Abstract

Objective: to determine the prevalence of orthostatic hypotension (OH) and associations with medication use in community-dwelling older women.

Design: cross-sectional analysis using data from the British Women’s Heart and Health Study.

Setting: general practices in 23 towns in the UK.

Participants: 3,775 women aged 60–80 years from 1999 to 2001.

Main outcome measure: orthostatic hypotension—drop of ≥20 mmHg in systolic and/or a drop of ≥10 mmHg in diastolic blood pressure on standing.

Results: prevalence of OH was 28% (95% confidence interval [CI] 26.6, 29.4), which increased with age and hypertension. Regardless of treatment status or diagnosed hypertension, raised blood pressure was strongly associated with OH (P < 0.001). OH was strongly associated with number of antihypertensives taken (none vs three or more: odds ratio [OR] 2.24, 95% CI 1.47–3.40, P < 0.001); the association was slightly attenuated after allowing for age and co-morbidities (OR 1.99; 95% CI 1.30, 3.05; P = 0.003). Women with multiple co-morbidities had markedly increased odds of OH independent of age, number and type of medications taken (none vs four or more diagnoses: OR 2.28, 95% CI 1.58–3.30, P = 0.005).

Conclusion: uncontrolled hypertension, use of three or more antihypertensives and multiple co-morbidities are predictors of OH in older women. Detection or monitoring of OH in these groups may prevent women from suffering its adverse consequences.

Introduction

Compensatory mechanisms are required in order to regulate blood pressure in old age. Baroreflexes, which control heart rate and vascular resistance, decline in efficiency with ageing, resulting in a tendency for systolic blood pressure and pulse pressure to fall upon standing [1, 2]. A reduction in systolic blood pressure of at least 20 mmHg and/or diastolic blood pressure of at least 10 mmHg within 3 min of standing is defined as orthostatic hypotension (OH) [3].

The reported prevalence of OH varies between 5% in community residents and 60% in people in institutions and acute-care settings [4]. OH is more prevalent in elderly people [2, 5], diabetic [6, 7] and hypertensive individuals [5–9]. An increased risk of falls [10], cerebrovascular events [6, 11] and mortality [2, 8, 12] have been reported in patients with OH.

Medications have been implicated in the development of OH, in particular anti-hypertensive agents [4, 10]. Polypharmacy refers to the prescription of more drugs than is clinically justified [13], and this occurs frequently in older adults. Presentation to the emergency department is more likely the greater the number of medications used and the older the patient [14], and a fall due to OH may be the precipitating factor. We examined the association between OH and patterns of medication use in community-dwelling older women.

Methods

The British Women’s Heart and Health Study (BWHHS) methodology has been fully described elsewhere [15]. Briefly, between 1999 and 2001, a cohort of 4,286 women was recruited from general practice lists in 23 nationally representative UK towns. Participants attended an interview where they were asked about diagnosed diseases, and were also asked to bring in all their current medications so that these could be recorded. Women completed a questionnaire collecting behavioural and lifestyle data, including smoking habit, alcohol consumption and indicators of socio-economic position. Their blood pressure levels were measured at the time of the interview, with two sitting measurements followed by two standing measurements (all at 1-min intervals). OH was defined as a drop of ≥10 mmHg in diastolic blood pressure and/or a drop of ≥20 mmHg in systolic blood pressure on standing (based on the differences between the first sitting and fourth standing measurements, within 3 min of standing). These measurements were taken in the morning for all patients who were also fasting. Participants were advised to take any regular medications on the day of examination.

Disease status was assessed by interview at baseline, self-completed questionnaire and by general practitioner record review. A diagnosis of coronary heart disease (CHD), stroke or diabetes was accepted using any of these methods. ‘Evidence of hypertension’ included women identified as hypertensive from self-reported doctor diagnosis of hypertension (regardless of blood pressure at medical examination) plus those without a self-reported diagnosis who had high blood pressure at medical exam. High blood pressure (BP) was defined as a systolic pressure of >140 mmHg and/or diastolic pressure of >90 mmHg according to British Hypertension Guidelines [16] as well as other international bodies [17, 18]. In addition, three groups were differentiated: ‘controlled hypertensives’, that is, known hypertension (self report) with normal blood pressure ≤140/90 mmHg on antihypertensive medication; ‘uncontrolled hypertensives’, that is, elevated BP (>140 mmHg systolic pressure and/or >90 mmHg diastolic pressure) and on antihypertensive medication; and ‘normotensives’. Circulatory disease was defined as any diagnosis of myocardial infarction, angina, stroke, transient ischaemic attack, aortic artery disease or peripheral artery disease. Chronic obstructive pulmonary disease (COPD) included asthma and chronic bronchitis, and eye disease included cataracts and glaucoma [15]. Diabetes was defined using the World Health Organization criteria [19] of a fasting blood glucose ≥7 mmol/L or report of a diagnosis of diabetes.

Analyses were confined to 3,775 women (88% of total women) who had blood pressure measurements made at examination and with information on medications (both prescribed and bought over the counter) collected at interview by a trained nurse. Characteristics of women, including disease status and medications taken, were compared according to presence or absence of OH. Odds ratios (OR) of OH were calculated for the different medications grouped according to class of drug using the British National Formulary categories [20] adjusted for age alone, for other medications and additionally for disease and lifestyle variables. These were calculated using conditional logistic regression analyses, adjusted for town of residence to account for the clustered sampling strategy used in recruitment.

