(Åblad, Nilsen). This study was performed at Department of Basic Animal and Veterinary Sciences, Faculty of Life Sciences, University of Copenhagen, Denmark, Din Veterinär Animal Hospital, Helsingborg, Sweden and Blå Stjärnans Animal Hospital, Gothenburg, Sweden.
Heart Rate, Heart Rate Variability, and Arrhythmias in Dogs with Myxomatous Mitral Valve Disease
Version of Record online: 13 DEC 2011
Copyright © 2011 by the American College of Veterinary Internal Medicine
Journal of Veterinary Internal Medicine
Volume 26, Issue 1, pages 76–84, January-February 2012
How to Cite
Rasmussen, C.E., Falk, T., Zois, N.E., Moesgaard, S.G., Häggström, J., Pedersen, H.D., Åblad, B., Nilsen, H.Y. and Olsen, L.H. (2012), Heart Rate, Heart Rate Variability, and Arrhythmias in Dogs with Myxomatous Mitral Valve Disease. Journal of Veterinary Internal Medicine, 26: 76–84. doi: 10.1111/j.1939-1676.2011.00842.x
Preliminary results of this study were presented as a research abstract for the 2011 American College of Veterinary Internal Medicine Forum in Denver, CO.
- Issue online: 10 JAN 2012
- Version of Record online: 13 DEC 2011
- Manuscript Accepted: 5 OCT 2011
- Manuscript Revised: 29 AUG 2011
- Manuscript Received: 24 FEB 2011
- Danish Council of Independent Research|Medical Sciences. Grant Numbers: 271-08-0998, 271-05-03335
- Synergy in Human and Animal Research
- Cavalier King Charles Spaniel;
- Mitral valve prolapse
Autonomic modulation of heart rhythm is thought to influence the pathophysiology of myxomatous mitral valve disease (MMVD).
(1) Holter-derived variables reflecting autonomic modulation of heart rhythm change with MMVD severity in Cavalier King Charles Spaniels (CKCS); (2) Holter-derived variables can identify MMVD severity in CKCS; and (3) Holter-derived variables in CKCS in congestive heart failure (CHF) secondary to MMVD differ from those in dogs of other breeds in CHF.
Ninety privately owned dogs: 70 CKCS with variable MMVD severity and 20 non-CKCS in CHF secondary to MMVD.
Dogs were prospectively recruited and divided into 5 MMVD severity groups based on history, breed, and physical and echocardiographic examination findings. Holter-derived variables included heart rate variability (HRV), heart rate (HR), and arrhythmia evaluated from 24-hour Holter recordings.
In CKCS, 18 of 26 HRV (all P < .0002) and 3 of 9 arrhythmia (all P < .0004) variables decreased with increasing MMVD, whereas minimum and mean HR (all P < .0001) increased with increasing MMVD severity. An arrhythmia variable representing sinus arrhythmia (“premature normals”) (P < .0001) and the HRV variable triangular index (TI) (P < .0001) could distinguish CKCS with moderate or severe mitral regurgitation from CKCS in CHF in specific intervals. Among dogs in CHF, Holter-derived variables did not differ among breeds.
Conclusions and Clinical Importance
In CKCS, Holter-derived variables changed with MMVD severity. “Premature normals” and TI showed diagnostic potential. Breed differences were not seen among dogs in CHF secondary to MMVD.
