The Effect of Age on Hypertension Control and Management

AJH

2006; 19:520 –527

Therapeutics

The Effect of Age on Hypertension Control and Management Ann M. Borzecki, Mark E. Glickman, Boris Kader, and Dan R. Berlowitz Background: Despite guidelines recommending similar blood pressure (BP) treatment goals regardless of age, controversy exists regarding treating those ⱖ80 years of age. Whether this affects current practice in terms of differences in BP control and number of prescribed antihypertensives by age is unknown. Methods: This was a cross-sectional study of 59,207 outpatients with hypertension treated at 10 Veterans Health Administration sites. Outcome measures were BP control (⬍140/90 mm Hg) and number of antihypertensive medications at the patient’s last study visit. Uncontrolled BP was also categorized by whether systolic, diastolic, or both were elevated. Results: Subjects 40 to 49 years and those 50 to 59 years of age had better BP control (adjusted odds ratios 1.35 [95% CI ⫽ 1.26 to 1.44] and 1.22 [CI ⫽ 1.17 to 1.28] respectively) compared with subjects 60 to 69 years of age; those 70 to 79 years of age and ⱖ80 years had worse control (OR ⫽ 0.92 for both; respective CIs ⫽ 0.88 to 0.96 and 0.86 to 0.99). Antihypertensive medication use increased by successive decade to age 80 years, after which

the trend reversed. Adjusted mean number of medications by age were: ⬍40 years, 2.60; 40 to 49, 2.82; 50 to 59, 2.91; 60 to 69, 3.01; 70 to 79, 3.03; ⱖ80 years, 2.90 (P ⬍ .05 in pairwise comparisons). The trend of number of medications by age did not vary across hypertension categories, despite systolic hypertension increasing and diastolic hypertension decreasing with age. Subjects ⬍40 years of age were taking the fewest medications, followed by subjects ⱖ80 years and then by those 40 to 49, 50 to 59, 70 to 79, and 60 to 69 years of age. Conclusions: The oldest hypertension patients, despite worse BP control, are being treated less aggressively with fewer medications than their younger counterparts (those 60 to 79 years of age). Our results suggest that current controversy in treating the oldest hypertensive patients is having an impact on actual practice. Am J Hypertens 2006;19:520 –527 © 2006 American Journal of Hypertension, Ltd. Key Words: Hypertension, guidelines, treatment, outcome assessment.

ypertension affects 29% of adults in the United States.1 Prevalence increases progressively with age, from 7% among individuals 18 to 39 years of age to 65% among those ⱖ60 years.1 Multiple clinical trials, including those of elderly patients with isolated systolic hypertension, have demonstrated the cardiovascular benefits of lower blood pressure (BP).2,3 Based on such trials, the last two Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure (JNC) reports have recommended controlling BP to at least ⬍140/90 mm Hg for all patients regardless of age.4,5 Notwithstanding the importance of treating hypertension and of dissemination of these national guidelines, many patients— especially elderly individuals— continue to have

H

inadequate BP control.1,6 In the 1999 to 2000 National Health and Nutrition Examination Survey (NHANES IV), only 31% of the population had a BP ⬍140/90.1 Control rates were lowest in the oldest age group both among treated and untreated subjects. Control rates are likely worst among elderly individuals, partly because BP is more difficult to manage.7 However, despite explicit JNC recommendations, considerable controversy exists regarding how or whether to treat hypertension in those ⱖ80 years and whether benefits outweigh risks.8,9 The existing clinical trials in the elderly population have had relatively few patients ⱖ80 years; and more recent data, including a meta-analysis examining this older group, plus preliminary results of a randomized trial in those ⬎80 years, have raised further doubts as to

Received March 22, 2005. First decision September 26, 2005. Accepted October 1, 2005. From the Center for Health Quality, Outcomes and Economic Research, Bedford Veterans Affairs Hospital, Bedford, Massachusetts; Department of Health Services, Boston University School of Public Health, Boston, Massachusetts.

This project was funded in part by grant SDR 99-300-1 from the the Department of Veterans Affairs Health Services Research and Development service. Address correspondence and reprint requests to Dr. Ann M. Borzecki, CHQOER, Bedford VA Hospital (152), 200 Springs Road, Bedford, MA 01730; e-mail: [email protected]

0895-7061/06/$32.00 doi:10.1016/j.amjhyper.2005.10.022

© 2006 by the American Journal of Hypertension, Ltd. Published by Elsevier Inc.