This study has approval from the 23 local research ethics committees covering our study population. All women gave signed informed consent at baseline.

Results

The characteristics of women with and without OH are shown in Table 1. There was a higher prevalence of OH amongst women who had evidence of hypertension (79% vs 64%, P < 0.001), but no associations were seen with coronary heart disease, circulatory disease, diabetes, chronic obstructive pulmonary disease, cancers and other chronic diseases.

Table 1

Characteristics of women with or without OH. Values are percentages of women (number of women) by risk factors, disease status and medications taken

 Without orthostatic hypotension (n = 2,716) With orthostatic hypotension (n = 1,059) P value adjusted for agea 
Risk factors:    
Age: mean (SD) 68.4 (5.4) 69.4 (5.5)  
Body mass index: mean (SD) 27.6 (4.8) 27.2 (4.8) 0.06 
Alcohol (over 14 units per week), % 5.9 (161) 4.1 (43) 0.03 
Manual social class, % 53.4 (1,452) 53.4 (566) 0.79 
Current smoker, % 11.4 (311) 9.4 (100) 0.14 
Disease status, %:    
Evidence of hypertensionb 64.1 (1,740) 79.2 (839) <0.001 
Arthritis 46.5 (1,264) 44.9 (476) 0.36 
Any circulatory disease 25.6 (695) 26.1 (276) 0.57 
Chronic obstructive lung disease 22.3 (606) 22.9 (243) 0.35 
Depression 15.4 (419) 15.7 (166) 0.47 
Coronary heart disease 15.6 (424) 16.0 (170) 0.69 
Eye disease 13.6 (370) 14.9 (148) 0.85 
Thyroid disease 10.3 (281) 10.9 (115) 0.76 
Diabetes 8.8 (239) 10.9 (115) 0.14 
Cancer 8.7 (236) 9.0 (95) 0.82 
Medication, %:    
On any BP-lowering medication 33.8 (919) 40.2 (426) <0.001 
Beta blockers 12.8 (347) 18.2 (193) <0.001 
Thiazide diuretic 12.9 (351) 16.2 (172) 0.008 
Calcium channel blockers 10.7 (291) 10.9 (115) 0.90 
Angiotensin-converting enzyme inhibitors 8.8 (238) 11.0 (116) 0.04 
Non-thiazide diuretic 8.5 (231) 9.7 (103) 0.25 
Nitrates 7.8 (213) 6.6 (70) 0.20 
Alpha-adrenoceptor-blocking drugs 1.4 (37) 2.5 (26) 0.02 
Central nervous system 28.5 (773) 26.1 (276) 0.14 
Endocrine system 21.3 (579) 22.6 (239) 0.40 
Antiplatelet drugs 13.5 (367) 14.5 (154) 0.40 
Respiratory system 11.2 (305) 12.6 (133) 0.20 
Lipid-lowering drugs 7.9 (215) 8.6 (91) 0.50 
Any other drug 31.2 (848) 30.8 (326) 0.80 
 Without orthostatic hypotension (n = 2,716) With orthostatic hypotension (n = 1,059) P value adjusted for agea 
Risk factors:    
Age: mean (SD) 68.4 (5.4) 69.4 (5.5)  
Body mass index: mean (SD) 27.6 (4.8) 27.2 (4.8) 0.06 
Alcohol (over 14 units per week), % 5.9 (161) 4.1 (43) 0.03 
Manual social class, % 53.4 (1,452) 53.4 (566) 0.79 
Current smoker, % 11.4 (311) 9.4 (100) 0.14 
Disease status, %:    
Evidence of hypertensionb 64.1 (1,740) 79.2 (839) <0.001 
Arthritis 46.5 (1,264) 44.9 (476) 0.36 
Any circulatory disease 25.6 (695) 26.1 (276) 0.57 
Chronic obstructive lung disease 22.3 (606) 22.9 (243) 0.35 
Depression 15.4 (419) 15.7 (166) 0.47 
Coronary heart disease 15.6 (424) 16.0 (170) 0.69 
Eye disease 13.6 (370) 14.9 (148) 0.85 
Thyroid disease 10.3 (281) 10.9 (115) 0.76 
Diabetes 8.8 (239) 10.9 (115) 0.14 
Cancer 8.7 (236) 9.0 (95) 0.82 
Medication, %:    
On any BP-lowering medication 33.8 (919) 40.2 (426) <0.001 
Beta blockers 12.8 (347) 18.2 (193) <0.001 
Thiazide diuretic 12.9 (351) 16.2 (172) 0.008 
Calcium channel blockers 10.7 (291) 10.9 (115) 0.90 
Angiotensin-converting enzyme inhibitors 8.8 (238) 11.0 (116) 0.04 
Non-thiazide diuretic 8.5 (231) 9.7 (103) 0.25 
Nitrates 7.8 (213) 6.6 (70) 0.20 
Alpha-adrenoceptor-blocking drugs 1.4 (37) 2.5 (26) 0.02 
Central nervous system 28.5 (773) 26.1 (276) 0.14 
Endocrine system 21.3 (579) 22.6 (239) 0.40 
Antiplatelet drugs 13.5 (367) 14.5 (154) 0.40 
Respiratory system 11.2 (305) 12.6 (133) 0.20 
Lipid-lowering drugs 7.9 (215) 8.6 (91) 0.50 
Any other drug 31.2 (848) 30.8 (326) 0.80 
a

P values are derived from conditional logistic regression of OH on age and each specified variable in turn.

b

Comprises of doctor diagnosis of hypertension and those without doctor diagnosis who had high blood pressure readings at baseline medical examination (>140 mmHg systolic and/or >90 mmHg diastolic pressure).

