Abstract
BACKGROUND
Although venous stasis is a risk factor for venous thromboembolism (VTE) and orthostatic hypotension (OH) can cause venous stasis, to our knowledge no study has examined the relationship between OH and VTE risk. We sought to quantify the association between OH and VTE (deep vein thrombosis or pulmonary embolism) using data from 2 large, prospective cohort studies: the Cardiovascular Health Study (CHS) and the Atherosclerosis Risk in Communities (ARIC) Study. We hypothesized that OH was positively associated with incident VTE.
METHODS
We measured OH—defined as a drop in systolic blood pressure (SBP) of at least 20mm Hg or diastolic blood pressure (DBP) of at least 10mm Hg within 3 minutes of standing—in participants aged 45–64 years in ARIC (n = 12,480) and ≥65 years in CHS (n = 5,027) at baseline visits (1987–1989 in ARIC; 1989–1990 and 1992–1993 in CHS), and followed participants for incident VTE (n = 568 in ARIC through 2011 and n = 148 in CHS through 2001). We calculated adjusted hazard ratios (HRs) and their 95% confidence intervals (CIs) for incident VTE in relation to OH status.
RESULTS
In CHS, there was a positive association between OH status and incident VTE (HR for VTE = 1.74 (95% CI: 1.20–2.51)). In contrast, there was no association between OH and VTE in the ARIC study (HR for VTE = 0.97 (95% CI: 0.70–1.33)).
CONCLUSIONS
Community-dwelling older adults with OH had a moderately increased risk of VTE. These results were not seen in a population-based middle-aged cohort.
Keywords: blood pressure, epidemiology, hypertension, orthostatic hypotension, orthostatic intolerance, risk factors, venous thromboembolism.
Orthostatic hypotension (OH) is a common condition in older adults. A current consensus statement has defined OH as a drop in systolic blood pressure (SBP) of at least 20mm Hg or diastolic blood pressure (DBP) of at least 10mm Hg within 3 minutes of standing from a supine position.1 OH is not part of a singular disease process, but can result from a variety of pathologies. If cerebral blood flow is affected, OH can cause dizziness, blurred vision, chronic fatigue, fainting, or pain in the neck and shoulders.2
When standing upright from supine position, about 500 to 1,000ml of blood is redistributed to lower extremities due to gravity, resulting in reduced venous return to the heart.2,3 Multiple compensatory responses are activated to maintain arterial pressure by favoring an increase in venous return.2,3 Conversely, in an individual with OH, at least one of these compensatory mechanisms is impaired. Blood pressure drops upon standing and venous pooling in the legs is sustained for a longer duration.2,3 Blood pooling can increase venous thromboembolism (VTE) risk.4 Therefore, a reasonable hypothesis is that OH may increase VTE risk through an increased duration of venous pooling.
Although it is known that VTE risk is related to physical characteristics of the leg that affect venous return, such as varicose veins5 and leg length,6 to our knowledge no study has examined the relationship between OH and VTE risk. Therefore, we tested whether OH is positively associated with incident VTE (leg deep vein thrombosis or pulmonary embolism) using data from 2 large, prospective cohort studies: the Cardiovascular Health Study (CHS) and the Atherosclerosis Risk in Communities (ARIC) Study.
METHODS
Study population
Between 1987 and 1989, ARIC recruited and examined 15,792 participants aged 45–64 years living in 4 US communities: Forsyth County, NC; Jackson, MS; suburban Minneapolis, MN; and Washington County, MD.7 CHS sampled from Medicare lists, recruited, and examined 5,888 community-dwelling participants aged 65 years or older living in 4 US communities (Pittsburgh, PA; Forsyth County, NC; Sacramento County, CA; and Washington County, MD) between 1989–1990 and 1992–1993.8
Individuals were excluded from all analyses if they had a history of VTE or anticoagulant use at baseline (n = 347 in ARIC, 419 in CHS), had missing data on any variable included in the analysis (n excluded in ARIC = 2,910, 442 in CHS), or were of a race other than African American or White (due to small numbers) (n = 0 in both ARIC and CHS due to previous exclusions). The final sample size was 12,480 in ARIC and 5,027 in CHS. Both studies were approved by the Institutional Review Boards of the collaborating institutions and informed consent was obtained from all participants.
