in CHS is similar to estimates for the general population29 but much lower than for the Strong Heart Study (9.25% had an AAI >1.4) and the hemodialysis cohort described by Ono and colleagues (10.9% had an AAI 1.3). These differences most likely reflect the higher prevalence of diabetes in these latter studies.
Despite strong associations with total and cardiovascular mortality, high AAI was not associated with a higher risk of cardiovascular events. Although the present analysis was underpowered to rule out such an association (we had 80% power to detect a 2.19 HR for cardiovascular events in the high AAI category), our results do demonstrate that the prognostic significance of a high AAI is greater for all-cause and cardiovascular mortality than for cardiovascular events, even among those without prevalent disease at baseline.
Limitations of the present study include the small number of participants with a high AAI, which renders inaccurate the point estimate for the association of high AAI with study outcomes in women. In addition, AAI was not recorded in 140 participants at the baseline examination. Information on why AAI was not recorded was not available for most of these participants, and criteria for noncompressibility were not applied. Thus, we were unable to measure event rates among those in whom an AAI could not be obtained owing to arterial incompressibility. A third limitation is that our results pertain specifically to community-dwelling elders and thus might not be generalizable to younger people or other elders. Finally, the AAI protocol for CHS used only right-arm blood pressure measurements. It is thus unclear how use of left-arm rather than right-arm blood pressure, or the average of right and left (arm) blood pressures, in calculating the AAI would have altered our results.
In conclusion, in this large cohort of community-dwelling elders, there was a nonlinear association between AAI and total and cardiovascular mortality. This confirms similar reports in Native Americans enrolled in the Strong Heart Study and in a large cohort of Japanese hemodialysis patients. Although AAI values 0.9 were associated with higher mortality among CHS participants, this was also true of AAI values between 0.9 and 1.0 and AAI values >1.4. Furthermore, the specific association of AAI with mortality and cardiovascular risk varied between men and women and between older and younger subjects and depended on whether we used the lowest or highest leg AAI.
Acknowledgments
Dr O’Hare is supported by a Research Career Development Award from the Department of Veterans Affairs Health Services Research and Development Service. Dr Shlipak is funded by R01 HL073208-01 and by the American Federation for Aging Research and National Institute on Aging (Paul Beeson Scholars Program) and the Robert Wood Johnson Foundation (Generalist Faculty Scholars Program). The CHS Study is supported by contracts N01-HC-85079 through N01-HC-85086, N01-HC-35129, and N01 HC-15103 from the National Heart, Lung, and Blood Institute. A full list of participating CHS investigators and institutions can be found at http://www.chs-nhlbi.org.
Disclosure
None.
References
Abbott RD, Petrovitch H, Rodriguez BL, Yano K, Schatz IJ, Popper JS, Masaki KH, Ross GW, Curb JD. Ankle/brachial blood pressure in men >70 years of age and the risk of coronary heart disease. Am J Cardiol. 2000; 86: 280–284.
Fishbane S, Youn S, Flaster E, Adam G, Maesaka JK. Ankle-arm blood pressure index as a predictor of mortality in hemodialysis patients. Am J Kidney Dis. 1996; 27: 668–672.
McDermott MM, Feinglass J, Slavensky R, Pearce WH. The ankle-brachial index as a predictor of survival in patients with peripheral vascular disease. J Gen Intern Med. 1994; 9: 445–449.
McKenna M, Wolfson S, Kuller L. The ratio of ankle and arm arterial pressure as an independent predictor of mortality. Atherosclerosis. 1991; 87: 119–128.
Murabito JM, Evans JC, Larson MG, Nieto K, Levy D, Wilson PW. The ankle-brachial index in the elderly and risk of stroke, coronary disease, and death: the Framingham Study. Arch Intern Med. 2003; 163: 1939–1942.
Newman AB, Siscovick DS, Manolio TA, Polak J, Fried LP, Borhani NO, Wolfson SK; Cardiovascular Heart Study (CHS) Collaborative Research Group. Ankle-arm index as a marker of atherosclerosis in the Cardiovascular Health Study. Circulation. 1993; 88: 837–845.
Newman AB, Tyrrell KS, Kuller LH. Mortality over four years in SHEP participants with a low ankle-arm index. J Am Geriatr Soc. 1997; 45: 1472–1478.