The prevalence of OH was 28% (95% confidence intervals [CI] 26.6, 29.4) amongst our women aged 60–80 years. Systolic OH was more frequent than diastolic OH (20.4% vs 12.4%). This prevalence varied substantially by age and by presence of diagnosed hypertension, from 21% amongst women aged 60–64 without hypertension to 32% amongst women aged 75–80 with hypertension. See the table Appendix 1 in the supplementary data on the journal website http://www.ageing.oxfordjournals.org/. The effects of blood pressure on the odds of OH do not differ by age group (P for interaction = 0.32). Overall, the prevalence of OH was 24% higher in women who reported a diagnosis of hypertension. The prevalence of OH in this diagnosed hypertension group was fairly similar between those who received blood-pressure-lowering treatment and those who did not (33% vs 29%). However, in the treatment group, OH was significantly higher among the ‘uncontrolled’ hypertensives (38% vs 21%, P value for difference <0.001). Among women who reported no diagnosis of hypertension, OH was nearly twice as prevalent (33% 95% CI 30.6, 36.1 vs 18% 95% CI 16.2, 20.6) in those who were categorised as ‘undiagnosed hypertensives’ compared to those who were ‘normotensives’ (P < 0.001).

Higher odds of OH were observed amongst women taking any blood-pressure-lowering medication (age-adjusted OR 1.26, 95% CI 1.09–1.47, P < 0.01). However, only evidence of associations between OH and beta blockers remained after adjustment for age, co-morbidity and health behaviours (OR 1.58; 95% CI 1.19, 2.09; P < 0.01; Table 2). There was evidence that taking nitrates reduced the odds of OH after adjusting for age and other drugs (OR 0.70; 95% CI 0.51, 0.96; P < 0.05). This association, however, was attenuated after adjusting for diseases and lifestyle variables (P = 0.08).

Table 2

Relationship between type of drug used and OH in 3,749 UK womena aged 60–80 years. Values are adjusted odds ratio of OH (95% CI)

 Adjusted for age Adjusted for age and for other drugs listed Additionally adjusted for disease and lifestyle variablesb 
On any BP-lowering medication 1.26** (1.09, 1.47) 0.95 (0.70, 1.30) 0.80 (0.58, 1.10) 
Beta blockers 1.47*** (1.23, 1.81) 1.59** (1.21, 2.10) 1.58** (1.19, 2.09) 
Thiazide diuretic 1.25* (1.02, 1.53) 1.18 (0.90, 1.55) 1.16 (0.88, 1.52) 
Calcium channel blockers 0.96 (0.76, 1.21) 0.95 (0.72, 1.27) 0.95 (0.71, 1.27) 
Angiotensin-converting enzyme inhibitors 1.27* (1.00, 1.61) 1.26 (0.94, 1.69) 1.23 (0.92, 1.66) 
Non-thiazide diuretic 1.07 (0.83, 1.37) 1.09 (0.84, 1.43) 1.13 (0.86, 1.49) 
Nitrates 0.77 (0.58, 1.02) 0.70* (0.51, 0.96) 0.72 (0.49, 1.05) 
Alpha-adrenoceptor-blocking drugs 1.81* (1.08, 3.03) 1.59 (0.92, 2.72) 1.51 (0.88, 2.58) 
Drugs for central nervous system 0.86 (0.73, 1.02) 0.86 (0.72, 1.01) 0.86 (0.72, 1.03) 
Drugs for endocrine system 1.14 (0.96, 1.35) 1.12 (0.94, 1.33) 1.13 (0.92, 1.38) 
Drugs for respiratory system 1.14 (0.91, 1.42) 1.24 (0.99, 1.56) 1.21 (0.93, 1.56) 
Antiplatelet drugs 0.96 (0.78, 1.19) 0.90 (0.71, 1.14) 0.88 (0.68, 1.13) 
Lipid-lowering drugs 1.10 (0.85, 1.43) 1.05 (0.79, 1.41) 1.00 (0.74, 1.34) 
Any other drugs 0.94 (0.81, 1.10) 0.94 (0.79, 1.11) 0.93 (0.78, 1.11) 
 Adjusted for age Adjusted for age and for other drugs listed Additionally adjusted for disease and lifestyle variablesb 
On any BP-lowering medication 1.26** (1.09, 1.47) 0.95 (0.70, 1.30) 0.80 (0.58, 1.10) 
Beta blockers 1.47*** (1.23, 1.81) 1.59** (1.21, 2.10) 1.58** (1.19, 2.09) 
Thiazide diuretic 1.25* (1.02, 1.53) 1.18 (0.90, 1.55) 1.16 (0.88, 1.52) 
Calcium channel blockers 0.96 (0.76, 1.21) 0.95 (0.72, 1.27) 0.95 (0.71, 1.27) 
Angiotensin-converting enzyme inhibitors 1.27* (1.00, 1.61) 1.26 (0.94, 1.69) 1.23 (0.92, 1.66) 
Non-thiazide diuretic 1.07 (0.83, 1.37) 1.09 (0.84, 1.43) 1.13 (0.86, 1.49) 
Nitrates 0.77 (0.58, 1.02) 0.70* (0.51, 0.96) 0.72 (0.49, 1.05) 
Alpha-adrenoceptor-blocking drugs 1.81* (1.08, 3.03) 1.59 (0.92, 2.72) 1.51 (0.88, 2.58) 
Drugs for central nervous system 0.86 (0.73, 1.02) 0.86 (0.72, 1.01) 0.86 (0.72, 1.03) 
Drugs for endocrine system 1.14 (0.96, 1.35) 1.12 (0.94, 1.33) 1.13 (0.92, 1.38) 
Drugs for respiratory system 1.14 (0.91, 1.42) 1.24 (0.99, 1.56) 1.21 (0.93, 1.56) 
Antiplatelet drugs 0.96 (0.78, 1.19) 0.90 (0.71, 1.14) 0.88 (0.68, 1.13) 
Lipid-lowering drugs 1.10 (0.85, 1.43) 1.05 (0.79, 1.41) 1.00 (0.74, 1.34) 
Any other drugs 0.94 (0.81, 1.10) 0.94 (0.79, 1.11) 0.93 (0.78, 1.11) 
*

