Blood pressures in healthy dogs have been evaluated in several studies1–4 via direct (intra-arterial) and indirect (Doppler ultrasonographic or oscillometric) methods. The effect of body position on SAP measurements in humans has been investigated extensively, and specific postural guidelines for blood pressure measurements of subjects with various medical conditions have been established.5–7
The effect of body position on blood pressure measurements in dogs has not been reported, with the exception of a study8 involving Irish Wolfhounds (n = 158), in which measurements obtained via oscillometry were not significantly different when the dogs were standing rather than recumbent. Studies in other species include one9 that showed giraffes have a higher mean carotid arterial blood pressure when their heads are in the raised position than when their heads are lowered. Another study10 involving anesthetized foals revealed a significant difference in direct measurements of arterial blood pressure obtained when the foals were in a hoisted versus laterally recumbent position.10
In veterinary practice, direct measurement of arterial blood pressure in dogs is commonly performed for routine health screening as well as for monitoring of hypertension.11 Reliable indirect measurements of blood pressure in dogs are important for accurate assessment of the risk of target organ damage due to hypertension.11 The objectives of the study reported here were to determine whether a difference existed in SAP measurement by Doppler ultrasonography in dogs in sitting versus laterally recumbent body positions and to establish the repeatability of blood pressure measurements in each of these positions. The hypothesis was that blood pressure measurements would be higher and more variable when dogs were sitting than when they were laterally recumbent.
Materials and Methods
Animals—Dogs were enrolled between November 2010 and January 2011 from the patient population examined at the Matthew J. Ryan Veterinary Hospital of the University of Pennsylvania as well as from the population of dogs owned by faculty, staff, and students. Potential candidates were identified when the attending clinician contacted the investigators or the investigators contacted the attending clinician regarding dogs available to participate in the study. Dogs that were imminently being discharged, unavailable because of diagnostic testing or treatment, or found to meet one of the exclusion criteria before measurements were attempted were not further pursued for inclusion. For the other dogs, the study protocol was reviewed with the dog's owner and owner consent was obtained prior to participation. The study protocol was approved by the University of Pennsylvania Institutional Animal Care and Use Committee.
Dogs were excluded from the study when they were < 1 year of age, unable to assume a sitting position, or had been sedated or anesthetized within 24 hours prior to blood pressure measurement. They were also excluded when they could not be restrained for blood pressure measurements or their SAP was higher than the upper limit measurable with the manometer (300 mm Hg).
Data collection—Information was collected on potential factors that might affect blood pressure measurements in the body positions. Dogs were classified as small (< 10 kg [22 lb]), medium (10 to 25 kg [22 to 55 lb]), or large (> 25 kg [> 55 lb]). Body condition score was assigned by 1 investigator (DAR), who used a 9-point scale12 (1 = emaciated; 5 = ideal; and 9 = grossly obese). Dogs owned by hospital staff that spent > 1 workday/wk at the hospital (accompanying their owners to work) were classified as having frequent hospital visits, and dogs spending 1 day/wk or less at the hospital were classified as having infrequent hospital visits.
Signs of anxiety in the dogs were subjectively scored by 1 investigator (DAR) during a 5-minute transition period following the blood pressure measurements obtained in the first body position. Scores were assigned on a scale of 1 to 3 (1 = no signs of anxiety; 2 = some signs of anxiety; and 3 = pronounced signs of anxiety), in which signs of anxiety were defined as panting, trembling, or resistance to restraint throughout blood pressure measurement in the first body position.
Owners were asked to complete a detailed questionnaire to provide information about their dog's past or current disease status and medications currently administered. Information requested included whether the dog currently had diabetes mellitus, hyperadrenocorticism, chronic or acute renal disease, protein-losing nephropathy, proteinuria, adrenal gland tumor or nodule, hypertension, primary hyperaldosteronism, hypothyroidism, or acromegaly and whether any of these had been previously diagnosed. Owners were also asked to list any other diseases not specifically mentioned as well as all current medications and their dose, route, frequency, and duration of administration.