Notably, 2,469 participants in ARIC were missing a measurement of OH at baseline, most of whom attended the baseline exam in the first 6 months of the ARIC study, before the postural blood pressure measurement protocol was implemented. We excluded these participants from this study, and considered whether selection bias would be present as a result. However, participants were randomly assigned examination dates, and those with a measurement of OH were not significantly different from those without a measurement in terms of seated SBP, anthropometric variables, or age, race, and sex distributions.9 Therefore, it is reasonable to believe that participants in this study represent a subsample of the entire ARIC cohort.
VTE ascertainment
Incident VTE was defined as the first occurrence of a validated deep vein thrombosis in the leg or pulmonary embolism from baseline through the end of follow-up: 31 December 2011 for the ARIC study and 31 December 2001 for CHS.
Hospitalizations were identified through participant report in both studies, through Medicare records in CHS, and through surveillance of community hospitals’ discharge lists in ARIC.10 Possible VTEs were identified using hospital discharge ICD codes and validated by physician review using standardized criteria.10 VTE events were categorized as provoked or unprovoked. Provoked VTE was defined as occurring within 90 days of major trauma, surgery, marked immobility, or within 1 year of active cancer. Unprovoked were all other confirmed VTE cases.
Measuring and modeling OH
In both CHS and ARIC, OH was defined as a SBP reduction upon standing of at least 20mm Hg, a DBP reduction of at least 10mm Hg, or both.1 In addition, participants in CHS who needed to abort the OH measurement procedure due to dizziness, lightheadedness, or faintness during standing were classified as having OH.
At ARIC’s baseline visit, after 20 minutes of resting in the supine position, blood pressure was measured supine approximately every 30 seconds for 2 minutes (2–5 measurements, 90% of participants had ≥4 measurements) using a Dinamap 1846 SX automated oscillometric device. Participants then stood, held hands on the chest, and blood pressure was measured repeatedly for the first 2 minutes after standing (2–5 measurements, 91% of participants had ≥4 measurements).11 Blood pressure change in ARIC was defined as the first supine blood pressure measurement minus the last standing blood pressure measurement (approximately 2 minutes after standing).
At CHS baseline visit, after at least 20 minutes of resting in the supine position, blood pressure was measured supine using a mercury sphygmomanometer (Baumanometer, W.A. Baum, Copiague, NY) according to a standardized protocol following recommendations.12 Participants then stood for 3 minutes, after which they rested a hand on a stand positioned at heart level, and blood pressure was measured. If dizziness, lightheadedness, or faintness occurred during standing, the OH measurement procedure was immediately aborted. Blood pressure change in CHS was defined as the supine blood pressure measurement minus the 3-minute standing blood pressure measurement.
Baseline measurements
Variables measured similarly in CHS and ARIC.
Age, race, sex, history of VTE, and perception of general health were self-reported. Body mass index was calculated as weight (kg) divided by height (m).2 Torso length was estimated as the difference between seated height and stool height; leg length was estimated as the difference between standing height and torso length. Medication use was identified from prescription bottles.13 “Use of medications that might induce OH” was defined as use of antihypertensives, anti-psychotics, antidepressants, anti-cholinergics, narcotics, nitro compounds, or sedatives.14 Diabetes mellitus was defined as fasting glucose ≥7 mmol/l (126mg/dl), nonfasting blood glucose ≥11.1 mmol/l (200mg/dl), a self-report of physician diagnosis, and/or current medication use for diabetes. Seated blood pressure was measured using a random-zero sphygmomanometer; the last 2 of 3 blood pressure measurements were averaged. Hypertension was defined as SBP ≥ 140mm Hg, DBP ≥ 90mm Hg, or taking antihypertensive medications.
Variables measured differently in CHS and ARIC.