angiotensin converting enzyme inhibitor
analyses of variance
atrial premature complex
area under the curve
beats per minute
congestive heart failure
Cavalier King Charles Spaniel
normalized high frequency
heart rate variability
interventricular septal thickness in diastole
interventricular septal thickness in systole
left atrium to aortic root ratio
Department of Basic Animal and Veterinary Sciences, Faculty of Life Sciences, University of Copenhagen, Denmark
normalized low frequency
left ventricular internal dimension in diastole
left ventricular internal dimension in systole
left ventricular free wall thickness in diastole
left ventricular free wall thickness in systole
mean of all NN-intervals
myxomatous mitral valve disease
mitral valve prolapse
interval between two QRS complexes of sinus or supraventricular origin
proximal isovelocity surface area
% of successive NN-intervals that differ more than 50 ms
square root of the mean squared differences of successive NN-intervals
receiver operating characteristic
standard deviation of the NN-intervals
ultra low frequency
very low frequency
ventricular premature complex
Heart rate variability (HRV) denotes rhythmic variations in intervals between adjacent QRS complexes (NN-intervals). The rhythmic variations reflect the sinoatrial node response to autonomic tone, thus autonomic modulation of heart rhythm. HRV is measured from the ECG and can be divided into frequency- and time-domain analysis. Frequency-domain analysis estimates how variance in heart rate (HR) is distributed as a function of frequency. The total power (TP) represents all frequencies and is divided into four frequency bands. The high frequency band (HF) is influenced by respiration and is known to reflect parasympathetic influence on the sinoatrial node, whereas the low frequency band (LF) reflects both parasympathetic and sympathetic activities.[1, 3, 4] The ultra low frequency (ULF) and very low frequency (VLF) bands are influenced by more long-term regulatory mechanisms such as the renin-angiotensin system, thermoregulation, and physical activity.[1, 5, 6] Time-domain analysis is based directly on NN-interval lengths, is simpler to calculate, and has higher reproducibility than frequency domain variables.[7, 8] The time domain variable RMSSD (square root of mean squared differences of successive NN-intervals) and pNN50 (percentage of successive NN-intervals differing more than 50 ms) represent parasympathetic activity and correlate with HF, whereas SD (standard deviation of NN-intervals) and TI (triangular index = total number of NN-intervals/maximum number of NN-intervals of equal length) represent overall HRV and correlate with TP.
Decreased HRV is reported in people in congestive heart failure (CHF) secondary to various cardiac diseases and often is explained as a sympathovagal imbalance.[2, 9-11] Myxomatous mitral valve disease (MMVD) is the most common naturally occurring heart disease in dogs. The Cavalier King Charles Spaniel (CKCS), a dog breed at high risk for developing MMVD,[12, 13] has decreased HRV measured on short-term ECG in CHF stages of MMVD.[14, 15] In early stages of MMVD, however, HRV measured on short-term ECG in Dachshunds, Poodles, Beagles, and CKCS seems to increase with increasing mitral valve prolapse (MVP), which is an early sign of MMVD. The increased HRV in early stages of MMVD is thought to be caused by hypervagal autonomic dysfunction. In addition, HRV measured in a 6-hour nightly period in Dachshunds increased with increasing MVP; however, the same study found a weak negative correlation between MVP and HRV measured on short-term ECG. A recent study showed no association between MVP and HRV measured over 24 hours in clinically healthy CKCS, Dachshunds, and Cairn Terriers, and finally, Beagles with experimentally induced mild mitral regurgitation (MR) had decreased HRV. Conflicting results regarding MVP and HRV also have been found in people.[20-22]
Cardiac arrhythmias are not common in dogs in early stages of MMVD,[16, 18] although an increased frequency of supraventricular arrhythmias has been associated with increasing MVP in young Dachshunds and with left atrial enlargement in older dogs with advanced stages of MMVD. The frequency of ventricular arrhythmias was higher in dogs in CHF secondary to MMVD compared with control dogs and ventricular arrhythmias tended to be associated with intra-myocardial arterial narrowing, myocardial interstitial fibrosis, and poor systolic function in dogs in CHF mainly secondary to MMVD.
The aims of this study were to investigate if Holter-derived variables reflecting autonomic modulation of heart rhythm changed with and could identify MMVD severity in CKCS. In addition, we investigated if Holter-derived variables from CKCS in CHF differed from those of dogs of other breeds in CHF secondary to MMVD. Holter-derived variables included HRV, HR, and arrhythmias evaluated from 24-hour Holter monitoring.
Materials and Methods
Privately owned dogs >4 years of age were prospectively included in the study. Dogs did not receive medication, except dogs in CHF, which were allowed medical CHF treatment. Dogs were not included if they had < 20 hours of readable data on the Holter recording or signs of clinically relevant systemic or organ-related disease, other than MMVD, based on history, clinical examination, echocardiography, serum biochemistry, CBC, or some combination of these. Dogs were divided into 5 MMVD severity groups (dog group) based on history, breed, clinical examination, and echocardiography: (1) CKCS with minimal or no MR (no clinical signs and MR jet ≤15% of the left atrial area), (2) CKCS with mild MR (no clinical signs and MR jet between 20 and 50%), (3) CKCS with moderate or severe MR (no clinical signs and MR jet >50%), (4) CKCS in CHF (clinical signs of CHF responsive to furosemide treatment and MR jet >50%), and (5) non-CKCS in CHF (other breeds of dogs with clinical signs of CHF responsive to furosemide treatment and MR jet >50%). All dogs were examined between January 2008 and May 2010 and only 1 examination from each dog was included in the study. Dogs in CHF were recruited from Department of Basic Animal and Veterinary Sciences, Faculty of Life Sciences, University of Copenhagen, Denmark (LIFE), Din Veterinär Animal Hospital, Sweden, and Blå Stjärnans Animal Hospital, Sweden, whereas CKCS with minimal or no, mild, or moderate to severe MR were recruited from LIFE. All owners gave their consent, and the study was approved by the Danish Animal Welfare Division and the Local Ethical Committee in Gothenburg, Sweden.