Newman AB, Shemanski L, Manolio TA, Cushman M, Mittelmark M, Polak JF, Powe NR, Siscovick D; the Cardiovascular Health Study Group. Ankle-arm index as a predictor of cardiovascular disease and mortality in the Cardiovascular Health Study. Arterioscler Thromb Vasc Biol. 1999; 19: 538–545.
Ono K, Tsuchida A, Kawai H, Matsuo H, Wakamatsu R, Maezawa A, Yano S, Kawada T, Nojima Y. Ankle-brachial blood pressure index predicts all-cause and cardiovascular mortality in hemodialysis patients. J Am Soc Nephrol. 2003; 14: 1591–1598.
Qiao JH, Doherty TM, Fishbein MC, Salusky IB, Luthringer DL, Fitzpatrick LA, Shah PK, Rajavashisth TB. Calcification of the coronary arteries in the absence of atherosclerotic plaque. Mayo Clin Proc. 2005; 80: 807–809.
Tsai AW, Folsom AR, Rosamond WD, Jones DW. Ankle-brachial index and 7-year ischemic stroke incidence: the ARIC study. Stroke. 2001; 32: 1721–1724.
Vliegenthart R, Oudkerk M, Hofman A, Oei HH, van Dijck W, van Rooij FJ, Witteman JC. Coronary calcification improves cardiovascular risk prediction in the elderly. Circulation. 2005; 112: 572–577.
Resnick HE, Lindsay RS, McDermott MM, Devereux RB, Jones KL, Fabsitz RR, Howard BV. Relationship of high and low ankle brachial index to all-cause and cardiovascular disease mortality: the Strong Heart Study. Circulation. 2004; 109: 733–739.
Fried LP, Borhani NO, Enright P, Furberg CD, Gardin JM, Kronmal RA, Kuller LH, Manolio TA, Mittelmark MB, Newman A. The Cardiovascular Health Study: design and rationale. Ann Epidemiol. 1991; 1: 263–276.
Tell GS, Fried LP, Hermanson B, Manolio TA, Newman AB, Borhani NO. Recruitment of adults 65 years and older as participants in the Cardiovascular Health Study. Ann Epidemiol. 1993; 3: 358–366.
Macy EM, Hayes TE, Tracy RP. Variability in the measurement of C-reactive protein in healthy subjects: implications for reference intervals and epidemiological applications. Clin Chem. 1997; 43: 52–58.
Cushman M, Arnold AM, Psaty BM, Manolio TA, Kuller LH, Burke GL, Polak JF, Tracy RP. C-reactive protein and the 10-year incidence of coronary heart disease in older men and women: the Cardiovascular Health Study. Circulation. 2005; 112: 25–31.
Maser RE, Wolfson SK Jr, Ellis D, Stein EA, Drash AL, Becker DJ, Dorman JS, Orchard TJ. Cardiovascular disease and arterial calcification in insulin-dependent diabetes mellitus: interrelations and risk factor profiles. Pittsburgh Epidemiology of Diabetes Complications Study-V. Arterioscler Thromb. 1991; 11: 958–965.
Everhart JE, Pettitt DJ, Knowler WC, Rose FA, Bennett PH. Medial arterial calcification and its association with mortality and complications of diabetes. Diabetologia. 1988; 31: 16–23.
Fox E, Harkins D, Taylor H, McMullan M, Han H, Samdarshi T, Garrison R, Skelton T. Epidemiology of mitral annular calcification and its predictive value for coronary events in African Americans: the Jackson Cohort of the Atherosclerotic Risk in Communities Study. Am Heart J. 2004; 148: 979–984.
Itani Y, Sone S, Nakayama T, Suzuki T, Watanabe S, Ito K, Takashima S, Fushimi H, Sanada H. Coronary artery calcification detected by a mobile helical computed tomography unit and future cardiovascular death: 4-year follow-up of 6120 asymptomatic Japanese. Heart Vessels. 2004; 19: 161–163.
McDermott MM, Liu K, Criqui MH, Ruth K, Goff D, Saad MF, Wu C, Homma S, Sharrett AR. Ankle-brachial index and subclinical cardiac and carotid disease: the multi-ethnic study of atherosclerosis. Am J Epidemiol. 2005; 162: 33–41.
Goodwin JS. Embracing complexity: a consideration of hypertension in the very old. J Gerontol A Biol Sci Med Sci. 2003; 58: 653–658.
Boshuizen HC, Izaks GJ, van Buuren S, Ligthart GJ. Blood pressure and mortality in elderly people aged 85 and older: community based study. BMJ. 1998; 316: 1780–1784.