In ARIC, prevalent coronary heart disease was defined as (i) electrocardiogram evidence of a previous myocardial infarction or (ii) a self-reported history of coronary revascularization or myocardial infarction; prevalent heart failure was defined as signs and symptoms of heart failure by the Gothenburg questionnaire.15 In CHS, prevalent coronary heart disease, prevalent heart failure, and prevalent stroke were based on self-report, which was validated with information from the baseline examination, medical records, or from physician questionnaires.16
Statistical analyses
Analyses were performed using SAS (version 9.2, SAS Institute, Cary, NC). A P-value of <0.05 on a 2-tailed test was considered statistically significant. We statistically evaluated whether the OH relation with VTE differed by study: In a dataset that contained both studies, we ran a VTE model that included the OH variable, a cohort variable (CHS, ARIC), and a multiplicative interaction term (OH by study). Because the interaction term indicated statistically significant differences in the OH–VTE relation by cohort (P-value = 0.02), all analyses were stratified by cohort and reasons for potential differences were explored. Baseline participant characteristics were presented by OH status. We computed person-years of follow-up as time elapsed from the baseline examination to whichever came first: VTE event, death, loss to follow-up, or the end of follow-up (31 December 2001 in CHS and 31 December 2011 in ARIC). VTE incidence rates were calculated by dividing the number of VTE events by person-years of follow-up, and 95% confidence intervals (CIs) were obtained using Rothman’s Episheet (krothman.hostbyet2.com/Episheet.xls).
Cox proportional hazards regression was used to calculate hazard ratios (HRs) and 95% CIs for incident VTE (total, provoked, and unprovoked) in relation to OH status (yes/no), using those without OH as the referent. We verified the proportional hazards assumption through inspection of ln(–ln) survival curves by OH status. Model 1 was adjusted for age (continuous), race (African American, White), and sex (male, female). Model 2, our main model, was adjusted for variables in model 1 plus body mass index (continuous), self-perceived health status (excellent, very good, good, fair, poor), and leg length (continuous). Model 3 is adjusted for all covariates that were found to be statistically different by OH status (Table 1). Analyses were also stratified by variables of interest.
Table 1.
Baseline characteristics of ARIC and CHS participants according to baseline orthostatic hypotension status
| ARIC | CHS | |||||
|---|---|---|---|---|---|---|
| Characteristics (means or prevalences) | Orthostatic hypotension (N = 934) | No orthostatic hypotension (N = 11,546) | P-value for differences by OH statusa | Orthostatic hypotension (N = 915) | No orthostatic hypotension (N = 4,112) | P-value for differences by OH statusa |
| Age, years ± SD | 56.3±5.7 | 53.9±5.7 | <0.0001 | 73.5±5.7 | 72.6±5.5 | <0.0001 |
| Male, % | 47.0 | 45.2 | 0.3 | 43.5 | 43.0 | 0.8 |
| African American, % | 29.9 | 26.1 | 0.01 | 15.2 | 16.6 | 0.3 |
| Diabetes mellitus, % | 20.7 | 11.3 | <0.0001 | 18.6 | 15.9 | 0.05 |
| History of heart failure, % | 12.4 | 9.7 | 0.007 | 11.5 | 10.4 | 0.3 |
| History of coronary heart disease, % | 7.4 | 4.5 | <0.0001 | 17.8 | 13.5 | 0.0007 |
| History of stroke, % | 4.2 | 1.5 | <0.0001 | 5.8 | 4.1 | 0.02 |
| Body mass index, kg/m2 ± SD | 27.9±5.9 | 27.6±5.2 | 0.06 | 26.1±4.5 | 26.8±4.7 | <0.0001 |
| Self-perceived health status—fair/poor, % | 30.5 | 17.9 | <0.0001 | 28.3 | 23.0 | 0.0007 |
| Leg length, cm ± SD | 80.2±5.6 | 80.0±5.8 | 0.3 | 79.2±6.1 | 79.6±5.7 | 0.04 |
| Hypertension, % | 52.4 | 32.8 | <0.0001 | 67.9 | 64.4 | 0.05 |
| Use of medications that might induce orthostatic hypotension, % | 48.2 | 33.1 | <0.0001 | 66.8 | 61.9 | 0.006 |
| Seated systolic blood pressure, mm Hg ± SD | 128±22 | 121±19 | <0.0001 | 138±24 | 136±21 | 0.04 |
| Mean change in systolic blood pressure upon standing, mm Hg ± SD | −24±12 | 3±11 | <0.0001 | −20±14 | −1±10 | <0.0001 |
| Seated diastolic blood pressure, mm Hg ± SD | 75±12 | 73±11 | <0.0001 | 69±12 | 71±11 | <0.0001 |
| Mean change in diastolic blood pressure upon standing, mm Hg ± SD | −5±8 | 5±6 | <0.0001 | −8±10 | 3±8 | <0.0001 |
Baseline ARIC Study: 1987–1989; Baseline CHS: 1989–1990 and 1992–1993.