Dogs were examined by means of a standardized protocol including blood sampling, physical examination, and echocardiography, in order of presentation. Relative to this examination, Holter recording was performed the day before (n = 4), the same day (n = 62), or within 1 month of presentation (n = 24). Dogs in CHF received the same treatment during echocardiographic examination and Holter monitoring.
None of the dogs was sedated during echocardiography and owners were present to calm the dogs. Two-dimensional echocardiography, with a Vivid i echocardiograph,1 was performed by 2 operators (LHO [n = 79] and TF [n = 11]). All echocardiographic variables were recorded and later evaluated by 1 individual masked to the identity of the dog (LHO). The left ventricular dimensions were measured by use of M-mode echocardiography performed on the short axis view. Left ventricular dimensions in this study are presented as percent increase from the expected normal dimensions. The degree of MVP was assessed from the right parasternal long axis 4-chamber view. The left atrial to aortic root ratio (LA/Ao) was measured in a short-axis view at the level of the aortic valve. MR jet size and proximal isovelocity surface area (PISA) were estimated with color Doppler mapping over the mitral valve area.[29, 30] MR was semiquantitatively estimated as the percentage area (to the nearest 5%) of the left atrium occupied by the largest jet, and PISA was measured as previously described.
A Holter recorder2 was placed on the dogs with a 2 lead precordial placement of electrodes.3 Before electrode placement, an area was prepared by shaving and cleaning the skin with alcohol. An elastic bandage and a specially designed vest were used to secure the Holter recorder and leads to the dog. All dogs wore the Holter monitor for 24 hours in their home environment and the owners noted activities in a diary.
Heart Rate, Heart Rate Variability, and Arrhythmia Analysis
A standardized protocol for semiautomatic analyses was performed with commercially available software designed for people.4,5 The observer was blinded to the identity of the dog and used the same procedure and software settings as previously described. The automatic HRV analysis excluded complexes of ventricular origin, but included intervals between adjacent QRS complexes (NN-intervals) comprising sinus complexes, atrial premature complexes (APCs), and supraventricular tachycardia (SVT). Holter recordings with <90% valid NN-intervals in an analysis period were excluded from HRV analysis. All HR, HRV, and arrhythmia variables were analyzed over a 24-hour period. During the 24-hour period, HRV variables also were analyzed in a 6-hour nightly period starting 30 minutes after the dog fell to sleep (bedtime). The diary was used to identify bedtime of the dog, or if not noted in the diary, bedtime was defined as the 1st period of 10 minutes after 9:00 pm with a HR below mean HR during Holter recording.
The frequency domain HRV variables comprised TP (0–0.4 Hz), ULF (0–0.00333 Hz), VLF (0.00333–0.04 Hz), LF (0.04–0.15 Hz), and HF (0.15–0.4 Hz). LF and HF were adjusted for TP and lower frequencies by a normalization equation (LFn = LF / (TP − ULF − VLF) × 100 and HFn = HF/(TP − ULF − VLF) × 100) .6 The HF to LF ratio (HF/LF) also was used as a frequency domain HRV variable. The following time domain HRV variables were included in the study: MEAN (average NN-interval), pNN50, RMSSD, SD, and TI. HRV variables were suffixed by 24 or 6 indicating the measuring period (24 or 6 hours). Minimum, mean, and maximum HR during Holter recording were used as HR variables. Arrhythmia variables included bradycardia (>4 successive sinus beats at a HR <45 beats per minute [bpm]), “dropped beats” (corresponds to a long diastolic pause, represents sinus arrhythmia and is defined as a NN-interval >180% longer than the previous NN-interval), sinus pauses (NN-intervals >2.0 seconds), “premature normals” (represents sinus arrhythmia and is defined as a premature sinus complex where premature is defined as a NN-interval 50% shorter than the previous NN-interval), APC (a P wave with abnormal morphology conducting a premature normal-appearing QRS complex), SVT (3 or more APCs at a HR > 150 bpm and with NN-intervals shorter than or equal to the previous NN-interval), sinus tachycardia (a sinus complex followed by 2 or more successive sinus complexes at a HR > 150 bpm and with NN-intervals shorter than or equal to the previous NN-interval), ventricular premature complex (VPC) (a premature wide and bizarre looking QRS complex, not associated with a P wave, but accompanying a large T wave of opposite polarity), ventricular escape complex (wide QRS complexes of different orientation occurring after a sinus pause and not associated with a P wave), fusion complex (normal P wave followed by an intermediately shaped QRS complex), and second-degree atrioventricular (AV) block (a P wave not related to a QRS complex). Second-degree AV blocks were not separated into Mobitz type I and II. APCs and VPCs were included in the statistical analyses as percentage of the total QRS complexes in 24 hours.