Hakala SM, Tilvis RS, Strandberg TE. Blood pressure and mortality in an older population: a 5-year follow-up of the Helsinki Ageing Study. Eur Heart J. 1997; 18: 1019–1023.
Mattila K, Haavisto M, Rajala S, Heikinheimo R. Blood pressure and five year survival in the very old. Br Med J (Clin Res Ed). 1988; 296: 887–889.
Rajala S, Haavisto M, Heikinheimo R, Mattila K. Blood pressure and mortality in the very old. Lancet. 1983; 2: 520–521.
Satish S, Freeman DH Jr, Ray L, Goodwin JS. The relationship between blood pressure and mortality in the oldest old. J Am Geriatr Soc. 2001; 49: 367–374.
Resnick HE, Foster GL. Prevalence of elevated ankle-brachial index in the United States 1999 to 2002. Am J Med. 2005; 118: 676–679.(Ann M. O’Hare, MA, MD; Ronit Katz, PhD; )
Collaborative Health Studies Coordinating Center (R.K.), Seattle, Wash
General Internal Medicine Section (M.G.S.), Veterans Affairs Medical Center, San Francisco, Calif, and Departments of Medicine, Epidemiology and Biostatistics, University of California, San Francisco, Calif
Departments of Medicine and Pathology (M.C.), University of Vermont, Burlington, Vt
Department of Epidemiology (A.B.N.), University of Pittsburgh Graduate School of Public Health and the Division of Geriatric Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pa.
Abstract
Background— A low ankle-arm index (AAI) is a strong predictor of mortality and cardiovascular events. A high AAI also appears to be associated with higher mortality risk in select populations. However, mortality and cardiovascular risk across the AAI spectrum have not been described in a more broadly defined population.
Methods and Results— We examined total and cardiovascular mortality and cardiovascular events across the AAI spectrum among 5748 participants in the Cardiovascular Health Study (CHS). The mean age of the sample population was 73±6 years, and the sample included 3289 women (57%) and 883 blacks (15%). The median duration of follow-up was 11.1 (0.1 to 12) years for mortality and 9.6 (0.1 to 12.1) years for cardiovascular events. There were 2311 deaths (953 of which were cardiovascular) and 1491 cardiovascular events during follow-up. After adjustment for potential confounders, AAI measurements 0.60 (hazard ratio [HR] 1.82, 95% CI 1.42 to 2.32), 0.61 to 0.7 (HR 2.08, 95% CI 1.61 to 2.69), 0.71 to 0.8 (HR 1.80, 95% CI 1.44 to 2.26), 0.81 to 0.9 (HR 1.73 95% CI 1.43 to 2.11), 0.91 to 1.0 (HR 1.40, 95% CI 1.20 to 1.63), and >1.40 (HR 1.57, 95% CI 1.07 to 2.31) were associated with higher mortality risk from all causes compared with the referent group (AAI 1.11 to 1.20). The pattern was similar for cardiovascular mortality. For cardiovascular events, risk was higher at all AAI levels <1 but not for AAI levels >1.4 (HR 1.00, 95% CI 0.57 to 1.74). The association of a high AAI with mortality was stronger in men than in women and in younger than in older cohort members.
Conclusions— In a cohort of community-dwelling elders, mortality risk was higher than the referent category of 1.11 to 1.2 among participants with AAI values above the traditional cutpoint of 0.9 (ie, 0.91 to 1.0 and >1.4), and the specific association of AAI with mortality varied by age and gender.
Key Words: epidemiology peripheral vascular disease mortality
Introduction
Many studies have now demonstrated that a low ankle-arm index (AAI) is strongly correlated with prevalent cardiovascular disease and is an independent predictor of incident cardiovascular events and mortality in a variety of different populations.1–12 Most prior studies focused on the predictive value of a low AAI and did not examine outcomes across the spectrum of AAI values. In fact, because of concerns about the reliability of the AAI as a measure of lower-extremity perfusion in the setting of arterial incompressibility, patients with a supranormal AAI were often excluded from the analysis of AAI with mortality.1,4,8 Recently, however, Resnick and colleagues13 reported that among Native Americans enrolled in the Strong Heart Study, the mortality risk associated with an AAI >1.4 (or arterial incompressibility) was comparable to that associated with an AAI 0.9. However, mortality risk across the AAI spectrum has not been examined in more broadly defined populations, and it is unknown whether a higher AAI is also associated with elevated risk of cardiovascular events. To address these questions, we analyzed the association of AAI with both mortality and cardiovascular events in a large cohort of community-dwelling elders.