Abbreviations: ARIC, Atherosclerosis Risk in Communities Study; CHS, Cardiovascular Health Study.
a T-tests were used to compare means and chi-squared tests were used to compare prevalences.
RESULTS
At baseline, the mean age of ARIC participants was 54 years, approximately 50% were female, one-quarter were African American, and the prevalence of OH was 7.5%. In comparison, the CHS cohort at baseline was older (mean age = 73), had a lower proportion of African Americans, a similar sex distribution, and a higher prevalence of OH (18.2%).
Among those with OH, the mean drop in SBP upon standing was 24 and 20mm Hg (ARIC and CHS, respectively). In contrast, the mean drop in DBP upon standing among those with OH was only 5 and 8mm Hg (ARIC and CHS, respectively), indicating that more participants fit the criteria for OH due to a 20mm Hg or more SBP reduction upon standing rather than a DBP reduction of 10mm Hg or more.
Several baseline characteristics—age, diabetes mellitus, history of heart failure, history of coronary heart disease, history of stroke, self-perceived health status, hypertension, use of medications that may induce OH, and seated SBP—were positively associated with OH (Table 1). Notably, with our large sample size P-values can be significant when differences between groups are trivial.
Over a median follow-up time of 17.8 years (245,620 person-years), 568 ARIC participants had VTE events; over a median follow-up time of 11.6 years (47,655 person-years), 148 CHS participants had VTE events. In ARIC, the crude incidence rate (95% CI) of VTE per 1,000 person-years was 2.54 (1.85–3.41) in those with OH and 2.30 (2.11–2.50) in those without OH; in CHS, the crude incidence rate (95% CI) of VTE per 1,000 person-years was 4.76 (3.44–6.44) in those with OH and 2.76 (2.28–3.32) in those without OH (Table 2).
Table 2.
HRs (95% CIs) for incident VTE among participants with OH at baseline compared to those without
| ARIC | CHS | |||
|---|---|---|---|---|
| Orthostatic hypotension (N = 934) | No orthostatic hypotension (N = 11,546) | Orthostatic hypotension (N = 915) | No orthostatic hypotension (N = 4,112) | |
| Total VTE | ||||
| N total VTE events | 41 | 527 | 39 | 109 |
| Median (IQR) follow-up, years | 21.2 (10.9) | 22.5 (5.1) | 10.3 (5.4) | 11.6 (4.0) |
| Unadjusted IR (95% CI)a | 2.54 (1.85, 3.41) | 2.30 (2.11, 2.50) | 4.76 (3.44, 6.44) | 2.76 (2.28, 3.32) |
| Model 1 HR (95% CI)b | 1.05 (0.76, 1.45) | 1 (referent) | 1.70 (1.18, 2.45) | 1 (referent) |
| Model 2 HR (95% CI)c | 0.97 (0.70, 1.33) | 1 (referent) | 1.74 (1.20, 2.51) | 1 (referent) |
| Model 3 HR (95% CI)d | 0.97 (0.71, 1.34) | 1 (referent) | 1.76 (1.22, 2.55) | 1 (referent) |
| Unprovoked VTE | ||||
| N unprovoked VTE events | 11 | 204 | 11 | 45 |
| Model 2 HR (95% CI)c | 0.67 (0.36, 1.23) | 1 (referent) | 1.14 (0.59, 2.21) | 1 (referent) |
| Model 3 HR (95% CI)d | 0.67 (0.36, 1.23) | 1 (referent) | 1.18 (0.61, 2.30) | 1 (referent) |
| Provoked VTE | ||||
| N provoked VTE events | 30 | 323 | 28 | 64 |
| Model 2 HR (95% CI)c | 1.16 (0.80, 1.69) | 1 (referent) | 2.19 (1.40, 3.44) | 1 (referent) |
| Model 3 HR (95% CI)d | 1.16 (0.80, 1.70) | 1 (referent) | 2.19 (1.39, 3.45) | 1 (referent) |
Abbreviations: ARIC, Atherosclerosis Risk in Communities Study; CHS, Cardiovascular Health Study; CI, confidence interval; HR, hazard ratio; IQR, interquartile range; IR, incidence rate; OH, orthostatic hypotension; VTE, venous thromboembolism.
aRate/1,000 years. bModel 1 is adjusted for age, sex, and race. cModel 2 is adjusted for model 1 covariates, body mass index, self-perceived health status, and leg length. dModel 3 is adjusted for model 2 covariates, seated systolic blood pressure, seated diastolic blood pressure, blood pressure medications, diabetes, history of heart failure, history of coronary heart disease, history of stroke, and use of medications that might induce orthostatic hypotension.