Dogs were classified as healthy when they had no known existing illness other than mild localized disease (eg, superficial pyoderma or degenerative joint disease) not causing clinical signs of systemic illness. This distinction between mild localized disease and systemic disease was made by the same investigator (DAR) for all enrolled dogs prior to blood pressure measurement.
Study design—A randomized, crossover study design was used. For each dog, blood pressure was measured in sitting and recumbent body positions, and these measurements were compared with one another. The position of the first set of blood pressure measurements was determined randomly for each dog by a coin flip.
Blood pressure measurements—All blood pressure measurements were obtained in the same quiet, isolated room by the same investigator (DAR), with the assistance of a certified veterinary technician, veterinary student, or other colleague.11 Owners were not permitted to be in the room with their dogs while measurements were made.
A Doppler probe was used for all measurements.a The same Doppler unit, probe, and sphygmomanometer were used throughout the study as well as the same 6 cuffs for various limb circumferences.b,c Limb circumference, measured at the mid to distal aspect of the antebrachium, was used to select a cuff with a width that was 30% to 40% of the circumference of the limb.11
For each dog, the palmar surface of the paw was shaved and the region cleansed with 70% isopropyl alcohol. Ultrasound coupling gel was applied to the probe to improve contact. After a pulse was audibly identified with the Doppler probe, the cuff was inflated until the pulse was no longer heard. The pressure in the cuff was then slowly released until the pulse was again heard. The blood pressure at which the pulse was first consistently audible was recorded.
When dogs were in the laterally recumbent position, the nondependent forelimb, most often the right forelimb, was used for blood pressure measurement. For those measurements, dogs were positioned with the left side down unless there was a physical obstruction to blood pressure measurement in the right forelimb, such as a bandage or IV catheter. The same limb was used for blood pressure measurement in both body positions.
For all measurements, the cuff was positioned at the estimated level of the right atrium.11 For measurements in the sitting position, this was accomplished by flexing the elbow joint and partially extending the shoulder of the measured limb with the dog sitting on its hind limbs, with the forelimb not used for measurement bearing weight in a vertical position. Measurements were repeated 7 times in each position, with the cuff completely deflated between each measurement.11 After measurements in the first position were completed, a 5-minute period of acclimation to the second position was provided before proceeding with the second set of measurements.
For data analysis, the first of 7 measurements in each position was discarded.11 Mean SAP was calculated for the remaining 6 measurements obtained in each position. The positional difference in SAP was defined as the difference between the mean blood pressures measured in the sitting and recumbent positions. A dog was defined as having moderate systolic hypertension when the mean of its 6 measurements in the laterally recumbent position was between 160 and 179 mm Hg or severe systolic hypertension when the mean of its 6 measurements in the laterally recumbent position was ≥ 180 mm Hg.11
Statistical analysis—Normality of data distribution was determined via the D'Agostino-Pearson test unless the sample size was too small (as identified by the statistical software), in which situation, the Kolmogorov-Smirnov test was used. For parametric tests requiring equality of variance, the Levene test for equality of variances was first used. Normally distributed data are reported as mean ± SD, and the remainder are reported as median and range.
The paired-samples t test was performed to compare mean sitting and laterally recumbent SAP values. The Wald test, which evaluates repeatability by testing the equality of within-subject coefficients of variation, was used to test the hypothesis that blood pressure measurements in the sitting position were more variable than blood pressure measurements in lateral recumbency.13 The mean coefficient of variation, which is another estimate of the degree of variability, was also calculated for dogs in each body position. Mean coefficient of variation values closer to zero indicate less variability in measurements. The intraclass correlation coefficient, which is a descriptive statistic used to assess the repeatability of repeated quantitative measurements, was also calculated for each position as an estimate of test precision.13 Intraclass correlation coefficient values closer to 1 indicate higher repeatability.
Positional difference data were also compared with dogs stratified into groups by their characteristics (eg, reproductive status, breed, anxiety score, body size, BCS, diagnosis, or health status). When ≥ 3 dogs represented a category of interest, positional differences for those dogs were compared by 1-way ANOVA when assumptions (normality and equality of variance) were satisfied. Otherwise, groups were compared via the Kruskal-Wallis test (for comparisons of > 2 groups) or Mann-Whitney test (for comparison of 2 groups).