Methods
Participants
The Cardiovascular Health Study (CHS) is an ongoing community-based, longitudinal study of Medicare-eligible adults aged 65 years and older at enrollment.14,15 The objective of the study was to evaluate risk factors for the development and progression of cardiovascular disease. An initial 5201 study participants were recruited between 1989 and 1990 from 4 US communities: Sacramento County, California; Forsyth County, North Carolina; Washington County, Maryland; and Allegheny County, Pennsylvania. An additional 687 blacks were recruited in 1992 and 1993. Follow-up interviews for events were conducted semiannually. The present analyses were conducted among the 5748 participants for whom AAI was recorded at the baseline examination.
Predictors
The primary predictor variable was the lowest leg AAI measured at study entry. The AAI is the ratio of ankle to arm systolic blood pressure. The protocol for AAI measurement in CHS has been described previously.8 Briefly, manual blood pressure measurements were obtained with a Doppler stethoscope (8 MHz, Huntleigh Technology, Inc) over the right brachial artery and both posterior tibial arteries. Duplicate blood pressure measurements were averaged for each location (right brachial and right and left posterior tibial arteries) to define an AAI for each leg. The reproducibility of duplicate blood pressure measurements in the baseline examination has been reported elsewhere.6 To evaluate the nature of the association of AAI with mortality and cardiovascular events, we categorized AAI into 0.1-U increments as follows: 0.6, 0.61 to 0.7, 0.71 to 0.8, 0.81 to 0.9, 0.91 to 1.0, 1.01 to 1.1, 1.11 to 1.2, 1.21 to 1.3, 1.31 to 1.4, and >1.4. The 1.11 to 1.2 category served as the referent for all analyses. In the primary analysis, we examined the association of the lowest leg AAI with each study outcome. In a secondary analysis, we also examined the association of the highest leg AAI with each outcome. AAI measurements were available for only 1 leg in 90 participants in the baseline examination. For these patients, we used the same AAI measurement for both analyses.
Baseline characteristics considered were age; gender; race; diabetes (defined as history of diabetes, use of hypoglycemic agent or insulin, or fasting glucose 126 mg/dL); systolic and diastolic blood pressure; use of antihypertensive medications; prior confirmed diagnoses of coronary heart disease, stroke, or congestive heart failure; smoking status (current versus past or never); body mass index; LDL and HDL cholesterol; triglycerides; serum creatinine; and C-reactive protein measured with stored baseline plasma in 1997, with a high-sensitivity ELISA developed at the CHS central blood laboratory.16,17 The analytical coefficient of variation was 6.2%.
Outcome
The primary outcome was time from enrollment to death. Secondary outcomes were, respectively, time to cardiovascular death (defined as death due to coronary heart disease, myocardial infarction, sudden cardiac death, or stroke) and time to first cardiovascular event (defined as myocardial infarction, stroke, angina, angioplasty, cardiac bypass surgery, or lower-extremity amputation/revascularization). CHS methods for surveillance and ascertainment of follow-up events have been described in detail elsewhere.14 Briefly, participants were contacted every 6 months and asked about hospitalizations and outpatient visits for specific cardiovascular diagnoses. Participant deaths and type of death were confirmed by examination of death certificates, inpatient records, nursing home or hospice records, physician questionnaires, and autopsy reports. The present analysis is based on follow-up data available through June 30, 2001.
Statistical Analysis
We examined the distribution of demographic characteristics, traditional cardiovascular risk factors, prevalent cardiovascular disease, and laboratory measures across AAI categories. We generated descriptive summary statistics for each AAI category using means (and SDs) for continuous variables and proportions for dichotomous variables, and we calculated probability values for linear trend across AAI categories.
We examined event rates across AAI categories in men and women separately because rates were substantially lower in women at all AAI levels. Associations of lowest leg AAI with all-cause and cardiovascular mortality and with cardiovascular events were assessed with unadjusted and adjusted Cox proportional hazard models. We tested for interactions with age, race, gender, and diabetes in all models. Analysis of incident cardiovascular events excluded participants with prevalent cardiovascular disease at baseline (defined as prior myocardial infarction, stroke, angina, angioplasty, bypass surgery, lower-extremity amputation, or revascularization). As a sensitivity analysis, we repeated the analysis of total mortality excluding those with prevalent cardiovascular disease or congestive heart failure at baseline. We also repeated all analyses using highest rather than lowest leg AAI as the primary predictor.