The association of OH with VTE differed between ARIC and CHS (P-value for interaction by study = 0.02). In ARIC, HRs for incident VTE (total, provoked, and unprovoked) among participants with OH at baseline compared to those without were approximately 1.0 in all regression models and by VTE type, indicating no association (Table 2). In CHS, there was a positive association between incident OH status and VTE in both models (model 2 HR for total VTE = 1.74 (95% CI: 1.20–2.51)), greater for provoked than unprovoked VTE.
As shown in Table 3, most subgroup-specific HRs for total VTE were not statistically different from each other, although power to detect a difference was low. However, the OH–VTE relation was marginally statistically different by (i) history of stroke status in ARIC (HR = 2.87 (95% CI: 0.83–9.92) in those with a history of stroke and 0.90 (95% CI: 0.64–1.26) in those without a history of stroke, P-value for interaction by stratum = 0.05), and (ii) self-reported health status in ARIC (HR = 0.73 (95% CI: 0.47–1.14) in those with a self-reported health status of excellent, very good or good and 1.56 (95% CI: 0.97–2.53) in those with a self-reported health status of fair or poor, P-value for interaction by stratum = 0.04). In CHS, the OH–VTE HRs were qualitatively different by history of stroke: HR = 34.26 (95% CI: 2.06–570.7) in those with a history of stroke and 1.60 (95% CI: 1.09–2.35) in those without a history of stroke, P-value for interaction by stratum = 0.10.
Table 3.
Adjusteda HRs (95% CIs) for incident total VTE among participants with orthostatic hypotension at baseline compared to those without, stratified by baseline characteristics
| Strata | ARIC | CHS | ||||
|---|---|---|---|---|---|---|
| Number of VTE events/total number in stratum | HR (95% CI) | P-value for interaction by stratumb | Number of VTE events/total number in stratum | HR (95% CI) | P-value for interaction by stratumb | |
| Overall | 568/12,480 | 0.97 (0.70, 1.33) | N/A | 148/5,027 | 1.74 (1.20, 2.51) | N/A |
| Age <55 years | 243/6,629 | 0.65 (0.33, 1.26) | 0.18 | 0 | N/A | N/A |
| Age ≥55 years | 325/5,851 | 1.14 (0.79, 1.65) | 100 | N/A | ||
| Age <70 years | 100 | N/A | N/A | 45/1,725 | 2.07 (1.06, 4.05) | 0.38 |
| Age ≥70 years | 0 | N/A | 103/3,302 | 1.61 (1.03, 2.50) | ||
| Female | 317/6,825 | 1.08 (0.72, 1.62) | 0.43 | 80/2,862 | 2.09 (1.28, 3.39) | 0.29 |
| Male | 251/5,655 | 0.85 (0.51, 1.45) | 68/2,165 | 1.40 (0.78, 2.49) | ||
| African American | 192/3,296 | 1.25 (0.75, 2.10) | 0.41 | 35/822 | 1.09 (0.44, 2.71) | 0.23 |
| White | 376/9,184 | 0.87 (0.58, 1.31) | 113/4,205 | 1.99 (1.32, 2.99) | ||
| Diabetes | 76/1,493 | 0.85 (0.39, 1.86) | 0.62 | 29/825 | 1.81 (0.81, 4.07) | 0.78 |
| No Diabetes | 492/10,987 | 0.99 (0.70, 1.4) | 119/4,202 | 1.69 (1.11, 2.57) | ||
| History of heart failure | 75/1,231 | 0.85 (0.37, 1.97) | 0.73 | 15/531 | 1.25 (0.34, 4.63) | 0.45 |
| No history heart failure | 493/11,249 | 0.98 (0.69, 1.38) | 133/4,496 | 1.81 (1.23, 2.67) | ||
| History of coronary heart disease | 20/584 | N/A | 0.95 | 12/717 | 1.95 (0.50, 7.56) | 0.55 |
| No history of coronary heart disease | 548/11,896 | 1.03 (0.75, 1.42) | 136/4,310 | 1.83 (1.24, 2.68) | ||
| History of stroke | 13/207 | 2.87 (0.83, 9.92) | 0.05 | 6/221 | 34.26 (2.06, 570.7) | 0.10 |
| No history stroke | 555/12,273 | 0.90 (0.64, 1.26) | 142/4,806 | 1.60 (1.09, 2.35) | ||
| Body mass index <25kg/m2 at baseline | 127/4,206 | 1.05 (0.55, 2.01) | 0.63 | 40/1,946 | 1.55 (0.77, 3.15) | 0.79 |