Simple linear regression analysis was used to evaluate correlations (r) between continuous variables (ie, age and body weight) and positional difference in SAP. The Spearman rank correlation coefficient (ρ) was calculated to evaluate correlations between ordinal variables (ie, BCS and anxiety score) and positional change in SAP A value of P < 0.05 was considered significant for all analyses. With the exception of Wald test calculations, which were performed manually, a commercially available statistical software programd was used for all analyses.
Results
Animals—Consent was obtained to enroll 57 dogs in the study, of which 6 were excluded, leaving 51 for inclusion. Four dogs were excluded prior to data collection because of an inability to restrain them. The fifth dog was excluded because of an inability to achieve appropriate positioning as a result of limb amputation, and the sixth was excluded because blood pressure measurements were higher than the upper limit measurable with the manometer.
The 51 dogs included the following breeds: mixed (n = 11), Golden Retriever (5), Labrador Retriever (5), and Australian Shepherd Dog, Beagle, Bichon Frise, Boxer, Soft-Coated Wheaton Terrier, Shih Tzu, Standard Poodle, and Whippet (2 each). The remainder represented 1 each of various other pure breeds. Mean ± SD age was 6.2 ± 3.7 years, mean body weight was 20.0 ± 11.8 kg (44.0 ± 26.0 lb), and median BCS was 5/9 (range, 3/9 to 8/9). The median anxiety score was 2. Fourteen of 51 (27%) dogs were classified as having frequent hospital visits.
Twenty-five (49%) dogs were classified as healthy Healthy dogs had diagnoses that included allergic skin disease (n = 4), mitral valve degeneration without cardiomegaly (2), and elbow dysplasia, benign cyst, pyoderma and otitis externa (concurrently), and a history of grape ingestion with unremarkable physical examination and serum biochemical findings (1 each). The remainder of the healthy dogs had been brought to the hospital for elective ovariohysterectomy (n = 1) or were owned by faculty, staff, or students who volunteered their dogs for the study (14).
Twenty-six (51%) dogs were classified as unhealthy. Their diagnoses included lymphoma (n = 6), possible hypothyroidism (2), adrenal gland mass or nodule of unknown type (2), pituitary-dependent hyperadrenocorticism (2), protein-losing nephropathy with hypertension and azotemia (2), hypoadrenocorticism (2), megaesophagus (2), and various other diseases (1 each). Some dogs had > 1 diagnosis. The 2 dogs with possible hypothyroidism had other illnesses (one with confirmed hyperadrenocorticism and tracheal collapse and the other with an adrenal mass) that resulted in their classification as unhealthy.
Medications had been administered to 29 (57%) dogs within 24 hours prior to blood pressure measurements. These medications included corticosteroids (n = 9), famotidine (9), an NSAID (8), ondansetron (8), metronidazole (5), omega-3 fatty acids (4), enalapril (4), proton pump inhibitor (4), tramadol (4), metoclopramide (3), vincristine (3), trimethoprim-sulfamethoxazole (3), and several additional medications (< 3/medication). Many dogs were receiving > 1 medication.
Blood pressure measurements—Six serial SAP measurements were recorded for each of the 51 dogs in sitting and recumbent body positions, yielding 612 SAP measurements for analysis. Mean SAPs in the sitting and laterally recumbent body positions were 172.1 ± 33.3 mm Hg and 147.0 ± 24.6 mm Hg, respectively. Mean SAP in the sitting position was significantly (P < 0.001) higher than in the recumbent position. Mean positional difference in SAP was 25.1 ± 28.5 mm Hg (range, −44 to 94 mm Hg).
Most dogs (44/51 [86%]) had a higher SAP reading when sitting than when recumbent. Seven of 51 (14%) dogs had a lower SAP reading when sitting. The mean positional difference in SAP for dogs that had a positive positional difference in SAP (increase in blood pressure reading when sitting) was 32.0 ± 23.5 mm Hg. The mean positional difference in SAP for dogs that had a negative positional difference in SAP (decrease in blood pressure when sitting) was −18.4 ± 16.0 mm Hg.