The proportional hazards assumption was tested by standard residual-based techniques. All analyses were performed with S-Plus (release 6.1, Insightful Inc) and SPSS statistical software (release 12.0.2, SPSS Inc). A probability value of <0.05 was considered statistically significant.
Results
Linear associations were noted between AAI, race, systolic pressure, smoking, body mass index, and LDL (Table 1). Several covariates demonstrated a U-shaped distribution across AAI categories (sex, use of antihypertensive medications, diabetes, congestive heart failure, coronary artery disease, stroke, and serum creatinine). HDL levels were lowest at both extremes of AAI. Women and blacks were underrepresented in the high AAI categories, with only 14 women and 6 blacks having an AAI >1.4.
A total of 53 246 person-years were available for mortality analysis, with a median follow-up duration of 11.1 years (range 0.1 to 12 years). During follow-up, there were 2311 deaths (953 cardiovascular). After exclusion of those with prevalent cardiovascular disease, a total of 37 144 years and a median of 9.6 (0.1 to 12.1) years were available for analysis. During this time, there were 1491 incident cardiovascular events.
All-cause and cardiovascular mortality and events were nonlinearly distributed across the lowest leg AAI categories. Rates were consistently highest among participants with an AAI 0.6, lowest among those with an AAI between 1.11 and 1.2, and intermediate among participants with AAI levels above this (Table 2). This pattern was most pronounced for all-cause mortality. Event rates were higher for men than for women across all AAI categories, and the upward trend in event risk above AAI values of 1.2 was more evident among men.
In unadjusted analysis, risks of all-cause and cardiovascular mortality were nonlinearly associated with lowest leg AAI (Table 3). After adjustment for confounders, lowest leg AAI measurements 1.0 and >1.4 were associated with higher risks of all-cause and cardiovascular mortality compared with the referent category (Table 3). AAI measurements 1.0 were also associated with a higher risk of fatal and nonfatal cardiovascular events, but this was not observed with measurements >1.4.
The association of AAI with all-cause mortality differed in men and women (P for interaction=0.018). An AAI >1.4 was associated with a statistically significantly higher risk of all-cause mortality in men (hazard ratio [HR] 1.78, 95% CI 1.18 to 2.68) but not in women (HR 0.67, 95% CI 0.17 to 2.71), although the accuracy of the point estimate for AAI >1.4 in women was limited by the small number of women with an AAI in this range (n=14).
An age interaction was present for both all-cause mortality (P<0.001) and cardiovascular mortality (P<0.001). Both low and high AAI values were more strongly associated with all-cause and cardiovascular mortality among cohort members <75 years old than among older cohort members, and this was most pronounced at high AAI levels. For example, the risk of death associated with an AAI >1.4 was 1.82 (95% CI 1.12 to 2.97) among participants <75 and 1.37 (95% CI 0.72 to 2.60) among those 75 years of age. Hazard ratios for all-cause mortality for an AAI 0.6 were 2.22 (95% CI 1.48 to 3.33) for those <75 years of age and 1.91 (95% CI 1.40 to 2.60) for those aged 75 years and older. There were no interactions for race or diabetes in any of the models.
Although only 66 participants had a lowest leg AAI >1.4, 178 participants had a highest leg AAI >1.4. Among these, most (n=89) had a lowest leg AAI between 1.21 and 1.4, and only 2 had an AAI 0.9. When we defined AAI categories by the highest rather than lowest leg AAI, associations with all-cause and cardiovascular mortality were stronger for low and weaker for high AAI groups (Table 4). In addition, the threshold level of AAI below which there was a statistically significant increase in mortality risk was higher (1.01 to 1.1) than for the primary analysis using lowest leg AAI. When we restricted the primary analysis of total mortality to those without baseline cardiovascular disease or congestive heart failure (n=4331), a high AAI (>1.4) was still associated with higher mortality risk (adjusted HR 1.88, 95% CI 1.23 to 2.89).
Discussion
We describe a nonlinear association between AAI and mortality in a large cohort of community-dwelling elderly men and women. When we categorized patients according to their lowest leg AAI, mortality and cardiovascular risk were higher among participants with AAI measurements 1 and >1.4 than among those in the 1.11 to 1.2 referent category (for all-cause and cardiovascular mortality). The specific association of AAI with mortality differed significantly between men and women, between older and younger patients, and depending on whether the highest or lowest leg AAI measurement was used.