| Body mass index ≥25kg/m2 at baseline | 441/8,274 | 0.96 (0.67, 1.39) | 108/3,081 | 1.79 (1.16, 2.77) | ||
| Self-reported health status - excellent/very good/good | 428/10,131 | 0.73 (0.47, 1.14) | 0.04 | 105/3,821 | 1.84 (1.18, 2.86) | 0.75 |
| Self-reported health status - fair/poor | 140/2,349 | 1.56 (0.97, 2.53) | 43/1,206 | 1.56 (0.80, 3.05) | ||
| Use of medications that might induce OH | 223/4,269 | 1.09 (0.71, 1.69) | 0.54 | 86/3,156 | 1.90 (1.19, 3.04) | 0.52 |
| No use of medications that might induce OH | 345/8,211 | 0.86 (0.53, 1.38) | 62/1,871 | 1.50 (0.82, 2.74) | ||
| Hypertension | 242/4,271 | 1.16 (0.78, 1.72) | 0.21 | 91/3,270 | 1.57 (0.98, 2.54) | 0.53 |
| No hypertension | 326/8,209 | 0.71 (0.41, 1.24) | 57/1,757 | 2.02 (1.12, 3.63) | ||
| Leg length ≥ 80 cm | 319/6,433 | 0.82 (0.52, 1.29) | 0.27 | 75/2,428 | 1.96 (1.16, 3.30) | 0.61 |
| Leg length < 80 cm | 249/6,047 | 1.19 (0.76, 1.86) | 73/2,599 | 1.52 (0.90, 2.58) | ||
Study- and stratum-specific HRs were computed from separate statistical models ran by study and stratum.
Abbreviations: ARIC, Atherosclerosis Risk in Communities Study; CHS, Cardiovascular Health Study; CI, confidence interval; HR, hazard ratio; VTE, venous thromboembolism.
aHRs adjusted for age, sex, race, body mass index, self-perceived health status, and leg length. bMultiplicative interactions of orthostatic hypotension with strata were evaluated by including cross-product terms in the models (dataset contained both strata).
Sensitivity analyses
Results did not appreciably change if we restricted follow-up time to 10 years so that VTEs—particularly in ARIC—were closer to the baseline OH measurement (HR = 1.02 (95% CI: 0.56–1.85) in ARIC and 1.55 (95% CI: 1.04–2.32) in CHS). Results also remained similar after excluding participants with VTE events within 90 days of marked immobility (HR = 1.50 (95% CI: 0.86–2.62) in ARIC and 2.02 (95% CI: 0.88–4.66) in CHS) or active cancer (HR = 1.16 (95% CI: 0.67–2.01) in ARIC and 2.27 (95% CI: 1.07–4.82) in CHS) in an attempt to remove these potential confounding variables. When we adjusted for model 2 covariates plus frailty (measured in the CHS cohort only), the HR was mildly attenuated compared to model 2 alone, but still statistically significant (HR = 1.64; 95% CI: 1.10–2.45). Results also did not appreciably change when we excluded VTE events from the first 5 years after baseline (HR = 0.95 (95% CI: 0.68–1.34) in ARIC and 1.98 (95% CI: 1.23–3.19) in CHS).
DISCUSSION
In the population-based CHS cohort of older adults, those with OH at baseline had an approximately 70% higher risk of incident total VTE than those without OH at baseline: risk was nominally greater for provoked than unprovoked VTE. In contrast, there was no association between OH and VTE in the ARIC study of initially middle-aged adults. Results did not appreciably change in several sensitivity analyses. In ARIC, the OH–VTE relation was marginally statistically different by history of stroke status (higher in those with a history of stroke) and self-reported health status (higher in those with worse self-reported health status). In CHS, OH–VTE estimates were qualitatively different by history of stroke stratum (also higher in those with a history of stroke). However, a large number of statistical tests were performed and, thus, subgroup analyses should be interpreted as hypothesis-generating and need to be replicated elsewhere.