AJH–May 2006 –VOL. 19, NO. 5

the benefit of treating hypertension in the so-called “very old,” with a trend toward higher all-cause mortality despite fewer cardiovascular events.2,3,10,11 Reflecting this uncertainty, surveys using hypothetical clinical scenarios have found that clinicians report treating elderly patients less aggressively and tolerating higher than recommended BPs.12,13 However we do not know how these issues translate into actual practice. Are older hypertensive individuals actually being treated less aggressively with fewer medications than their younger counterparts? We addressed this question by examining: 1) whether BP control rates differ by age among veterans with hypertension; 2) the association between age and number of prescribed antihypertensive medications, a proxy for treatment aggressiveness; and 3) the pattern of hypertension by age (eg, isolated systolic or diastolic hypertension), and determining whether this affects aggressiveness of prescribing practices.

Methods Study Population We identified outpatients from 10 Veterans Health Administration (VA) sites across the country through 2000 using the VA’s National Patient Care Database (NPCD). Eligible subjects: 1) had at least two NPCD-listed hypertension diagnoses, ICD-9-CM code 401, between July 1, 1997, and June 30, 1999; and 2) were regular VA users (ie, had three or more NPCD-listed visits to a general medicine or subspecialty medical clinic between July 1, 1999, and December 31, 2000). The first visit during this period was considered as the index visit. The study protocol was approved by the Bedford VA Hospital’s institutional review board. Data Collection Data sources were the Veterans Health Information Systems and Technology Architecture (VISTA), the VA’s electronic record system, and the NPCD. VISTA, maintained at each site, contains multiple files including clinical data such as vital signs, laboratory results, pharmacy records, and provider notes. (We previously found the vitals file very complete, with provider notes contributing minimal additional BP information.14) VISTA also contains diagnoses and procedure information from all outpatient visits and inpatient stays; this information is transferred to a central data repository and incorporated into the NPCD.15 Demographic characteristics and comorbidities were obtained from the NPCD (1998 to 2000 inclusive). Hypertension-related comorbidities (HTN-RC), conditions associated with either more aggressive treatment (diabetes mellitus, renal disease, congestive heart failure, and coronary artery, cerebrovascular, and peripheral vascular disease), or conditions for which treatment of the primary

AGE AND HYPERTENSION MANAGEMENT

521

condition also treats hypertension (atrial fibrillation, benign prostatic hypertrophy), were identified by the presence of specified ICD-9-CM codes in the 12 months preceding and including the index visit.16 Height, weight, and BP were obtained from VISTA vitals, and medications from VISTA pharmacy files (July 1, 1999, to December 31, 2000). Statistical Analyses Outcome measures were BP control (⬍140/90 mm Hg, yes/no) and number of antihypertensive medications at each patient’s last study visit. We initially examined bivariate associations between BP control and age via logistic regression and ␹2 tests as appropriate, using age first as a continuous variable and then as a categorical variable categorized by decade, to determine the most appropriate variable form. We then used logistic regression and adjusted for characteristics associated with control: ethnicity, gender, body mass index (BMI), presence of diabetes and renal disease at the index visit (both yes/no), and number of antihypertensive medications at the last visit.7,17 To determine the appropriate BMI form to use, we also examined its association with BP control, (as well as with number of antihypertensive medications), both as a continuous and categorical variable.18 Analysis of BMI categories suggested a negative linear association with BP control (and a positive linear association with medication number). Thus we used a continuous BMI variable in multivariate analyses. To examine the robustness of BP model findings, we tested for the inclusion of additional HTN-RC and their two-way interactions with age, plus the interaction between antihypertensive medications and age. Next we examined the association between number of antihypertensive medications and age. For unadjusted analyses we examined the mean number of medications among all age groups using one-way analysis of variance (ANOVA) and then compared medication numbers between successive age groups using t tests. We performed this for the full sample and both controlled and uncontrolled subgroups. We also compared medication use between controlled and uncontrolled groups in a given age category using t tests. We next examined the association between age and medication number using linear regression, controlling for ethnicity, gender, BMI, presence of specific HTN-RC, and BP control at the last visit. To account for conditions such as malignancy that might result in less aggressive treatment, we performed an analysis including a “modified” Charlson index that subtracted out cardiovascular conditions, renal disease, and diabetes,19 as these were already in the model and were expected to have the opposite effect on antihypertensive medication prescribing practices. We also included interactions between age and: HTN-RC, Charlson index, and BP control. We compared adjusted pairwise age effects using the Tukey multiple comparison procedure (␣ ⫽ 0.05). Given the skewness of the medica-