These findings confirm previous reports of a nonlinear association of AAI with mortality in select populations. In the Strong Heart Study, Native Americans with an AAI >1.4 or incompressible arteries had a higher risk of all-cause and cardiovascular mortality than those with an AAI between 0.9 and 1.4.13 Similarly, among a cohort of Japanese hemodialysis patients, Ono and colleagues9 noted a higher risk of cardiovascular mortality at AAI levels below 1.1 and above 1.3 compared with the referent category of 1.1 to 1.3. Collectively, these findings suggest that because the association of AAI with mortality is nonlinear, the use of the traditional AAI cutpoint of 0.9 to define high- and low-risk groups may not allow for optimal mortality and cardiovascular risk stratification.
A nonlinear association between AAI and mortality probably exists because AAI measurements are the result of at least 2 independent lower-extremity large-vessel characteristics: perfusion and compressibility. A high AAI has traditionally been thought to reflect decreased arterial compliance associated with medial arterial calcification.18 In support of this possibility, high AAI measurements are more common among those with diabetes, in whom the prevalence of medial arterial calcification is known to be increased.19 Furthermore, lower-extremity medial arterial calcification, coronary artery calcification, and cardiac valvular calcification have all been shown to be independent predictors of mortality.10,12,19–21
Consistent with recent data from the Multi-Ethnic Study of Atherosclerosis, the distribution of baseline AAI measurements differed between men and women.22 Women were overrepresented in the 0.91 to 1.2 AAI group and underrepresented at AAI measurements above this. Differences in the association between AAI and mortality in men and women reported here may relate to gender differences in the association of AAI with calcification and with subclinical atherosclerosis in other locations.22
Differences in the association of AAI with outcomes in older and younger CHS participants were unexpected. However, consistent with this, differences across age groups in the prognostic importance of blood pressure measurements have been documented previously.23–28 Possible mechanisms for alterations in the prognostic importance of AAI with advancing age include a higher prevalence of other comorbidities at older ages (lessening the potential impact of a single condition on mortality) and the presence of a survival advantage among elderly patients with AAI measurements ordinarily associated with higher mortality risk.
Variation in the specific association of high AAI with mortality between different studies may be explained, at least in part, by differences in the demographic composition of study populations. For example, in contrast to the present analysis, in which mortality risk was clearly highest among participants with a low AAI, in the Strong Heart Study, the risk of death in the high AAI group rivaled that in the low AAI group.13 Given our observation that a high AAI was a stronger predictor of mortality in men than in women and in younger than in older CHS participants, the greater prognostic significance of a high AAI in the Strong Heart Study may be explained, at least in part, by the younger age and smaller percentage of female participants in that cohort.
Differences between study results may also reflect differences in the AAI classification method selected. In the Strong Heart Study, the high AAI group included participants with incompressible arteries and those with a high AAI in either leg. All other participants were classified on the basis of the lowest leg AAI. In the present analysis, we did not have information on arterial incompressibility, and patients were assigned to each AAI category on the basis of their lowest leg AAI (in the primary analysis) or highest leg AAI (in the secondary analysis). Thus, risk stratification based on AAI measurement should take into account the specific method of AAI classification selected.
The overall clinical importance of a high AAI is likely to vary depending not only on the prognostic importance of a given AAI measurement within the population but also with the distribution of AAI measurements in that population. The prevalence of high AAI measurements in CHS is similar to estimates for the general population29 but much lower than for the Strong Heart Study (9.25% had an AAI >1.4) and the hemodialysis cohort described by Ono and colleagues (10.9% had an AAI 1.3). These differences most likely reflect the higher prevalence of diabetes in these latter studies.
Despite strong associations with total and cardiovascular mortality, high AAI was not associated with a higher risk of cardiovascular events. Although the present analysis was underpowered to rule out such an association (we had 80% power to detect a 2.19 HR for cardiovascular events in the high AAI category), our results do demonstrate that the prognostic significance of a high AAI is greater for all-cause and cardiovascular mortality than for cardiovascular events, even among those without prevalent disease at baseline.