We are unsure why the relation of OH with VTE differed between ARIC and CHS. This difference was not explained by longer follow-up time in ARIC, as results were similar when we restricted follow-up time to 10 years. To verify that the OH–VTE relation in CHS was independent of confounding variables, we not only adjusted for several VTE risk factors, but also conducted analyses that excluded participants with VTE events within 90 days of marked immobility or within 1 year of active cancer. Since the association remained, it is likely that these variables did not confound our results. Further, when we excluded VTE events from the first 5 years after baseline, the association remained similar, making it unlikely that an undetected condition at baseline (e.g., cancer) could explain the association. Notably, the 2 cohorts had different, nonoverlapping age ranges at baseline. However, the OH–VTE relation did not differ by age categories within ARIC and CHS, providing no evidence for interaction by age. It is possible, but untestable, that differences in arm position contributed to the different results between ARIC and CHS: CHS used a Mayo stand for arm support at heart level and ARIC had hands placed on the chest. Having one’s arms in the incorrect position (not at heart level) overestimates standing blood pressure, and therefore underestimates the prevalence of OH.17,18 It is also possible, but untestable, that discrepancies in timing of standing blood pressure measurements contributed to the different results between ARIC and CHS. Indeed, if longer duration of venous pooling is a risk factor for VTE—as it is shown to be for long-distance travel19—then a drop in blood pressure that remains 3 minutes after standing, as in CHS, would be expected to have a stronger relation with VTE than a drop in blood pressure only 2 minutes or less after standing, as in ARIC. Future research could examine the relation between “delayed OH”—a variant of OH that occurs beyond 3 minutes of standing1—and VTE. Delayed OH is at least as prevalent as the “classical” OH that we measured in this study20 and some believe21 that the optimal time duration of orthostatic challenge for the diagnosis of OH is longer than the 3 minutes called for in the consensus definition.
If our findings in CHS are replicated, they could be explained through the increased duration of venous pooling associated with OH. The venous system of the buttocks and lower extremities is a large capacitance system, and upon standing from supine position, blood pools. This results in a decrease in venous return to the heart or a decrease in cardiac output. To maintain cardiac output, the normal physiological response is to activate compensatory mechanisms: an interplay between neurohumoral effects, the skeletal muscle pump, neurovascular compensation, and cerebral blood flow regulation.2,3 These compensatory mechanisms cause a reduction in venous pooling and an increase in venous return.2,3 Conversely, in an individual with OH, at least one of the compensatory mechanisms is impaired and there is a longer duration of venous pooling.2,3 Blood pooling (i.e., venous stasis) is an element of Virchow’s Triad—which includes factors that are thought to contribute to thrombosis—and can increase VTE risk.4
Although previous studies have examined the relation between OH and other cardiovascular diseases, to our knowledge no other study has examined the relation between OH and VTE risk.22 In conclusion, community-dwelling older adults with OH have a moderately increased risk of VTE. These results were not seen in a population-based middle-aged cohort.
DISCLOSURE
The authors declared no conflict of interest.
ACKNOWLEDGMENTS
We thank staff and participants of the ARIC and CHS studies for their important contributions.
The ARIC Study was carried out as a collaborative study supported by National Heart, Lung, and Blood Institute (NHLBI) contracts (HHSN268201100005C, HHSN268201100006C, HHSN268201100007C, HHSN268201100008C, HHSN2682 01100009C, HHSN268201100010C, HHSN268201100011C, and HHSN268201100012C). CHS was supported by contracts HHSN268201200036C, HHSN268200800007C, N01HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083, N01HC85086, and grant U01HL080295 from the NHLBI, with additional contribution from the National Institute of Neurological Disorders and Stroke (NINDS). Additional support was provided by R01AG023629 from the National Institute on Aging (NIA). E.J.B. was supported by NHLBI training grant T32HL007779.