522

AJH–May 2006 –VOL. 19, NO. 5

AGE AND HYPERTENSION MANAGEMENT

tion data, we also examined the robustness of our findings by log transforming the medication variable, log(1⫹ [number of medications]). Given the significance of BP control in regression models, we also performed separate analyses for controlled/uncontrolled subgroups (minus the BP control variable). Finally we examined the pattern of hypertension among patients with poor control (BP ⱖ140/90 mm Hg). We divided “poor control” into four categories: 1) isolated systolic hypertension (systolic BP ⱖ160 mm Hg and diastolic ⱕ90 mm Hg); 2) “borderline” isolated systolic hypertension (systolic BP 140 to 159 mm Hg and diastolic ⱕ90 mm Hg); 3) diastolic hypertension (systolic BP ⬍140 and diastolic ⱖ90 mm Hg); or 4) combined hypertension (systolic BP ⱖ140 and diastolic ⱖ90 mm Hg).20 First we compared each category of hypertension (yes/no) across age groups using 2 ⫻ 6 ␹2 tests in four separate analyses. Next we examined the mean number of medications across the four hypertension categories using one-way ANOVA,

and performed pairwise comparisons using the Tukey test (␣ ⫽ 0.05). Lastly we examined the effect of age on medication numbers within each hypertension category using one-way ANOVA and the Tukey test to examine means between age groups. All analyses were performed using SAS software, version 8.02 (SAS Institute Inc., Cary, NC).

Results Baseline sample characteristics are shown in Table 1. The mean age of the subjects was 65 years, and ⬎97% of the subjects were male. The majority were of white ethnicity, and in 24% the ethnicity was not identified in the VA databases. Overall 40.8% of subjects had BP ⬍140/90 mm Hg, and most were on at least two antihypertensive medications. The mean number of medical visits with a BP measurement was 5.95 ⫾ 4.49 over the 18-month study period.

Table 1. Baseline characteristics of hypertensive patient sample Characteristic Age, mean (SD), years Age by decade, no. (%), years ⬍40 (range 17–39) 40–49 50–59 60–69 70–79 ⱖ 80 Male sex, no. (%) Ethnicity, no. (%) Non-Hispanic white Non-Hispanic African American Hispanic Other* Unknown Body mass index, mean (SD) kg/m2 Blood pressure at last visit, mean (SD), mm Hg Systolic Diastolic Overall control (⬍140/90 mm Hg) Number of antihypertensive medications at last visit Mean (SD) No. (%) 0 1 2 ⱖ3 Selected co-existing conditions, no. (%) Diabetes mellitus Renal disease Coronary artery disease Cerebrovascular disease Congestive heart failure Peripheral vascular disease Atrial fibrillation Benign prostatic hypertrophy All characteristics are as noted at index visit unless otherwise specified. * Includes Asians and Native Americans.

N ⴝ 59,207 65.1 (11.1) 683 4,855 12,247 17,732 19,163 4,304 57,513

(1.2) (8.2) (21.1) (29.9) (32.4) (7.3) (97.1)

34,407 8,475 1,393 550 14,382 28.2

(58.1) (14.3) (2.4) (0.9) (24.3) (5.5)

144 (21) 77 (13) 40.8% 2.3 (1.2) 1,632 14,912 19,844 22,819

(2.8) (25.2) (33.5) (38.5)

19,385 3,677 20,175 7,008 6,042 5,002 3,970 11,444

(32.7) (6.2) (34.1) (11.8) (10.2) (8.4) (6.7) (19.3)

AJH–May 2006 –VOL. 19, NO. 5

AGE AND HYPERTENSION MANAGEMENT

Percent BP<140/90

50%

45%

40%

35% <40 (683)

40-49* (4,855)

50-59* (12,470)

60-69* (17,732)

70-79* (19,163)

>/=80 (4,304)

Age (n)

FIG. 1. Blood pressure control by age. Number of subjects in each age category is shown. Error bars represent 95% confidence intervals. *Unadjusted proportion of those controlled is significantly different (P ⬍ .05) in successive decade comparisons using 2 ⫻ 2 ␹2 analyses, (nonoverlapping confidence intervals), except for subjects ⱖ80 years v 70 to 79 years of age.

Age and Control of BP In bivariate models, increasing age was associated with lower prevalence of BP control, although categorization by decade revealed a nonlinear association (Fig. 1). Thus we used a categorical age variable in final analyses. By decade, control was lowest in subjects 70 to 79 years of age at 38.7%, and highest in those 40 to 49 years at 45.6%. From age 50 to 59 years onward, the proportion of subjects in which BP was controlled dropped significantly compared to the prior decade, with each subsequent decade of age, except for those ⱖ80 years of age (proportions were not significantly different between those 70 to 79 and ⱖ80 years at 38.7% and 39.6% respectively; P ⫽ .25). Figure 2 shows mean systolic BP of the study subjects by age. Adjusted model results were similar to bivariate findings. Subjects 40 to 49 years and those 50 to 59 years of age had significantly better control, whereas subjects 70 to 79 years and ⱖ80 years had worse control rates than those 60 to 69 years of age (Table 2). Inclusion of additional HTN-RC strengthened the age-related associations, with little change seen by adding interaction terms despite a significant age by congestive heart failure interaction and age by number of medications interaction (Table 2).