Statistical analyses were performed with commercially available software.7 The significance level of P < .05 was Bonferroni adjusted for each type of statistical test.
Group-Wise Comparisons in CKCS
Analyses of variance (ANOVA) were used to test differences among CKCS dog groups in HRV and HR variables. ANOVAs were performed with an HRV or HR variable as response variable, and dog group, age, sex, and interaction between age and dog groups were explanatory variables. If dog group was statistically significant, posthoc group-wise comparisons (Student‘s t-tests) were performed for each response variable. Kruskal–Wallis tests were used to test for differences among CKCS dog groups for arrhythmia variables. If dog group was statistically significant, posthoc group-wise comparisons (Wilcoxon signed rank tests) were performed for each response variable. Spearman's correlations were performed to correlate arrhythmia variables with age and Wilcoxon signed rank tests were used to test if arrhythmia variables showed differences between males and females. If Holter-derived variables differed significantly between dog groups 1 and 2, 2 and 3, and 3 and 4, multiple logistic regressions (MRL) were performed to explore which Holter-derived variables best indicated difference between dog groups. The MRL response variable was dog group and explanatory variables included age, sex, body weight, and Holter-derived variables selected based on results of posthoc group-wise comparisons. In addition, MRL were performed with and without mean HR as a fixed effect. The diagnostic efficacy for all Holter-derived variables that were significant from MRL, were evaluated by receiver operating characteristic (ROC) curves, sensitivity, specificity, and area under the curve (AUC) with a second commercially available software program.8
Association between Holter-Derived Variables and Echocardiographic Indices of MMVD in CKCS
ANOVA was used to test if HRV or HR variables were associated with echocardiographic indices of MMVD. ANOVA models consisted of a HRV or HR variable as response variable and the following explanatory variables: echocardiographic indices of MMVD, age, sex, and interaction between age and echocardiographic indices of MMVD. Spearman's correlations were performed to test association between echocardiographic indices of MMVD and arrhythmia variables. Echocardiographic indices of MMVD were included separately in statistical models and consisted of left ventricular internal dimensions in diastole and systole (LVIDd and LVIDs), MVP, MR, LA/Ao, and PISA.
Difference between CKCS and non-CKCS among dogs in CHF
ANOVA was used to investigate if HRV and HR variables in CKCS in CHF were different from non-CKCS in CHF. A HRV or HR variable was included as a response variable in ANOVA models and dog group, age, sex, and interaction between age and dog groups were explanatory variables. Kruskal–Wallis tests were used to investigate if arrhythmia variables in CKCS in CHF were different from those of non-CKCS in CHF. Among dogs in CHF, Spearman's correlations were used to correlate arrhythmia variables with age and finally, Wilcoxon signed rank test was used to test if an arrhythmia variable showed differences between males and females.
In all of the above ANOVA models, stepwise backward elimination was used until only significant effects remained, and residuals were tested for homogeneity of variation and Gaussian distribution with residual plot and a Shapiro–Wilks test, respectively. If necessary, logarithmic or power transformations were applied.