P < 0.05.

**

P < 0.01.

***

P < 0.001.

a

Restricted to 3,749 women with medication, OH, disease and lifestyle variables available.

b

Disease and lifestyle variables adjusted for: high blood pressure, any circulatory disease, CHD (myocardial infarction or angina), diabetes, cancer, COPD, arthritis, thyroid disease, eye disease, depression, heavy alcohol drinker (daily or ≥14 units/week), social class and smoking status.

The odds of OH did not increase with the overall number of drugs being taken. See the table Appendix 2 in the supplementary data on the journal website http://www.ageing.oxfordjournals.org/. However, the odds of OH did increase in those on multiple antihypertensive drugs (none vs three or more: crude OR 2.24, 95% CI 1.47–3.40, P < 0.001). This association was modestly attenuated by adjustment for age (OR 2.09, 95% CI 1.37–3.18, P < 0.001) and number of diseases (OR 1.99, 95% CI 1.30–3.05, P = 0.003). To determine whether the number of drugs or the number of co-morbid diseases was the more important factor, an analysis of odds ratios for OH by number of co-morbidities was carried out (Table 3). In age-adjusted analyses, women with an increasing number of diseases (from none up to four or more diseases) had increased odds of having OH (OR 2.40; 95% CI 1.70, 3.38; P = 0.007). This association remained when adjusted for both age and number of drugs taken (OR 2.28, 95% CI 1.58–3.30, P = 0.005). We also examined potential adverse outcomes associated with OH. OH was not significantly associated with falls (age-adjusted OR 0.96; 95% CI 0.79, 1.17; P = 0.70) or cerebrovascular events (age-adjusted OR 1.15; 95% CI 0.88, 1.50; P = 0.30). However, OH was strongly associated with all-cause mortality over the 7.6 years of follow-up since baseline blood pressure measurement (age-adjusted OR 1.10; 95% CI 1.07, 1.14; P < 0.001).

Table 3

Relationship between number of diseases and OH. Values are crude and adjusted OR of OH (95% CI)

Number of diseasesa Odds ratios (95% CI)
 
% with OH n = 3,780 Crude Adjusted for age Adjusted for age and number of drugs 
16.4 292 1.00 1.00 1.00 
27.3 938 1.82 (1.19, 2.78) 1.68 (1.10, 2.58) 1.84 (1.31, 2.60) 
28.5 1,005 2.09 (1.37, 3.18) 1.87 (1.22, 2.85) 1.95 (1.38, 2.74) 
28.5 740 2.26 (1.47, 3.47) 1.94 (1.26, 3.00) 1.93 (1.35, 2.77) 
≥4 32.0 805 3.03 (1.99, 4.62) 2.48 (1.62, 3.80) 2.28 (1.58, 3.30) 
P valueb 0.0001  <0.001 <0.001 0.005 
Number of diseasesa Odds ratios (95% CI)
 
% with OH n = 3,780 Crude Adjusted for age Adjusted for age and number of drugs 
16.4 292 1.00 1.00 1.00 
27.3 938 1.82 (1.19, 2.78) 1.68 (1.10, 2.58) 1.84 (1.31, 2.60) 
28.5 1,005 2.09 (1.37, 3.18) 1.87 (1.22, 2.85) 1.95 (1.38, 2.74) 
28.5 740 2.26 (1.47, 3.47) 1.94 (1.26, 3.00) 1.93 (1.35, 2.77) 
≥4 32.0 805 3.03 (1.99, 4.62) 2.48 (1.62, 3.80) 2.28 (1.58, 3.30) 
P valueb 0.0001  <0.001 <0.001 0.005 
a

Number of diseases includes diabetes, any circulatory disease, coronary heart disease, high blood pressure, thyroid disease, eye disease, arthritis, cancer, depression and COPD.

b

P value from Mann–Whitney U test for percent with OH; P value for trend for ORs from logistic regression analysis.