523

years (P ⬍ .001 for all 2 ⫻ 6 [presence/absence of condition ⫻ age group] ␹2 analyses). However, after adjustment for HTN-RC and BP control, subjects ⱖ80 years were on significantly fewer medications (adjusted least squares mean 2.90 ⫾ 1.85) than those 60 to 69 years (3.01 ⫾ 3.36) or 70 to 79 years of age (3.03 ⫾ 3.37, P ⬍ .001) (Table 3). As with unadjusted results, medication use increased significantly with each successive decade up to age 80 years (P ⬍ .05 for pairwise comparisons). Adjusted mean medication numbers for the remaining age groups were as follows: ⬍40 years, 2.60 ⫾ 1.31; 40 to 49 years, 2.82 ⫾ 2.05; 50 to 59 years, 2.91 ⫾ 2.94) Inclusion of the “modified” Charlson index and interaction terms for age by comorbidities and by BP control in the models strengthened the observed association (Table 3), with decreased medication use in the oldest age group. In addition subjects 70 to 79 years of age were now on significantly fewer medications than those 60 to 69 years (respective adjusted means with increasing age: 3.10, 2.98, 2.78). Log transformation of the number of medications yielded similar results. This age relationship also held among controlled and uncontrolled subsamples (data not shown for subsamples). Pattern of Hypertension and Number of Antihypertensive Medications Examining the pattern of hypertension by age, as expected, isolated systolic hypertension became more prevalent with increasing age, and isolated diastolic hypertension became less prevalent (Fig. 4). In all age groups, borderline isolated systolic hypertension (systolic BP 140 to 159, diastolic BP ⬍90) was the most common finding. Subjects with isolated systolic hypertension were on the most medications (mean 2.53 ⫾ 1.31), whereas subjects with isolated diastolic hypertension were on the fewest (2.01 ⫾ 1.09, P ⬍ .001) (Table 4). The pattern of medication numbers by age did not vary appreciably based on hypertension category (data not shown). Subjects ⬍40 years of age were on the fewest medications, followed by those ⱖ80 years of age, and then by those 40 to 49, 50 to 59, 70 to 79, and 60

Number of Antihypertensive Medications by Age 147 146 144 143

146

146

70-79

>/=80

145

145

mm Hg

Examining medication use, uncontrolled subjects were on significantly more medications than controlled subjects (mean ⫾ SD: 2.35 ⫾ 1.25 v 2.24 ⫾ 1.20, P ⬍ .001). This association held across all age groups up to age 70 (Fig. 3). Medication use increased by successive decade up to age 80 years for the entire sample, and for controlled and uncontrolled subsamples. After age 80 years the trend reversed; those ⱖ80 years of age were on fewer medications than individuals in the preceding decade (P ⬍ .001). The prevalence of HTN-RC increased with each successive age group (as did the modified Charlson index score; data not shown but available from authors), except for diabetes, for which prevalence was highest (36.2%) in subjects 60 to 69 years of age and was 25.1% in those ⱖ80

142

142

142 141

141 140 139 138 137 <40

40-49

50-59

60-69 Age

FIG. 2. Mean systolic blood pressure by age. *Error bars represent 95% confidence intervals. *Unadjusted mean blood pressures; means are significantly different in successive decades (P ⬍ .05 for all t test comparisons), nonoverlapping confidence intervals.

524

AJH–May 2006 –VOL. 19, NO. 5

AGE AND HYPERTENSION MANAGEMENT

Table 2. Unadjusted and adjusted blood pressure control by age Unadjusted odds ratio (95% CI)*

Variable Age (referent: 60–69 years)§ ⬍40 40–49 50–59 70–79 ⱖ80 Sex (referent: male) Female Ethnicity (referent: non-Hispanic white) Non-Hispanic African American Hispanic Other Unknown Body mass index Diabetes (referent: No) Renal disease (referent: No) Antihypertensive medications (per 1-unit increase)

1.01 1.26 1.18 0.95 0.99

(0.87–1.19) (1.19–1.35)储 (1.13–1.24)储 (0.91–0.99)储 (0.92–1.06)

Base model odds ratio (95% CI)† 1.08 1.33 1.21 0.93 0.93

(0.92–1.27) (1.25–1.42)储 (1.15–1.27)储 (0.89–0.97)储 (0.87–0.99)储

Final model odds ratio (95% CI)‡ 1.40 1.58 1.44 0.81 0.74

(1.03–1.91)储 (1.38–1.82)储 (1.30–1.60)储 (0.73–0.89)储 (0.63–0.87)储

0.95 (0.86–1.05)

1.02 (0.92–1.12)