Characteristics of the 90 included dogs are shown in Table 1. The non-CKCS consisted of 6 Dachshunds, 2 Bichon Frises, and Italian Greyhound, King Charles Spaniel, Jack Russell Terrier, Boston Terrier, Kelpie, Maltese, Norfolk Terrier, Border Collie, Cocker Spaniel, Nova Scotia Duck Tolling Retriever, Bichon Havanais and a mixed breed dog all represented by 1 dog. The treatments of CHF dogs are summarized in Table 2. Two non-CKCS in CHF had sustained atrial fibrillation and were excluded from further statistical analyses of Holter-derived variables. Because of <90% valid NN-intervals, 3 and 5 dogs were excluded from the 24- and 6-hour nightly HRV analyses, respectively. Therefore, the 24-hour HRV analyses include 16 CKCS with minimal or no MR, 15 CKCS with mild MR, 21 CKCS with moderate to severe MR, 15 CKCS in CHF and 17 non-CKCS in CHF. Furthermore, the 6-hour nightly HRV analyses include 16 CKCS with minimal or no MR, 14 CKCS with mild MR, 21 CKCS with moderate to severe MR, 15 CKCS in CHF and 16 non-CKCS in CHF. Sixty-nine percent of all dogs had APCs and 43% had second-degree AV-block. VPCs occurred in 50% of the dogs including, 3 CKCS and 2 non-CKCS with ventricular couplets, 1 CKCS and 1 non-CKCS with ventricular triplets, 2 CKCS and 3 non-CKCS with ventricular bigeminy, and 2 CKCS with ventricular tachycardia. Four CKCS and 3 non-CKCS had fusion complexes and finally 1 non-CKCS had ventricular escape complexes.
|Group||No/minimal MR||Mild MR||Moderate/Severe MR||CHF||CHF|
|Age (years)||5.6 (4.3–6.1)||6.0 (4.8–6.9)||7.11,2 (6.2–8.4)||9.71,2 (7.9–11.7)||11.54 (10.5–13.1)|
|Body weight (kg)||9.7 (8.2–10.7)||9.4 (8.4–10.9)||10.1 (8.6–11.7)||10.5 (8.9–11.7)||10.3 (7.3–13.4)|
|HRex (bpm)||120.0 (107.0–124.0)||118.0 (106.0–120.0)||128.0 (120.0–144.0)||140.01,2 (134.5–148.5)||130.0 (112.0–140)|
|IVSTd (%)||10.6 (−0.1–17.4)||8.3 (−1.3–15.7)||7.1 (−1.6–18.8)||3.8 (−12.0–14.1)||15.24 (9.4–33.0)|
|IVSTs (%)||−1.0 (−13.7–6.5)||−6.1 (−12.0–1.7)||4.8 (−11.8–21.3)||6.1 (−13.4–25.8)||19.7 (7.5–36.0)|
|LVIDd (%)||−0.84 (−4.6–4.4)||−4.0 (−7.9–9.6)||11.51,2 (1.2–21.1)||29.41,2,3 (22.0–46.7)||40 (30.5–50.9)|
|LVIDs (%)||7.5 (−3.8–22.0)||9.6 (2.8–23.4)||14.6 (7.3–29.4)||23.71 (13.6–28.2)||38.3 (17.8–53.6)|
|LVPWd (%)||11.2 (3.3–22.3)||12.0 (4.1–16.4)||14.3 (1.1–24.1)||12.8 (−5.9–25.1)||21.1 (7.8–29.1)|
|LVPWs (%)||−8.9 (−15.2–5.4)||−12.6 (−15.3–5.1)||−5.4 (−10.6–1.9)||−7.0 (−10.5–3.3)||1.8 (−4.2–10.4)|
|LA/Ao||1.2 (1.2–1.3)||1.3 (1.2–1.4)||1.5 (1.3–1.7)||1.81,2,3 (1.5–2.3)||1.9 (1.8–2.0)|
|PISA (mm)||0 (0.0–0.0)||4.0 (2.8–5.0)||8.01,2 (5.5–10.0)||13.51,2,3 (12.0–15.0)||13 (10.3–15.8)|
|FS (%)||30.4 (22.9–35.7)||26.0 (21.8–31.3)||32.32 (27.3–37.2)||36.61,2,3 (34.5–42.0)||36.4 (30.6–41.8)|
|MVP (mm)||4.0 (3.0–5.0)||5.0 (4.0–6.8)||8.01,2 (5.8–10.3)||9.01,2 (8.0–10.75)||9.0 (8.0–12.0)|
|Medical treatment||CKCS n||CKCS %||Non-CKCS n||Non-CKCS %|
|F + ACE-I||3||18.8||1||5.0|
|F + Pimo||0||0||1||5.0|
|F + ACE-I + β-blocker||0||0||1||5.0|
|F + ACE-I + Pimo||5||31.3||6||30.0|
|F + ACE-I + Pimo + S||1||6.3||3||15.0|
|F + ACE-I + Pimo + S + Dig||2||12.5||2||10.0|
|F + Pimo + Dig||1||6.3||0||0.0|
|F + Pimo + S||0||0||2||10.0|
|F + Pimo + Dig + β-blocker||0||0||1||5.0|
Group-Wise Comparisons in CKCS