Discussion

The prevalence of OH in British women was high when compared to other community elderly study populations. Associations with increasing age [2, 5], hypertension [5, 7–9] and death [2, 8, 12] were confirmed. Use of beta blockers and three or more antihypertensives were independently associated with OH. Multiple co-morbidities and uncontrolled hypertension were also strong predictors of OH in older women. No association of OH with cerebrovascular events, diabetes or falls was found, failing to support earlier findings [6, 7, 10, 11].

The risk of OH was higher in women on antihypertensives, as shown in smaller studies [7, 10], and rose with the number of antihypertensive drugs taken. However, contrary to expectations, there was no evidence to suggest that polypharmacy was associated with OH. Given the increased emphasis on achieving blood pressure control through use of at least two antihypertensives, our findings indicate the need for clinicians to be cautious and provide close monitoring of symptoms and signs of OH when multiple agents are used in hypertensive older adults. In frail elderly people, efforts to achieve optimal therapeutic control should be balanced with consideration of their complex and multiple co-morbidities, which render them at greater risk of OH and its consequences. In contrast to other studies [5, 6, 21], OH was strongly associated with the number of co-morbidities but not with each individual co-morbidity. This was independent of the number of drugs taken. This may be partly explained by the higher number of co-morbid diseases in our older participants compared with those in other studies. This association may relate to pathophysiological changes such as peripheral sympathetic failure or atherosclerosis with diabetes [7] as well as other age-associated diseases that impair autonomic function [22].

OH is associated with hypertension-related alterations in blood pressure regulation. With increasing age and regardless of treatment status or doctor diagnosis of hypertension, we found that OH was strongly associated with high blood pressure at baseline, above the recommended guidelines for treatment [16–18]. This indicates that undetected high blood pressure, rather than antihypertensive drugs, is a sufficient cause of OH. The combination of hypertension and OH in older adults is becoming increasingly recognised but may be due to the association of both OH and high blood pressure with increasing age [21]. Older hypertensive patients often display lability in blood pressure possibly via altered sympathovagal balance, increased left ventricular wall thickness, decreased left ventricular preload and alterations of left ventricular diastolic filling [23]. Improvement in postural blood pressure changes may occur with prolonged treatment with certain antihypertensive medications, possibly related to their action on reducing left ventricular wall thickness and enhancement of stroke volume [9, 24]. However, owing to the cross-sectional design of the study, we were unable to determine whether this association was related to duration of treatment with the various antihypertensives or not.

The mechanism for drug-induced OH is related to interference with reflex responses by drugs that may limit vasoconstriction, heart rate or cardiac output adjustments or exaggerate venous pooling such as vasodilators, alpha- and beta-adrenergic blockers, calcium channel blockers (CCB) and nitrates [25]. OH occurs via volume depletion with diuretics. It has been suggested that it may be a misconception to attribute OH to any particular type of antihypertensive due to a relatively high prevalence of OH among elderly patients in general [26]. However, in keeping with other studies [7, 10], our study showed strong age-adjusted associations with OH in women taking beta blockers, alpha-adrenoceptor blockers, thiazide diuretics and angiotensin-converting enzyme (ACE) inhibitors. Despite its intrinsic sympathomimetic activity and the reported effects of longstanding beta-blocker treatment in protecting the elderly from OH [22], the stronger relationship between OH and beta blockers may be due to its use in women with (as yet undetected) underlying coronary artery disease and left ventricular impairment. Expected associations with OH in those taking the older alpha-adrenoceptor blockers and CCBs [4, 7, 10] were not observed. This may reflect use of CCBs in patients suffering from coronary artery disease in addition to hypertension [20]. Despite having a more gradual onset of action and less postural hypotensive effects, women taking newer alpha-adrenoceptor blockers such as doxazosin still had a particularly strong age-adjusted odds of OH, although few were on these types of antihypertensive. Although nitrates have been associated with OH, especially in conjunction with other antihypertensives such as ACE inhibitors [20], there was modest evidence for a negative association with OH (P = 0.03). This ‘protective effect’ could be related to nitrate tolerance in those on sustained therapy [27]. There is evidence suggesting that some nitrates, particularly Glyceryl trinitrate (GTN), are associated with reduced bioavailability of GTN-derived nitrous oxide, impairing its vasodilator activity and, possibly, directly counteracting GTN-induced vasorelaxation [28].

Strengths and limitations

The BWHHS is a large, population-based cohort which is representative of British women. It has nurse-verified data on prescribed medications which reflect prescribing practices in a wide range of locations in the UK. We were also able to obtain blood pressure measurements according to a precise protocol applied by trained nurses, enabling us to measure OH objectively. Although our response rate (60%) is moderate, it is consistent with other large contemporary epidemiological surveys [29]. The respondents in our study were slightly younger than non-respondents, hence the results on the prevalence of OH in the general community may be an underestimation, especially since OH and medication use are generally more common in older age groups(>80 years) who were not part of our study. However, the prevalence we observed was considerably higher than that reported in earlier community studies [2, 12]. This higher prevalence may be due to a higher sitting systolic blood pressure taken at baseline. This ‘elevated’ first reading may be related to ‘white coat hypertension’ usually experienced when blood pressure is first taken upon consultation. Furthermore, the women were also told to take their medication before attending the medical examination and hence this too affected the prevalence of OH among them.