Repeatability of the blood pressure measurements was significantly (P < 0.001; Wald test) worse in the sitting position than in the laterally recumbent position. The intraclass correlation coefficient was 0.82 for blood pressure readings in the sitting position and 0.95 for blood pressure readings in the laterally recumbent position. Similarly, the mean coefficients of variation for blood pressure measured in sitting and laterally recumbent positions were 0.06 and 0.03, respectively.
No significant effect on positional difference in SAP was identified for any of the dog characteristics assessed, including severity of hypertension, first position of measurement, reproductive status, breed, frequency of hospital visits, body weight, BCS, anxiety score, health status, or medications (Table 1). There was also no significant correlation between positional difference in SAP and age (r = 0.26; P = 0.07), anxiety score (ρ = −0.15; P = 0.29), BCS (ρ = 0.001; P = 0.99), or body weight (r = 0.23; P = 0.10).
Differences between repeated Doppler oscillometric measurements of SAP (6 values/position) in dogs in sitting versus laterally recumbent body positions.
| Characteristic | No. of dogs | Difference (mm Hg) | P value |
|---|---|---|---|
| All dogs | 51 | 25.1 ± 28.5 | — |
| Blood pressure classification | 0.06 | ||
| Severe hypertension (≥ 180 mm Hg) | 4 | –2.5 (−18.1 to 17.0) | — |
| Moderate hypertension (160–179 mm Hg) | 11 | 23.1 ± 33.5 | — |
| Normotensive (< 160 mm Hg) | 36 | 28.6 ± 27.0 | — |
| First position of measurement | 0.21 | ||
| Sitting | 25 | 28.5 ± 20.7 | — |
| Recumbent | 26 | 21.8 ± 34.5 | — |
| Reproductive status | 0.78 | ||
| Spayed female | 29 | 25.9 ± 25.7 | — |
| Sexually intact female | 2 | 23.4 (17.0–29.9) | — |
| Neutered male | 16 | 26.3 ± 37.9 | — |
| Sexually intact male | 4 | 12.1 (9.8–26.0) | — |
| Breed | — | ||
| Golden Retriever | 5 | 34.9 ± 44.3 | 0.43 |
| Non–Golden Retriever | 46 | 25.5 ± 26.8 | — |
| Labrador Retriever | 5 | 45.4 ± 20.8 | 0.09 |
| Non–Labrador Retriever | 46 | 22.8 ± 28.5 | — |
| Frequency of hospital visits* | 0.29 | ||
| Frequent | 14 | 18.1 ± 30.1 | — |
| Infrequent | 37 | 27.7 ± 27.9 | — |
| Body weight | 0.11 | ||
| < 10 kg | 15 | 13.7 (−44.0 to 93.9) | — |
| 10–25 kg | 19 | 24.1 ± 24.1 | — |
| > 25 kg | 17 | 34.0 ± 30.7 | — |
| BCS | 0.85 | ||
| 1–3 | 3 | 31.1 (9.8 to 51.6) | — |
| 4–6 | 37 | 24.8 ± 29.1 | — |
| 7–9 | 11 | 24.6 ± 30.1 | — |
| Anxiety score | 0.52 | ||
| 1 (no signs of anxiety) | 10 | 29.8 ± 38.6 | — |
| 2 (some signs of anxiety) | 22 | 28.1 ± 25.8 | — |
| 3 (pronounced signs of anxiety) | 19 | 19.1 ± 26.0 | — |
| Health status | — | ||
| Healthy | 25 | 19.8 ± 31.3 | 0.20 |
| Unhealthy | 26 | 30.2 ± 25.2 | — |
| Lymphoma | 6 | 41.7 ± 24.8 | 0.13 |
| Nonlymphoma | 45 | 22.9 ± 28.5 | — |
| Medications administered in preceding 24 h† | — | ||
| Corticosteroid | 9 | 32.5 ± 26.5 | 0.40 |
| Famotidine | 9 | 28.2 ± 24.5 | 0.72 |
| NSAID | 8 | 32.1 ± 34.0 | 0.45 |
| Ondansetron | 8 | 41.6 ± 21.7 | 0.07 |
| Metronidazole | 5 | 32.1 ± 12.5 | 0.57 |
| Omega-3 fatty acid | 4 | 37.3 (–3.7 to 63.0) | 0.51 |
| Enalapril | 4 | 28.8 (–3.7 to 63.0) | 0.79 |
| Proton pump inhibitor | 4 | 49.5 (18.5 to 67.3) | 0.06 |
| Tramadol | 4 | 6.7 (–5.3 to 93.7) | 0.43 |
| Metoclopramide | 3 | 51.6 (17.0 to 68.6) | 0.16 |
| Vincristine | 3 | 67.3 (6.3 to 68.6) | 0.23 |
| Trimethoprim-sulfamethoxazole | 3 | 47.4 (38.1 to 68.6) | 0.06 |
Data are reported as mean ± SD if normally distributed or median (range) if not.