0.75 1.08 1.08 0.78 0.99 0.99 1.04

0.80 1.05 1.14 0.85 0.99 0.93 0.81

(0.71–0.79)储 (0.91–1.28) (0.91–1.28) (0.75–0.81)储 (0.98–0.99)储 (0.95–1.02) (0.97–1.12)

0.94 (0.93–0.96)

(0.76–0.84)储 (0.95–1.18) (0.96–1.35) (0.81–0.88)储 (0.98–0.99)储 (0.87–0.99)储 (0.71–0.93)储

0.90 (0.88–0.92)储

CI ⫽ confidence interval. * Odds ratios indicate likelihood of having blood pressure ⬍140/90 associated with a given age group compared to referent, persons 60–69 years of age. Unadjusted proportion of persons whose blood pressure was controlled also significantly different (P ⬍ .05) in successive decade comparisons using 2 ⫻ 2 ␹2 analyses, (confidence intervals nonoverlapping), except for subjects ⱖ80 years v 70–79 years of age; † Base multivariate logistic regression model adjusted for sex, ethnicity (white, African American Hispanic, other, unknown), body mass index (kg/m2), presence of diabetes mellitus (yes/no) and renal disease (yes/no), and number of antihypertensive medications at the last visit; ‡ Full model also includes adjustment for coronary artery disease, cerebrovascular disease, congestive heart failure, peripheral vascular disease, atrial fibrillation, and benign prostatic hypertrophy, as well as interactions between age and hypertension-related conditions, and between age and number of antihypertensive medications. Age*congestive heart failure and age*number of medications were significant in the model; § Subjects 60–69 years of age used as reference group for logistic models because mean sample age was 65 years; 储 Significantly different from referent.

to 69 years. This ordering held for all except borderline isolated systolic hypertension: subjects 40 to 49 years were on fewer medications than those ⱖ80 years (P ⬍ .05), and subjects 60 to 69 years of age were on fewer medications than those 70 to 79 years, although for this

Mean # Medications

2.6 2.4 2.2 2.0

Controlled Uncontrolled

1.8 1.6 1.4 1.2 <40

40-49

50-59

60-69

70-79

>/=80

Age

FIG. 3. Comparison of number of medications by age group and blood pressure (BP) control. Means shown with surrounding bars representing 95% confidence intervals. Confidence intervals overlap between lines for those ⬍40 and those ⱖ70, however, by Student t test, means of controlled versus uncontrolled for those ⬍40 are significantly different (P ⫽ .02). Comparing mean medications between successive decades, there is a significant difference between decades within subsamples (P ⬍ .001 for all t test comparisons), except between 60 and 69 years and between 70 and 79 years for the uncontrolled subsample.

latter comparison the differences were not significant across any of the hypertension categories.

Discussion In this large sample of hypertension patients, BP control was significantly associated with age, although this association was nonlinear. The proportion of subjects whose BP was controlled was highest in middle-aged group (40 to 59 years) and was lowest in the youngest and oldest groups. Medication use also varied with age, increasing up to age 79 years then decreasing in subjects ⱖ80 years, despite the greater HTN-RC burden in these oldest patients. Patients with uncontrolled BP were on more medications than their controlled age counterparts, likely because they had more resistant hypertension. However this relationship held only up to age 69 years; in persons ⱖ80 years the trend actually started to reverse with uncontrolled subjects on fewer medications than controlled subjects. The association between older age and fewer prescribed medications was even stronger after adjustment for conditions that should result in more aggressive prescribing, as well as adjustment for other age-associated comorbidities that were associated with less aggressive prescribing practices. Consistent with other studies, most

AJH–May 2006 –VOL. 19, NO. 5

AGE AND HYPERTENSION MANAGEMENT

525

Table 3. Models of predictors of number of antihypertensive medications (full sample) Variable

Base model coefficients*

P value

⫺0.30 ⫺0.09 ⫺0.01 0.11 0.13

⬍.001 ⬍.001 .74 ⬍.001 ⬍.001

0.37 1.06 0.83 0.71 0.39

.68 ⬍.001 ⬍.001 ⬍.001 ⬍.001

⫺0.15

⬍.001

⫺0.15

⬍.001

0.23 ⫺0.01 0.03 ⫺0.01 0.03

⬍.001 .67 .52 .33 ⬍.001

0.22 ⫺0.02 0.03 ⫺0.02 0.03

⬍.001 .50 .63 .31 ⬍.001

0.16 0.15 0.11 0.26 0.52 0.26 0.13 0.52 ⫺0.14

⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001

0.23 0.12 0.09 0.15 0.49 0.20 0.15 0.33 ⫺0.14 ⫺0.08

⬍.001 .002 .07 ⬍.001 ⬍.001 ⬍.001 .006 ⬍.001 ⬍.001 ⬍.001

Age (referent: 80⫹ years)‡ ⬍40 40–49 50–59 60–69 70–79 Sex (referent: male) Female Ethnicity (referent: non-Hispanic white) Non-Hispanic African American Hispanic Other Unknown Body mass index Hypertension-related comorbidities Atrial fibrillation (ref: No) Benign Prostatic hypertrophy (ref: No) Cerebrovascular disease (ref: No) Coronary artery disease (ref: No) Congestive heart failure (ref: No) Diabetes mellitus (ref: No) Peripheral vascular disease (ref: No) Renal disease (ref: No) Blood pressure control (ref: No) Modified Charlson index