As this analysis is cross-sectional, we cannot examine the effects of changes in the prescribing of drugs, and it is possible, but unlikely, that the associations observed are susceptible to reverse causality. Also, the lack of relationship of OH with the use of one drug relative to another may reflect the play of chance, and the confidence intervals include the possibility of a modest increase in risk of OH. There were no available data on how many of these had symptomatic OH at the time of measurement. Post-prandial OH was not considered as all respondents were fasting at time of blood pressure measurements.

Implications

Although treatment of hypertension in older people is effective in cardiovascular disease prevention [30], OH or its symptoms are a major reason for withdrawal of antihypertensives. We recommend that judicious prescribing of specific classes of antihypertensives (i.e. beta blockers and use of three or more antihypertensives) is required if OH and its consequences are to be avoided. Checking for OH as part of routine detection and control of hypertension in primary care in view of the strong association of OH among those with previously unrecognised raised blood pressure would be a useful step in assessing the need for antihypertensive treatment.

Key points

  • Multiple co-morbidities are a more powerful independent determinant of orthostatic hypertension than the overall number of drugs taken.

  • Orthostatic hypotension is particularly common in women with previously unrecognised raised blood pressure, so checking for this as part of routine surveillance in primary care is necessary.

  • Blood pressure lowering with beta blockers and the use of three or more antihypertensives are independently associated with orthostatic hypotension, regardless of co-morbidities, and therefore close monitoring of signs and symptoms of OH is needed.

The British Women's Heart and Health Study is co-directed by Shah Ebrahim and D.A. Lawlor. We thank Carol Bedford, Alison Emerton, Nicola Frecknall, Karen Jones, Mark Taylor and Katherine Wornell for collecting and entering data, all the general practitioners and their staff who supported data collection and the women who participated in the study.

Contributors

All authors developed the study's aim and design and managed its data. All authors contributed to the final version. S.E. will act as guarantor for the paper.

Declaration of sources of funding

The British Women's Heart and Health Study was funded by the Department of Health Policy Research Programme and the British Heart Foundation.

S. Kamaruzzaman is a Lecturer in Geriatric Medicine at the University of Malaya and funded by Public Services Department of the Government of Malaysia.

Conflicts of interest

None.

Ethical approval

Local ethics committees approved the study, and 99.4% of participants gave written informed consent for their medical records to be available.

Supplementary data

Supplementary data mentioned in the text is available to subscribers at the journal website http://ageing.oxforjournal.org

References

1
Lipsitz
L
Orthostatic hypotension in the elderly
N Engl J Med
 , 
1989
, vol. 
321
 (pg. 
952
-
957
)
2
Masaki
KH
Schatz
IJ
Burchfiel
CM
, et al.  . 
Orthostatic hypotension predicts mortality in elderly men: the Honolulu Heart Program
Circulation
 , 
1998
, vol. 
98
 (pg. 
2290
-
5
)
3
Consensus statement on the definition of orthostatic hypotension, pure autonomic failure, and multiple system atrophy. The Consensus Committee of the American Autonomic Society and the American Academy of Neurology
Neurology
 , 
1996
, vol. 
46
 pg. 
1470
 