In CKCS, 21 of 38 Holter-derived variables were significantly influenced by dog group (all P < .0001). The significant differences in Holter-derived variables between dog groups are shown in Table 3. Episodes of “premature normals,” sinus tachycardia, and bradycardia decreased with increasing age (all P < .0001) in CKCS, otherwise none of the Holter-derived variables in Table 3 was associated with age, sex or interaction between dog group and age. The Holter-derived variables not included in Table 3 were not associated with dog group, age, sex, or interaction between age and dog group in the data set including only CKCS. No Holter-derived variable was significantly different between dog groups 1 and 2 or 2 and 3 (Table 3). Therefore, MRL were only performed to differentiate between dog groups 3 and 4. In the MRL including mean HR (P = .3) as a fixed effect (fixed model), pNN5024 (P = .009), and “premature normals” (P = .002) could distinguish dog group 3 from 4. However, if MRL did not include mean HR (non-fixed model), TI24 (P = .004), and “premature normals” (P = .01) could distinguish dog group 3 from 4. All AUC from ROC curves plotted of pNN5024, TI24, “premature normals,” and mean HR were significantly different from 0.5 (all P < .003) (Fig 1). However, only TI24 and “premature normals” showed diagnostic potential, because the 95% confidence intervals were higher than the line of no effect (AUC = 0.5) in a specific interval (Fig 1B,C).
|Dog group||1 versus 2||1 versus 3||1 versus 4||2 versus 3||2 versus 4||3 versus 4|
|HRV frequency domain variables||TP24b||TP24b|
|VLF24a||VLF24 and 6b|
|HRV time domain variables||MEAN24 and 6b||MEAN24 and 6b|
|pNN5024 and 6b||pNN5024 and 6b||MEAN24 and 6b|
|RMSSD24 and 6a||RMSSD24 and 6b||pNN5024 and 6b||pNN5024a|
|SD24b||SD24 and 6b||SD24b||TI24b|
|TI24 and 6b||TI24 and 6b|
Association between Holter-Derived Variables and Echocardiographic Indices of MMVD Severity in CKCS
In CKCS, 23 of 38 Holter-derived variables were significantly influenced by ≥1 echocardiographic indices of MMVD (Table 4). The remaining Holter-derived variables were not associated with echocardiographic indices of MMVD, age, sex, or interaction between age and echocardiographic indices of MMVD.
|HRV frequency domain variables||TP24||<0.0001||ns||<0.0001||<0.0001||<0.0001||<0.0001|
|HRV time domain variables||MEAN24||<0.0001||nsa||nsa||<0.0001||<0.0001||<0.0001|
|Arrhythmia variables||Sinus tachycardiaa||<0.0001||ns||0.0004||0.0002||<0.0001||<0.0001|
|“premature normals” a||0.0002||ns||<0.0001||<0.0001||<0.0001||<0.0001|
Difference between CKCS and Non-CKCS among Dogs in CHF
Among dogs in CHF, none of the Holter-derived variables were significantly influenced by dog group, age, sex or interaction between dog group and age.
Holter-derived variables changed with MMVD severity in CKCS and changes were most pronounced in advanced stages of the disease. HRV variables, “premature normals,” bradycardia, and sinus tachycardia decreased with increasing MMVD severity, whereas minimum and mean HR measured over 24 hours increased. “Premature normals” and TI24 had diagnostic potential in distinguishing between CKCS with moderate to severe MR and CKCS in CHF. CKCS in CHF showed no significant difference in Holter-derived variables compared to other breeds of dogs in CHF.
This study suggests that HRV decreases with increasing MMVD severity in CKCS with early stages of MMVD because MVP was inversely associated with multiple HRV variables. Previous studies have shown no, a positive, or an inverse association between HRV and MVP in early stages of MMVD.[7, 16, 18] The discrepancies may be explained by differences in breed, age, recording periods, environment, and equipment.