Dogs owned by hospital staff that spent > 1 workday/wk at the hospital (accompanying their owners to work) were classified as having frequent hospital visits, and dogs spending 1 workday/wk or less at the hospital were classified as having infrequent visits.
Referent group is dogs not treated with the drug.
— = Not calculated.
Values of P < 0.05 were considered significant. Seven SAP readings were obtained in each position for each dog; however, the first value was excluded from calculations.
Forty-seven (92%) dogs had their left side down when blood pressure measurements were performed for the laterally recumbent position, and 4 (8%) dogs had their right side down. Mean SAP measured with the left side down (26.9 ± 28.6 mm Hg) was not significantly (P = 0.07) different than that measured with the right side down (median, 3.1 mm Hg; range, −18.1 to 26 mm Hg).
Twenty-eight of 51 (55%) dogs had blood pressure measurements that were classified differently when measured in the 2 body positions. Ten of 36 (28%) dogs that had unremarkable SAP readings when in the laterally recumbent position had readings consistent with moderate systolic hypertension when in the sitting position, and 9 of 36 (25%) dogs that had unremarkable SAP readings when laterally recumbent had readings consistent with severe systolic hypertension when sitting. Seven of 11 dogs that had SAP readings consistent with moderate systolic hypertension based in the laterally recumbent position had values consistent with severe systolic hypertension in the sitting position, and 1 of 11 dogs with readings suggestive of moderate systolic hypertension when laterally recumbent had readings considered unremarkable when sitting. Of the 4 dogs with SAP readings suggestive of severe systolic hypertension when laterally recumbent, 3 continued to be classified as severely hypertensive when in the sitting position and the other had readings consistent with moderate systolic hypertension when sitting.
Discussion
Doppler oscillometric measurements of SAP in 86% of dogs in present study were significantly higher when the dogs were in a sitting rather than laterally recumbent position. The mean difference between positions was 25 mm Hg. It follows that body position affects SAP measurements in dogs and that, in most dogs, indirect measurement of SAP when dogs are sitting will yield values significantly higher than would be acquired when dogs are laterally recumbent.
The difference of 25 mm Hg in a blood pressure measurement may have important clinical relevance, as it could change the classification of hypertension and its potential likelihood to cause target organ damage.11 Indeed, 55% of the study dogs had blood pressure measurements that led to different classifications between the 2 body positions, and 53% of those that were classified as having an unremarkable blood pressure were classified as hypertensive when measured in the sitting position. According to the 2007 American College of Veterinary Internal Medicine consensus statement, the risk of future target organ damage is mild with an SAP of 150 to 159 mm Hg, moderate with an SAP of 160 to 179 mm Hg, and severe with an SAP ≥ 180 mm Hg.11 As such, the perceived risk of future target organ damage could, in many situations, change when SAP is measured in sitting versus laterally recumbent dogs. Inappropriate assessment of SAP and the risk of future target organ damage may lead to inappropriate therapeutic intervention. We therefore recommend that body position be recorded at the time of blood pressure measurement, and follow-up measurements should be performed in the same body position as they were in the past.