4
Hajjar
I
Postural blood pressure changes and orthostatic hypotension in the elderly patient: impact of antihypertensive medications
Drugs Aging
 , 
2005
, vol. 
22
 (pg. 
55
-
68
)
5
Rutan
G
Hermanson
B
Bild
D
Kittner
S
LaBaw
F
Tell
G
Orthostatic hypotension in older adults. The Cardiovascular Health Study. CHS Collaborative Research Group
Hypertension
 , 
1992
, vol. 
19
 (pg. 
508
-
19
)
6
Wu
J
Wu
N
Lu
F
Chang
C
Factors associated with orthostatic hypotension in the Chinese population in Taiwan
Am J Hypertens
 , 
1996
, vol. 
9
 (pg. 
999
-
1005
)
7
Luukinen
H
Koski
K
Laippala
P
Kivela
S
Prognosis of diastolic and systolic orthostatic hypotension in older persons
Arch Intern Med
 , 
1999
, vol. 
159
 (pg. 
273
-
80
)
8
Raiha
I
Luutonen
S
Piha
J
Seppanen
A
Toikka
T
Sourander
L
Prevalence, predisposing factors, and prognostic importance of postural hypotension
Arch Intern Med
 , 
1995
, vol. 
155
 (pg. 
930
-
5
)
9
Ooi
WL
Barrett
S
Hossain
M
Kelley-Gagnon
M
LA
L
Patterns of orthostatic blood pressure change and their clinical correlates in a frail, elderly population
JAMA
 , 
1997
, vol. 
277
 (pg. 
1299
-
304
)
10
Poon
I
Braun
U
High prevalence of orthostatic hypotension and its correlation with potentially causative medications among elderly veterans
J Clin Pharm Ther
 , 
2005
, vol. 
30
 (pg. 
173
-
8
)
11
Eigenbrodt
M
Rose
K
Couper
D
Arnett
D
Smith
R
Jones
D
Orthostatic hypotension as a risk factor for stroke: the atherosclerosis risk in communities (ARIC) study, 1987–1996
Stroke
 , 
2000
, vol. 
31
 (pg. 
2307
-
13
)
12
Rose
K
Eigenbrodt
M
Biga
R
, et al.  . 
Orthostatic hypotension predicts mortality in middle-aged adults: the Atherosclerosis Risk In Communities (ARIC) Study
Circulation
 , 
2006
, vol. 
114
 (pg. 
630
-
6
)
13
Reid MA
CP
Polypharmacy: causes and effects in older people
Prescriber
 , 
2005
(pg. 
57
-
62
)
14
Baena
MI
Faus
MJ
Fajardo
PC
, et al.  . 
Medicine-related problems resulting in emergency department visits
Eur J Clin Pharmacol
 , 
2006
, vol. 
62
 (pg. 
387
-
93
)
15
Lawlor
D
Patel
R
Ebrahim
S
Association between falls in elderly women and chronic diseases and drug use: a cross sectional study
BMJ
 , 
2003
, vol. 
327
 (pg. 
1
-
6
)
16
Williams
B
Poulter
NR
Brown
MJ
Davis
M
McInnes
GT
Potter
JF
Guidelines for management of hypertension: report of the fourth working party of the British Hypertension Society
J Human Hypertension
 , 
2004
, vol. 
18
 (pg. 
139
-
85
)
17
Guidelines Committee. European Society of Hypertension–European Society of Cardiology guidelines for the management of arterial hypertension
J Hypertension
 , 
2003
, vol. 
21
 (pg. 
1011
-
53
)
18
Guidelines Subcommittee
World Health Organization–International Society of Hypertension guidelines for the management of hypertension
J Hypertension
 , 
1999
, vol. 
17
 (pg. 
151
-
83
)
19
Alberti
KGMM
Aschner
P
Assal
J-P
, et al.  . 
Definition, diagnosis and classification of diabetes mellitus and its complications
Report of a WHO Consultation
 , 
1999
(pg. 
1
-
65
Part 1: diagnosis and classification of diabetes mellitus
20
Committee
J. British National Formulary
 , 
2007
pg. 
54
 
21
Harris
T
Lipsitz
LA
Kleinman
JC
Cornoni-Huntley
J
Postural change in blood pressure associated with age and systolic blood pressure
Journal of Gerontology: MEDICAL SCIENCES
 , 
1991
, vol. 
46
 (pg. 
M159
-
M163
)
22
Cleophas
TJ
van Marum
R
Age-related decline in autonomic control of blood pressure: implications for the pharmacological management of hypertension in the elderly
Drugs Aging
 , 
2003
, vol. 
20
 (pg. 
313
-
9
)
23
Gottdiener
JS
Yanez
D
Rautaharju
P
Gardin
JM
Bild
DE
Lima
J
Newman
AB
for the Cardiovascular Health Study
Orthostatic hypotension in the elderly: contributions of impaired LV filling and altered sympathovagal balance
Am J Geriatr Cardiol
 , 
2000
, vol. 
9
 pg. 
273
 
24
Gottlieb
S
Antihypertensives reduce left ventricular hypertrophy
BMJ
 , 
1999
, vol. 
318
 pg. 
1164
 
25
Hopson
JR
Rea
RF
Kienzle
MG
Alterations in reflex function contributing to syncope: orthostatic hypotension, carotid sinus hypersensitivity and drug-induced dysfunction
Herz
 , 
1993
, vol. 
18
 (pg. 
164
-
74
)
26
Meredith
P
Is postural hypotension a real problem with antihypertensive medication?
Cardiology
 , 
2001
, vol. 
96
 (pg. 
19
-
24
)
27
Parker
JD
Parker
JO
Nitrate therapy for stable angina pectoris
N Engl J Med
 , 
1998
, vol. 
338
 (pg. 
520
-
31
)
28
Gori
T
Parker
JD
The puzzle of nitrate tolerance: pieces smaller than we thought?
Circulation
 , 
2002
, vol. 
106
 (pg. 
2404
-
8
)
29
London. DoH
Health Survey for England: Cardiovascular Disease: The Stationary Office
 , 
1999
30
Insua
JT
Sacks
HS
Lau
TS
, et al.  . 
Drug treatment of hypertension in the elderly
Ann Intern Med
 , 
1994
, vol. 
121
 (pg. 
355
-
62
)