In more advanced stages of MMVD in dogs, this study and others confirm that HRV decreases and HR increases with increasing MMVD severity.[14, 15, 23] These changes in CHF are commonly described as increased sympathetic or decreased parasympathetic modulation of HR, or both, caused by altered activity of the autonomous nervous system.[2, 9-11] However, because HRV indirectly denotes autonomic tone, changes in HRV related to CHF also could be explained by decreased sinoatrial node responsiveness or altered baroreflex function, which is seen in dogs with tachycardia-induced CHF. As expected, this study indicated decreasing parasympathetic activity with increasing MMVD severity, because HF24 and HF6 significantly decreased with echocardiographic indices of MMVD severity. LF24 also decreased with increasing MMVD severity, but LF is more difficult to interpret because LF reflects both parasympathetic and sympathetic activity. Minimum and mean HR increased significantly with increasing MMVD severity, which could indicate increased sympathetic activity. However, decreased parasympathetic activity would also increase HR, which is more likely in this study because LF24 decreased with increasing MMVD severity. LFn24, LFn6, HRn24, and HFn6 were not associated with MMVD severity, but an association might be masked by the normalization equation (LFn = LF / (TP − ULF − VLF) × 100), where both numerator and denominator decreased with MMVD severity. Maximum HR was not related to MMVD severity, but maximum HR was shown previously to be the least reproducible HR variable. In this study, all time-domain variables were significantly influenced by dog group and several echocardiographic indices of MMVD. Thus, both frequency and time domain HRV and HR variables from this study indicate that overall HRV and parasympathetic tone decrease with increasing MMVD severity. Supraventricular arrhythmias previously have been associated with MMVD in dogs.[18, 23] However, supraventricular arrhythmias were not associated with the degree of MVP in clinically healthy CKCS, Dachshunds, and Cairn Terriers, and the present study showed no association between supraventricular arrhythmias and MMVD severity in CKCS. Episodes of “premature normals,” sinus tachycardia, and bradycardia were associated with MMVD severity in this study. However, these variables were also associated with age in nonparametric statistical models and therefore it was not possible to determine whether the 3 arrhythmia variables were influenced by age, MMVD severity or both. APCs, VPCs, and second-degree AV-block were common findings in CKCS in this study, but the frequencies were not associated with MMVD severity.
In agreement with Häggström et al, HRV could not be used to differentiate between early stages of MMVD. The Holter-derived variables pNN5024, TI24, mean HR, and “premature normals” could differentiate between CKCS with moderate to severe MR and CKCS in CHF. However, ROC curves demonstrated that only “premature normals” and TI24 had diagnostic potential in a specific interval. The ability only to show diagnostic potential in specific interval might be caused by a low number of observations (ie, lack of power) or the fact that the majority of CHF dogs were receiving medical treatment. Only 3 CKCS in CHF were not receiving medical treatment during the examinations and excluding these dogs from the statistical analyses (data not shown), resulted in fewer Holter-derived variables being significantly different between CKCS in CHF and CKCS with other stages of MMVD. Thus, it can be suggested that “premature normals” and TI24 have diagnostic value; these 2 Holter-derived variables also could have prognostic value. The definition of “premature normals” corresponds to severe sinus arrhythmia measured on short-term ECG, which previously has been associated with MVP in dogs.[16, 18] TI encompasses all HRV and can predict risk of death or sustained ventricular tachycardia in people after acute myocardial infarction.
Breed differences in Holter-derived variables have been described in clinically healthy dogs.[7, 38] Clinically healthy CKCS have decreased HRV and increased minimum HR compared to other breeds of dogs. According to the present study, CKCS and other breeds of dogs have similar HRV and HR in CHF. Therefore, other breeds of dogs may have a larger change in HRV and minimum HR when progressing from clinically healthy to CHF secondary to MMVD. These differences could be linked to the finding that CKCS in CHF have longer survival compared to other breeds of dogs in CHF. However, further studies are needed to evaluate this hypothesis.
In conclusion, Holter-derived variables reflecting autonomic modulation of heart rhythm changed with MMVD severity in CKCS and changes were most pronounced in advanced stages of MMVD. In CKCS, “premature normals” and TI24 showed diagnostic potential in a specific interval. There was no difference in Holter-derived variables between CKCS in CHF and other breeds of dogs in CHF secondary to MMVD.