REFERENCES
- 1. Freeman R, Wieling W, Axelrod FB, Benditt DG, Benarroch E, Biaggioni I, et al. Consensus statement on the definition of orthostatic hypotension, neurally mediated syncope and the postural tachycardia syndrome. Auton Neurosci 2011; 161:46–48. [DOI] [PubMed] [Google Scholar]
- 2. Fedorowski A, Melander O. Syndromes of orthostatic intolerance: a hidden danger. J Intern Med 2013; 273:322–335. [DOI] [PubMed] [Google Scholar]
- 3. Perlmuter LC, Sarda G, Casavant V, Mosnaim AD. A review of the etiology, associated comorbidities, and treatment of orthostatic hypotension. Am J Ther 2013; 20:279–291. [DOI] [PubMed] [Google Scholar]
- 4. Esmon CT. Basic mechanisms and pathogenesis of venous thrombosis. Blood Rev 2009; 23:225–229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Heit JA. The epidemiology of venous thromboembolism in the community. Arterioscler Thromb Vasc Biol 2008; 28:370–372. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Lutsey PL, Cushman M, Heckbert SR, Tang W, Folsom AR. Longer legs are associated with greater risk of incident venous thromboembolism independent of total body height. The Longitudinal Study of Thromboembolism Etiology (LITE). Thromb Haemost 2011; 106:113–120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am J Epidemiol 1989; 129:687–702. [PubMed] [Google Scholar]
- 8. 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. [DOI] [PubMed] [Google Scholar]
- 9. Nardo CJ, Chambless LE, Light KC, Rosamond WD, Sharrett AR, Tell GS, Heiss G. Descriptive epidemiology of blood pressure response to change in body position. The ARIC study. Hypertension 1999; 33:1123–1129. [DOI] [PubMed] [Google Scholar]
- 10. Cushman M, Tsai AW, White RH, Heckbert SR, Rosamond WD, Enright P, Folsom AR. Deep vein thrombosis and pulmonary embolism in two cohorts: the longitudinal investigation of thromboembolism etiology. Am J Med 2004; 117:19–25. [DOI] [PubMed] [Google Scholar]
- 11. The ARIC Investigators. ARIC Manual 11: Sitting Blood Pressure and Postural Changes in Blood Pressure and Heart Rate. ARIC Coordinating Center: Chapel Hill, NC, 1987. [Google Scholar]
- 12. Recommendations for human blood pressure determination by sphygmomanometers. Circulation 1988; 77:501A–514A. [PubMed] [Google Scholar]
- 13. Psaty BM, Lee M, Savage PJ, Rutan GH, German PS, Lyles M. Assessing the use of medications in the elderly: methods and initial experience in the Cardiovascular Health Study. The Cardiovascular Health Study Collaborative Research Group. J Clin Epidemiol 1992; 45:683–692. [DOI] [PubMed] [Google Scholar]
- 14. Agarwal SK, Alonso A, Whelton SP, Soliman EZ, Rose KM, Chamberlain AM, et al. Orthostatic change in blood pressure and incidence of atrial fibrillation: results from a bi-ethnic population based study. PLoS One 2013;8:e79030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Loehr LR, Rosamond WD, Chang PP, Folsom AR, Chambless LE. Heart failure incidence and survival (from the Atherosclerosis Risk in Communities study). Am J Cardiol 2008; 101:1016–1022. [DOI] [PubMed] [Google Scholar]
- 16. Psaty BM, Kuller LH, Bild D, Burke GL, Kittner SJ, Mittelmark M, et al. Methods of assessing prevalent cardiovascular disease in the Cardiovascular Health Study. Ann Epidemiol 1995; 5:270–277. [DOI] [PubMed] [Google Scholar]
- 17. Ogedegbe G, Pickering T. Principles and techniques of blood pressure measurement. Cardiol Clin 2010; 28:571–586. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Mariotti G, Alli C, Avanzini F, Canciani C, Di Tullio M, Manzini M, et al. Arm position as a source of error in blood pressure measurement. Clin Cardiol 1987; 10:591–593. [DOI] [PubMed] [Google Scholar]
- 19. Philbrick JT, Shumate R, Siadaty MS, Becker DM. Air travel and venous thromboembolism: a systematic review. J Gen Intern Med 2007; 22:107–114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Gibbons CH, Freeman R. Delayed orthostatic hypotension: a frequent cause of orthostatic intolerance. Neurology 2006; 67:28–32. [DOI] [PubMed] [Google Scholar]
- 21. Naschitz JE, Rosner I. Orthostatic hypotension: framework of the syndrome. Postgrad Med J 2007; 83:568–574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Casiglia ED, Tikhonoff V, Caffi S, Boschetti G, Giordano N, Guidotti F, et al. Orthostatic hypotension does not increase cardiovascular risk in the elderly at a population level. Am J Hypertens 2014; 27:81–88. [DOI] [PubMed] [Google Scholar]