Supplementary data

Comments

3 Comments
Withdrawing antihypertensives on the basis of orthostatic hypotension
23 February 2010
Margaret Mac Mahon
The recent BWHH Study regarding orthostatic hypotension (OH) and medications suggests that clinicians should be cautious in the context of older adults taking antihypertensives and that OH is a ‘major reason for withdrawal of antihypertensives’ [1]. The corresponding editorial queries whether our treating BP in the elderly can ‘result in OH that could be potentially detrimental’ and proposes a large study to determine the prevalence of OH. We query these conclusions because the study actually found that the prevalence of OH in those taking and not taking antihypertensives was quite similar and that OH was not significantly associated with falls. Although an association with mortality was demonstrated this is commonly attributed to coexistent hypertension and ageing rather than any causal relationship. The most notable findings in their study were the strong correlation between OH and hypertension itself, and that after subgroup analysis, thiazides, ACE Inhibitors and calcium channel blockers were not associated with OH. In practice, clinicians often withdraw antihypertensives on the basis of OH, on the assumption that it results in falls or syncope but there have been some excellent reviews challenging the association between antihypertensives and falls [2] and syncope [3] and a recent meta-analysis assessing the impact of medications on falls in almost 80,000 elderly found the association between antihypertensives and falls was not highly significant [4]. We are not aware either of any evidence supporting the withdrawal of antihypertensives in the context of falls reduction or any other outcome benefit. We agree with previous authors who reported that postural symptoms correlate much more strongly with falls than does OH. There is good evidence that elderly subjects with postural instability are more likely to have ischaemic lesions on neuroimaging and to have multiple vascular risk factors so we would hypothesise that reducing antihypertensives in such patients could actually increase their risk of vascular ischaemia, and therefore aggravate postural instability even further. We therefore do not feel that OH alone should be the basis for withdrawing antihypertensives and that reducing such treatment requires greater justification. 1. Kamaruzzaman S, Watt H, Carson C, Ebrahim S. The association between orthostatic hypotension and medication use in the British Women’s Heart and Health Study. Age Ageing 2010; 39: 51-56. 2. Vogt T. Falls, fallacies and hypertension. J Am Geriatr Soc 1995; 43: 1169 3. Wynne H, Schofield S. Drug-induced hypotension. In: Syncope in the elderly (editor RA Kenny) 1996 Chapman & Hall pp 137-154. 4. Woolcott JC et al. Meta-analysis of the impact of 9 medication classes on falls in elderly persons. Arch Intern Med 2009; 169(21): 1952-60

Conflict of Interest:

None declared

Submitted on 23/02/2010 7:00 PM GMT
Re: Withdrawing antihypertensives on the basis of orthostatic hypotension
25 April 2010
Shahrul Kamaruzzaman

We appreciated Dr MacMahon and colleagues' letter commenting on our paper and the accompanying editorial [1].We agree that there is no strong evidence on the effects on falls of withdrawing antihypertensive medication in people with postural symptoms or in those with asymptomatic orthostatic hypotension (OH).We did not conclude that OH alone should be a criterion for withdrawing antihypertensive treatment.Data from British Women's Heart & Health Study has shown that falls are more strongly associated with co-morbidity than with polypharmacy [2, 3].We did show that OH was twice as prevalent in women with undiagnosed and uncontrolled high blood pressure. On the basis of the findings presented in our paper we concluded that:"uncontrolled hypertension, use of three or more antihypertensives and multiple co-morbidities are predictors of OH in older women. Detection or monitoring of OH in these groups may prevent women from suffering its adverse consequences".

Shah Ebrahim, Claire Carson, Hilary Watt, Sharul Kamaruzzaman

1. Margaret MacMahon et al Age and Ageing 24 February 2010 (letter).

2. Kamaruzzaman S, Watt H, Carson C, Ebrahim S. The association between orthostatic hypotension and medication use in the British Women’s Heart and Health Study. Age and Ageing 2010; 39: 51-56.

3. Lawlor DA, Patel R, Ebrahim S. Association between falls in elderly women and chronic diseases and drug use: cross sectional study. BMJ Sep 2003; 327: 712 - 717

Conflict of Interest:

None declared

Submitted on 25/04/2010 8:00 PM GMT
WITHDRAWING ANITHYPERTENSIVES ON THE BASIS OF ORTHOSTATIC HYPOTENSION
27 July 2010
Edmund J Dunstan

Neither the original paper from the British Women’s Heart and Health Study (1) nor subsequent correspondence (2,3) acknowledge the fundamental weakness of this work: that it only recruited women up to the age of 80. It thus excluded the group at most potential risk of falls, those over 80, in whom the balance of future benefit (a reduced chance of stroke) against present harm (symptomatic orthostatic hypotension) might be expected to be most crucial. Using a sensitive definition of orthostatic hypotension (20mm Hg drop in systolic pressure, or 10 mm Hg in diastolic) it is not surprising that Choulerton et al. quote little relationship between postural hypotension and falls: clinical experience is that only substantially (and variably) greater drops are associated with symptoms. As so often, the people about whom we really need to know are the ones we don’t.

Unfortunately, this study is analogous to research on perinatal mortality that excludes all births below 38 weeks of gestation.

References 1. Kamarruzzaman S, Watt H, Carson C & Ebrahim S: The association between orthostatic hypotension and medication use in the British Women’s Heart and Health Study. Age and Ageing 2010; 39: 51-56 2. Choulerton J, Mudd P& MacMahon M Withdrawing antihypertensives on the basis of orthostatic hypotension. Age and Ageing 2010;39:518 3. Ebrahim S, Carson C, Watt H & Kamaruzzaman S. Re: Withdrawing anithypertensives on the basis of orthostatic hypotension. Age and Ageing 2010;39:518

Conflict of Interest:

None declared

Submitted on 27/07/2010 8:00 PM GMT