Data from dogs in CHF were collected at 3 different locations, and 2 different operators performed echocardiography. However, the same equipment was used and only 1 operator evaluated the echocardiographic data. Furthermore, CHF dogs were treated with different types of medication, including β-blockers and digoxin, which could influence Holter-derived variables. However, it is difficult to recruit a sufficient number of privately owned dogs in CHF secondary to MMVD on a standardized medical treatment protocol.
The age difference among dog groups is a limitation to this study. However, the influence of age on Holter-derived variables was accounted for by including age as an explanatory variable in the statistical models.
The Holter analysis software is designed for people and therefore the HRV analyses exclude sinus pauses >2.5 seconds, which are common in dogs. In addition, APCs and SVT were included in the HRV analysis. It is assumed that software limitations in relation to HRV analysis only had a minor influence on the results, because sinus pauses were not influenced by MMVD severity and inclusion of APCs and SVT in statistical models was not significantly associated with any HRV variables (data not shown). Another software limitation is that the definitions of some arrhythmia variables were based on reference values in people. The authors have tried to accommodate the settings to dogs. However, publications including reference values of Holter monitoring in dogs are sparse.
The study was supported by the Danish Council of Independent Research|Medical Sciences (project no. 271-08-0998 and 271-05-03335) and Synergy in Human and Animal Research (SHARE), Faculty of Life Sciences, University of Copenhagen.
The authors thank Birgitte Holle, Christina Kjempff, Vibeke Christensen, Dennis Jensen, and Hanne Carlsson at the Department of Basic Animal and Veterinary Sciences, University of Copenhagen, Frederiksberg, Denmark for their indispensable technical assistance.
Vivid i echocardiograph, GE-medical, Milwaukee, WI
Lifecard CF Digital Holter recorder, SPACELABS Healthcare Company (previously Delmar Reynolds), Issaquah, WA
3MTM Red DotTM electrodes, 3M, St. Paul, MN
Pathfinder digital Holter analysis system V8.701, SPACELABS Healthcare Company
HRV Tools software package version 1.73, SPACELABS Healthcare Company
HRV Tools. Installation and instruction manual. ©2004 Delmar Reynolds Medical Limited, Hertford, Denland. Drawing No. 038/0369/0 Issue 2 CN 4657. Part No 18-0369
SAS statistical software, version 9.1, SAS Institute, Cary, NC
MedCalc statistical software, version 126.96.36.199, Mariakerke, Belgium
- 2Insights from the study of heart rate variability. Annu Rev Med 1999;50:261., .
- 5Beat-to-beat variations of heart rate reflect modulation of cardiac autonomic outflow. News Physiol Sci 1990;5:32–37..
- 6Physical activity is a major contributor to the ultra low frequency components of heart rate variability. Heart 1999;82:9e., , .
- 16Mitral Valve Prolapse in the Dog. Pathogenesis, Pathophysiology, Diagnosis and Comparative Aspects of Early Myxomatous Mitral Valve Disease. Copenhagen, Denmark: The Royal Veterinary and Agricultural University; 2000. Thesis..
- 22Heart rate variability assessment in patients with mitral valve prolapse syndrome. Am J Noninvas Card 1993;7:210–214., , , et al.
- 33Diagnosis and treatment of arrhythmias (dysrhythmias). In: Kittleson MD, Kienle RD, eds. Small Animal Cardiovascular Medicine, St. Louis, MO: Mosby; 1998:449–494..
- 34Arrhythmia in mitral valve prolapse. In: Podrid PJ, Kowey PR, eds. Cardiac Arrhythmias: Mechanisms, Diagnosis and Management. Philadelphia, PA: Williams and Wilkins; 1995:1253–1266., .
- 35Analysis of common canine cardiac arrhythmias. In: Cann CC, Hunsberger SL, eds. Essentials of Canine and Feline Electrocardiography. Malvern, PA: Lea & Febiger; 1992:127–207..
- 38Standardization of the 24-hour electrocardiographic monitoring in dogs. Arq Bras Med Vet Zootec 2002;54:133–138., , , .
- 39Effect of pimobendan or benazepril hydrochloride on survival times in dogs with congestive heart failure caused by naturally occurring myxomatous mitral valve disease: The QUEST study. J Vet Intern Med 2008;22:1124–1135., , , et al.