Final model coefficients†

P value

Ref ⫽ referent. * Base model includes adjustment for sex, ethnicity, body mass index, presence of individual hypertension related comorbidities and blood pressure (BP) control; † Final model also includes adjustment for the modified Charlson index and interactions between age and: hypertension-related conditions, Charlson index, and BP control. Age*coronary artery disease, age*cerebrovascular disease, age*congestive heart failure, age*diabetes, age*renal disease and age*BP control were significant in the model; ‡ The same associations held when the model was run on sample with controlled BP and sample with uncontrolled BP. In pairwise comparisons there was a significant difference between decades (P ⬍ .05) with each successive decade for the whole sample and the uncontrolled subsample. For the controlled subsample there was no significant difference between subjects 60–69 and those ⱖ80 years of age.

patients with uncontrolled BP had borderline systolic hypertension, with isolated systolic BP elevation more common with increasing age.20,21 The pattern of hypertension did not alter the age-related medication trend. Our results demonstrate that the so-called “very old” individuals with hypertension, despite having worse BP control, higher systolic hypertension rates, and more cardiovascular comorbidity, are in fact being treated less

aggressively than the “young elderly,” whereas the socalled “young elderly” are being treated more aggressively than younger subjects. Thus these findings suggest that the current controversy in BP treatment of the “very old” individuals is having an impact on actual practice. There are few comparable studies. These mostly address BP control, use fewer age groupings and don’t specifically distinguish those aged 80 and older from younger subjects. No other studies compare medication use by age or by pattern of BP control. Hajar et al, using

100% 90%

Isolated diastolic HTN, <140/>/=90 mm Hg

80% 70%

Combined systolic & diastolic HTN, >/=140/>/=90 mm Hg

60% 50%

Table 4. Number of medications by pattern of hypertension*

Isolated systolic HTN, >/=160/<90 mm Hg

40% 30%

Borderline isolated systolic HTN, 140159/<90 mm Hg

20% 10% 0% <40

40-49 50-59

60-69

70-79

>/=80

Overall

Age

FIG. 4. Pattern of hypertension in those with blood pressure (BP) ⱖ140/90 by age. *HTN ⫽ hypertension. *P ⬍ .001 for all 2 ⫻ 6 ␹2 analyses (eg, elevated BP ⱖ160/⬍90, yes/no by age groupings).

Blood pressure

N

ⱖ160/⬍90 mm Hg 140–159/⬍90 mm Hg ⱖ140/ⱖ90 mm Hg ⬍140/ⱖ90 mm Hg

8,001 18,074 7,893 994

Mean no. of medications (SD) 2.53 2.28 2.36 2.01

(1.31) (1.20) (1.28) (1.09)

* In pairwise comparisons all groups were significantly different from each other (P ⬍ .05).

526

AJH–May 2006 –VOL. 19, NO. 5

AGE AND HYPERTENSION MANAGEMENT

data from the fourth National Health and Nutrition Examination Survey (NHANES IV) and three age groups, similarly found a lower unadjusted proportion controlled in younger (age 18 to 39 years) and older (ⱖ60 years) groups compared with the middle age group (40 to 59 years), accounting for either all subjects with hypertension or, more comparable to our study, for all treated subjects (respective proportions with increasing age were 52%, 66% and 44%).1 Using these same age categories, our proportions were 40%, 44%, and 39%. A Women’s Health Initiative analysis found that among hypertensive women 50 to 79 years, those 70 to 79 years of age were more likely to have uncontrolled BP but were as likely to receive BP medications as those 50 to 59 years of age.22 That study did not otherwise examine treatment aggressiveness. The only study to address BP control expressly in individuals ⱖ80 years of age is published only as an abstract. Of 8735 patients seen in a hypertension clinic examined by 20-year spans, the proportion controlled was highest in the youngest age group (20 to 39 years) at 75% and decreased with increasing age.23 Proportions were 45% in the 60 to 79 and 30% in the group ⱖ80 years. Our findings are strengthened by the very large sample size: almost tenfold the number of NHANES IV subjects and sixfold the number in either NHANES III phase.1 We also have a much larger elderly population than was represented in NHANES. Thus we were able to compare many more age groups. This study adds to the relative paucity of epidemiologic age-related data with respect to BP control and medication use. Although some of the unadjusted absolute differences were small, the consistency and strengthening of these observed differences when accounting for additional important covariates suggests that such differences are real. Clinicians are treating the so-called “very old” hypertensive patients differently from “young elderly” patients. They are prescribing fewer medications in the oldest group despite more uncontrolled BP, higher systolic BP, and more HTN-RC than in the “young elderly” and younger age groups. This has important implications, inasmuch as randomized trials and observational studies demonstrate that even small BP differences can have a relatively large impact on cardiovascular morbidity and mortality, especially in subjects with existing target-organ damage.24,25 This study was performed in a sample of predominantly male veterans with good access to medical care and medications. Therefore findings may not be generalizable to other settings. However our age-related BP control trends were similar to other populations.1,22,23 Our aggressiveness-of-therapy measure is based on number of medications and does not capture dosage. We also lack information on medications dispensed outside the VA. This is probably a rare occurrence, as VA pharmacy benefits are much more generous than most drug plans (medications are available free or with a small co-payment) and unlikely to be an age-specific issue. We lack medication side effect and adherence information. It is possible that increased side