A study limitation is that indirect blood pressure measurements were not compared with direct blood pressure measurements. Therefore, the body position in which Doppler ultrasonographic measurement would best correlate with direct blood pressure measurements is unknown. A follow-up study measuring SAP directly in different positions concurrently with indirect blood pressure measurements would help to determine the body position in which indirect SAP measurements are most accurate. However, findings in the present study suggested that SAP measurements obtained in laterally recumbent dogs are more reproducible than measurements obtained in sitting dogs.
One possible explanation for higher blood pressure measurements in sitting versus laterally recumbent dogs is that fluctuations in the muscle tone of the extended limb could alter the forces applied by the cuff within the measurement period. Invasive methods of blood pressure measurement would be needed to evaluate this hypothesis further.
No associations were identified between positional difference in SAP and any of the signalment or historical factors evaluated. This may have been because those factors truly do not influence positional difference in SAP or because the study sample size was too small for detection of such associations once the data were grouped within factors.
ABBREVIATIONS
| BCS | Body condition score |
| SAP | Systolic arterial blood pressure |
Doppler Flow Detector, model 811-B, Parks Medical Electronics Inc, Aloha, Ore.
Sphygmomanometer, model 148 standard, Omron Healthcare Inc, Lake Forest, Ill.
Critikon Classic-cuf (sizes 1 to 5) and Dura-cuf (larger cuff sizes), GE Healthcare, Piscataway, NJ.
MedCalc for Windows, version 11.5.0, MedCalc Software, Mariakerke, Belgium.
References
1. Bodey AR, Michell AR. Epidemiological study of blood pressure in domestic dogs. J Small Anim Pract 1996; 37: 116–125.
2. Chalifoux A, Dallaire A, Blais D, et al. Evaluation of the arterial blood pressure of dogs by two noninvasive methods. Can J Comp Med 1985; 49: 419–423.
3. Coulter DB, Keith JC Jr. Blood pressures obtained by indirect measurement in conscious dogs. J Am Vet Med Assoc 1984; 184: 1375–1378.
4. Remillard RL, Ross JN, Eddy JB. Variance of indirect blood pressure measurements and prevalence of hypertension in clinically normal dogs. Am J Vet Res 1991; 52: 561–565.
5. Nardo CJ, Lloyd EC, Light KC, et al. Descriptive epidemiology of blood pressure response to change in body position. Hypertension 1999; 33: 1123–1129.
6. Davis BR, Langford HD, Blaufox MD, et al. The association of postural changes in systolic blood pressure and mortality in persons with hypertension: the Hypertension Detection and Follow-up Program experience. Circulation 1987; 75: 340–346.
7. Chobanian AV, Bakris GL, Black HR, et al. Seventh report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure. Hypertension 2003; 42: 1206–1252.
8. Bright JM, Dentino M. Indirect arterial blood pressure measurement in nonsedated Irish Wolfhounds: reference values for the breed. J Am Anim Hosp Assoc 2002; 38: 521–526.
9. Brøndum E, Hasenkam JM, Secher NH, et al. Jugular venous pooling during lowering of the head affects blood pressure of the anesthetized giraffe. Am J Physiol Regul Integr Comp Physiol 2009; 297: R1058–R1065.
10. Braun C, Trim CM, Eggleston RB. Effects of changing body position on oxygenation and arterial blood pressures in foals anesthetized with guaifenesin, ketamine and xylazine. Vet Anaesth Analg 2009; 36: 18–24.
11. Brown S, Atkins C, Bagley R, et al. Guidelines for the identification, evaluation and management of systemic hypertension in dogs and cats; 2007 ACVIM Forum Consensus Statement. J Vet Intern Med 2007; 21: 542–558.
12. Laflamme DP. Development and validation of a body condition score system for dogs. Canine Pract 1997; 22(4):10–15.
13. Shoukri MM, Colak D, Kaya N, et al. Comparison of two dependent within subject coefficients of variation to evaluate the reproducibility of measurement devices. BMC Med Res Methodol 2008; 8: 24.