Limitations of the present study include the small number of participants with a high AAI, which renders inaccurate the point estimate for the association of high AAI with study outcomes in women. In addition, AAI was not recorded in 140 participants at the baseline examination. Information on why AAI was not recorded was not available for most of these participants, and criteria for noncompressibility were not applied. Thus, we were unable to measure event rates among those in whom an AAI could not be obtained owing to arterial incompressibility. A third limitation is that our results pertain specifically to community-dwelling elders and thus might not be generalizable to younger people or other elders. Finally, the AAI protocol for CHS used only right-arm blood pressure measurements. It is thus unclear how use of left-arm rather than right-arm blood pressure, or the average of right and left (arm) blood pressures, in calculating the AAI would have altered our results.
In conclusion, in this large cohort of community-dwelling elders, there was a nonlinear association between AAI and total and cardiovascular mortality. This confirms similar reports in Native Americans enrolled in the Strong Heart Study and in a large cohort of Japanese hemodialysis patients. Although AAI values 0.9 were associated with higher mortality among CHS participants, this was also true of AAI values between 0.9 and 1.0 and AAI values >1.4. Furthermore, the specific association of AAI with mortality and cardiovascular risk varied between men and women and between older and younger subjects and depended on whether we used the lowest or highest leg AAI.
Acknowledgments
Dr O’Hare is supported by a Research Career Development Award from the Department of Veterans Affairs Health Services Research and Development Service. Dr Shlipak is funded by R01 HL073208-01 and by the American Federation for Aging Research and National Institute on Aging (Paul Beeson Scholars Program) and the Robert Wood Johnson Foundation (Generalist Faculty Scholars Program). The CHS Study is supported by contracts N01-HC-85079 through N01-HC-85086, N01-HC-35129, and N01 HC-15103 from the National Heart, Lung, and Blood Institute. A full list of participating CHS investigators and institutions can be found at http://www.chs-nhlbi.org.
Disclosure
None.
References
Abbott RD, Petrovitch H, Rodriguez BL, Yano K, Schatz IJ, Popper JS, Masaki KH, Ross GW, Curb JD. Ankle/brachial blood pressure in men >70 years of age and the risk of coronary heart disease. Am J Cardiol. 2000; 86: 280–284.
Fishbane S, Youn S, Flaster E, Adam G, Maesaka JK. Ankle-arm blood pressure index as a predictor of mortality in hemodialysis patients. Am J Kidney Dis. 1996; 27: 668–672.
McDermott MM, Feinglass J, Slavensky R, Pearce WH. The ankle-brachial index as a predictor of survival in patients with peripheral vascular disease. J Gen Intern Med. 1994; 9: 445–449.
McKenna M, Wolfson S, Kuller L. The ratio of ankle and arm arterial pressure as an independent predictor of mortality. Atherosclerosis. 1991; 87: 119–128.
Murabito JM, Evans JC, Larson MG, Nieto K, Levy D, Wilson PW. The ankle-brachial index in the elderly and risk of stroke, coronary disease, and death: the Framingham Study. Arch Intern Med. 2003; 163: 1939–1942.
Newman AB, Siscovick DS, Manolio TA, Polak J, Fried LP, Borhani NO, Wolfson SK; Cardiovascular Heart Study (CHS) Collaborative Research Group. Ankle-arm index as a marker of atherosclerosis in the Cardiovascular Health Study. Circulation. 1993; 88: 837–845.
Newman AB, Tyrrell KS, Kuller LH. Mortality over four years in SHEP participants with a low ankle-arm index. J Am Geriatr Soc. 1997; 45: 1472–1478.
Newman AB, Shemanski L, Manolio TA, Cushman M, Mittelmark M, Polak JF, Powe NR, Siscovick D; the Cardiovascular Health Study Group. Ankle-arm index as a predictor of cardiovascular disease and mortality in the Cardiovascular Health Study. Arterioscler Thromb Vasc Biol. 1999; 19: 538–545.
Ono K, Tsuchida A, Kawai H, Matsuo H, Wakamatsu R, Maezawa A, Yano S, Kawada T, Nojima Y. Ankle-brachial blood pressure index predicts all-cause and cardiovascular mortality in hemodialysis patients. J Am Soc Nephrol. 2003; 14: 1591–1598.
Qiao JH, Doherty TM, Fishbein MC, Salusky IB, Luthringer DL, Fitzpatrick LA, Shah PK, Rajavashisth TB. Calcification of the coronary arteries in the absence of atherosclerotic plaque. Mayo Clin Proc. 2005; 80: 807–809.
Tsai AW, Folsom AR, Rosamond WD, Jones DW. Ankle-brachial index and 7-year ischemic stroke incidence: the ARIC study. Stroke. 2001; 32: 1721–1724.