effects in very elderly individuals might be accounting for less aggressive treatment, although if this were the case one might expect older patients to be taking lower doses of more drugs. With regard to adherence, providers may either not be adding medications because they are trying to improve adherence to a current regimen or, conversely, they may be adding medications because they do not realize that a patient is not taking medications as prescribed. It is unclear why the first scenario as opposed to the second would be more likely to apply to elderly patients. Of note, studies have found elderly veterans (ⱖ65 years) more adherent to medication regimens than younger patients, although this has not been specifically addressed among very old patients.26 Our data predates the most recent JNC report and publication of two large antihypertensive clinical trials.5,27,28 Although both trials included subjects ⱖ80 years, they were designed to determine optimal first-line therapy, not whether older patients should receive differential treatment. Furthermore the last two JNC reports (published in 1997 and 2003) have recommended treating older hypertensive patients to goals identical to those for younger patients.4,5 The JNC 7 also emphasizes that most patients, regardless of age, will require more than one antihypertensive drug to achieve control.5 In our study using data through the end of 2000, the majority of patients across all age groups were already taking at least two medications. Thus we would not expect substantially different findings with more current data. Improving hypertension care and BP control is a national priority, both inside and outside the VA.29 It is a common condition, the prevalence of which is increasing as our population ages. The oldest age groups with lower control rates are at highest risk for catastrophic events such as myocardial infarction and stroke. Although guidelines recommend treating all patients the same regardless of age, the evidence for this in terms of benefit in very old patients is less solid.30 It is likely at least in part because of this uncertainty, as well as concern about medication side effects, that clinicians are treating these patients less aggressively than younger patients. We, and the rest of the medical community, eagerly await the final results of the Hypertension in the Very Elderly Trial to help resolve this dilemma.11

Acknowledgment The authors especially thank Marshall Goff for his editorial assistance.

References 1.

2.

Hajjar I, Kotchen TA: Trends in prevalence, awareness, treatment, and control of hypertension in the United States, 1988 –2000. J Am Med Assoc 2003;290:199 –206. SHEP Cooperative Research Group: Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension. Final results of the Systolic Hypertension in the Elderly Program (SHEP). J Am Med Assoc 1991;265:3255–3264.

AJH–May 2006 –VOL. 19, NO. 5

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

Amery A, Birkenhager W, Brixko P, Bulpitt C, Clement D, Deruyttere M, De Schaepdryver A, Dollery C, Fagard R, Forette F, Hamdy R, Joossens JV, Lund-Johansen P, Petrie J, Tuomilehto J, Williams B: Mortality and morbidity results from the European Working Party on High Blood Pressure in the Elderly trial. Lancet 1985;1: 1349 –1354. Joint National Committee: The sixth report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. Arch Intern Med 1997;157:2413–2446. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL Jr, Jones DW, Materson BJ, Oparil S, Wright JT Jr, Roccella EJ: The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. J Am Med Assoc 2003;289:2560 –2572. Berlowitz DR, Ash AS, Hickey EC, Friedman RH, Glickman M, Kader B, Moskowitz MA: Inadequate management of blood pressure in a hypertensive population. N Engl J Med 1998;339:1957– 1963. Cushman WC, Ford CE, Cutler JA, Margolis KL, Davis BR, Grimm RH, Black HR, Hamilton BP, Holland J, Nwachuku C, Papademetriou V, Probstfield J, Wright JT Jr, Alderman MH, Weiss RJ, Piller L, Bettencourt J, Walsh SM: Success and predictors of blood pressure control in diverse North American settings: the antihypertensive and lipid-lowering treatment to prevent heart attack trial (ALLHAT). J Clin Hypertens 2002;4:393– 405. Goodwin JS: Embracing complexity: a consideration of hypertension in the very old. J Gerontol A Biol Sci Med Sci 2003;58:653– 658. Aronow WS, Denson S, Hajjar I, Harris TB, Lowenthal DT, Michel JP, Grab B, Perrenoud JJ, Newman AB, Thomas DR: Commentary on “Embracing complexity: a consideration of hypertension in the very old.”J Gerontol A Biol Sci Med Sci 2003;58:659 – 668. Gueyffier F, Bulpitt C, Boissel JP, Schron E, Ekbom T, Fagard R, Casiglia E, Kerlikowske K, Coope J: Antihypertensive drugs in very old people: a subgroup meta-analysis of randomised controlled trials. INDANA Group. Lancet 1999;353:793–796. Bulpitt CJ, Beckett NS, Cooke J, Dumitrascu DL, Gil-Extremera B, Nachev C, Nunes M, Peters R, Staessen JA, Thijs L: Results of the pilot study for the Hypertension in the Very Elderly Trial. J Hypertens 2003;21:2409 –2417. Hyman DJ, Pavlik VN: Self-reported hypertension treatment practices among primary care physicians: blood pressure thresholds, drug choices, and the role of guidelines and evidence-based medicine. Arch Intern Med 2000;160:2281–2286. Ekpo EB, Shah IU, Fernando MU, White AD: Isolated systolic hypertension in the elderly: survey of practitioners’ attitude and management. Gerontology 1993;39:207–214. Borzecki AM, Wong AT, Hickey EC, Ash AS, Berlowitz DR: Can we use automated data to assess quality of hypertension care? Am J Manag Care 2004;10:473– 479. VA Information Resource Center: Toolkit for new users of VA data. Available at: http://www.virec.research.med.va.gov/Support/TrainingNewUsersToolkit/Toolkit.htm. (accessed March 10, 2005). Borzecki AM, Wong AT, Hickey EC, Ash AS, Berlowitz DR:

AGE AND HYPERTENSION MANAGEMENT

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29. 30.

527

Identifying hypertension-related comorbidities from administrative data: what’s the optimal approach? Am J Med Qual 2004;19: 201–206. Berlowitz DR, Ash AS, Hickey EC, Kader B, Friedman RH, Moskowitz MA: Profiling outcomes of ambulatory care: casemix affects perceived performance. Med Care 1998;36:928 –933. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. WHO Technical Report Series, No. 894. Geneva: World Health Organization; 2000. Charlson ME, Pompei P, Ales KL, MacKenzie CR: A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 1987;40:373–383. Psaty BM, Manolio TA, Smith NL, Heckbert SR, Gottdiener JS, Burke GL, Weissfeld J, Enright P, Lumley T, Powe N, Furberg CD: Time trends in high blood pressure control and the use of antihypertensive medications in older adults: the Cardiovascular Health Study. Arch Intern Med 2002;162:2325–2332. Borzecki AM, Wong AT, Hickey EC, Ash AS, Berlowitz DR: Hypertension control: how well are we doing? Arch Intern Med 2003;163:2705–2711. Wassertheil-Smoller S, Anderson G, Psaty BM, Black HR, Manson J, Wong N, Francis J, Grimm R, Kotchen T, Langer R, Lasser N: Hypertension and its treatment in postmenopausal women: baseline data from the Women’s Health Initiative. Hypertension 2000;36: 780 –789. Graves JW, Darby CH, Bailey KR, Sheps SG: The effect of increasing age on hypertension control rates. Am J Hypertens 2001;14: A232. Lewington S, Clarke R, Qizilbash N, Peto R, Collins R: Agespecific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet 2002;360:1903–1913. Ogden LG, He J, Lydick E, Whelton PK: Long-term absolute benefit of lowering blood pressure in hypertensive patients according to the JNC VI risk stratification. Hypertension 2000;35:539 – 543. Ren XS, Kazis LE, Lee A, Zhang H, Miller DR: Identifying patient and physician characteristics that affect compliance with antihypertensive medications. J Clin Pharm Ther 2002;27:47–56. ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group: Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). J Am Med Assoc 2002;288:2981–2997. Wing LM, Reid CM, Ryan P, Beilin LJ, Brown MA, Jennings GL, Johnston CI, McNeil JJ, Macdonald GJ, Marley JE, Morgan TO, West MJ, Group ftSANBPS: A comparison of outcomes with angiotensin-converting-enzyme inhibitors and diuretics for hypertension in the elderly. N Engl J Med 2003;348:583–592. Chobanian AV: Control of hypertension—an important national priority. N Engl J Med 2001;345:534 –535. Chaudhry SI, Krumholz HM, Foody JM: Systolic hypertension in older persons. J Am Med Assoc 2004;292:1074 –1080.