Vliegenthart R, Oudkerk M, Hofman A, Oei HH, van Dijck W, van Rooij FJ, Witteman JC. Coronary calcification improves cardiovascular risk prediction in the elderly. Circulation. 2005; 112: 572–577.
Resnick HE, Lindsay RS, McDermott MM, Devereux RB, Jones KL, Fabsitz RR, Howard BV. Relationship of high and low ankle brachial index to all-cause and cardiovascular disease mortality: the Strong Heart Study. Circulation. 2004; 109: 733–739.
Fried LP, Borhani NO, Enright P, Furberg CD, Gardin JM, Kronmal RA, Kuller LH, Manolio TA, Mittelmark MB, Newman A. The Cardiovascular Health Study: design and rationale. Ann Epidemiol. 1991; 1: 263–276.
Tell GS, Fried LP, Hermanson B, Manolio TA, Newman AB, Borhani NO. Recruitment of adults 65 years and older as participants in the Cardiovascular Health Study. Ann Epidemiol. 1993; 3: 358–366.
Macy EM, Hayes TE, Tracy RP. Variability in the measurement of C-reactive protein in healthy subjects: implications for reference intervals and epidemiological applications. Clin Chem. 1997; 43: 52–58.
Cushman M, Arnold AM, Psaty BM, Manolio TA, Kuller LH, Burke GL, Polak JF, Tracy RP. C-reactive protein and the 10-year incidence of coronary heart disease in older men and women: the Cardiovascular Health Study. Circulation. 2005; 112: 25–31.
Maser RE, Wolfson SK Jr, Ellis D, Stein EA, Drash AL, Becker DJ, Dorman JS, Orchard TJ. Cardiovascular disease and arterial calcification in insulin-dependent diabetes mellitus: interrelations and risk factor profiles. Pittsburgh Epidemiology of Diabetes Complications Study-V. Arterioscler Thromb. 1991; 11: 958–965.
Everhart JE, Pettitt DJ, Knowler WC, Rose FA, Bennett PH. Medial arterial calcification and its association with mortality and complications of diabetes. Diabetologia. 1988; 31: 16–23.
Fox E, Harkins D, Taylor H, McMullan M, Han H, Samdarshi T, Garrison R, Skelton T. Epidemiology of mitral annular calcification and its predictive value for coronary events in African Americans: the Jackson Cohort of the Atherosclerotic Risk in Communities Study. Am Heart J. 2004; 148: 979–984.
Itani Y, Sone S, Nakayama T, Suzuki T, Watanabe S, Ito K, Takashima S, Fushimi H, Sanada H. Coronary artery calcification detected by a mobile helical computed tomography unit and future cardiovascular death: 4-year follow-up of 6120 asymptomatic Japanese. Heart Vessels. 2004; 19: 161–163.
McDermott MM, Liu K, Criqui MH, Ruth K, Goff D, Saad MF, Wu C, Homma S, Sharrett AR. Ankle-brachial index and subclinical cardiac and carotid disease: the multi-ethnic study of atherosclerosis. Am J Epidemiol. 2005; 162: 33–41.
Goodwin JS. Embracing complexity: a consideration of hypertension in the very old. J Gerontol A Biol Sci Med Sci. 2003; 58: 653–658.
Boshuizen HC, Izaks GJ, van Buuren S, Ligthart GJ. Blood pressure and mortality in elderly people aged 85 and older: community based study. BMJ. 1998; 316: 1780–1784.
Hakala SM, Tilvis RS, Strandberg TE. Blood pressure and mortality in an older population: a 5-year follow-up of the Helsinki Ageing Study. Eur Heart J. 1997; 18: 1019–1023.
Mattila K, Haavisto M, Rajala S, Heikinheimo R. Blood pressure and five year survival in the very old. Br Med J (Clin Res Ed). 1988; 296: 887–889.
Rajala S, Haavisto M, Heikinheimo R, Mattila K. Blood pressure and mortality in the very old. Lancet. 1983; 2: 520–521.
Satish S, Freeman DH Jr, Ray L, Goodwin JS. The relationship between blood pressure and mortality in the oldest old. J Am Geriatr Soc. 2001; 49: 367–374.
Resnick HE, Foster GL. Prevalence of elevated ankle-brachial index in the United States 1999 to 2002. Am J Med. 2005; 118: 676–679.(Ann M. O’Hare, MA, MD; Ronit Katz, PhD; )