Is There Increased Sympathetic Activity inPatients with Mitral Valve Prolapse?

GEORGE E. KOGHIADAKIS, FRAGISKOS I. PARTHENAKIS,EMMANUEL G. ZURIDAKIS, AMALIA T. ROMBOLA,STAVROS I. GHRYSOSTOMAKIS, and PANOS E. VARDAS

From the Cardiology Department, Heraklion University Hospital, Crete, Greece

KOCHIADAKIS, G.E., ET AL.: IS There Increased Sympathetic Activity in Patients with Mitral Valve Pro-lapse? The aim of this study was to investigate autonomic nervous system tone in patients with mitralvalve prolapse (MVP). Heart rate variability (HRV) was assessed from 24-hour ambulatory Holter record-ings in 28 patients with primary MVP and in 28 age and sex matched normal control subjects in a drug-free state. Sixteen of the MVP patients were symptomatic and 12 asymptomatic. Spectral HRV was calcu-lated in terms of low (LF: 0.06-0.15 Hz) and high (HF: 0.15-0.40 Hz) frequency components using fastFourier transform analysis, and the ratio LF/HF was calculated. Spectral analysis of HRV showed that theMVP patients, taken as a single group, had lower HF and LF and a higher LF/HF ratio than the controls.No significant difference in HRV was found between the 16 symptomatic and the 12 asymptomatic pa-tients, but the symptomatic patients had a significantly higher LF/HF ratio than the controls. Our obser-vations suggest that, during normal daily activities, patients with MVP experience a significant deviationin autonomic nervous system tone with predominance of the sympathetic branch. This predominance ismore marked in symptomatic patients. (PACE 1996; 19[Pt. II]:1872-1876)

mitral valve prolapse, autonomic nervous system, heart rate variability

Introduction

Previous studies suggest that mitral valve pro-lapse (MVP) is due to a neuroendocrine distur-bance of which the cardiovascular system is a ma-jor target, and that anatomical MVP is a marker ofa heritable disorder that includes an abnormal,centrally mediated autonomic adrenergic activ-ity.^ On the other hand, other studies have re-ported no difference in autonomic testing of pa-tients with MVP compared with control subjectsand have suggested that autonomic abnormalitiesmay be caused by symptoms rather than the MVPitself.̂ '̂ Furthermore, the type of autonomic ab-normality involved has not been precisely charac-terized. Some investigators have suggested thatthe abnormality is predominantly mediated by al-pha-adrenergic receptors,"*'̂ whereas others founda significant contribution from abnormal beta-adrenergic receptor mediated function.^ Simi-

Address for reprints: Prof. Panos E. Vardas, M.D., Ph.D. (Lon-don), FESC, FACC, Cardiology Department, Heraklion Univer-sity Hospital, P.O. Box 1352, Stavrakia, Heraklion, Crete,Greece. Fax: 30-81-542-055.

larly, both normal^ and diminished^ parasympa-thetic tones have been reported in MVP patients.

In view of these contradictory reports, weused the analysis of heart rate variability (HRV) ina group of patients with MVP and a group of ageand sex matched normal subjects to determinewhether there is altered autonomic activity duringdaily life in these patients and if such autonomicabnormalities are related to the presence of symp-toms.

Patients and Methods

Study Population

The study population consisted of 28 selectedpatients (17 men and 11 women: mean age 43 ±5.7 years) with primary MVP, 16 of whom weresymptomatic (chest pain, palpitations, dyspnea,fatigue, dizziness, syncope, anxiety). These pa-tients had no cardiovascular disease other thanMVP. Patients with diabetes mellitus, neuropathy,obstructive pulmonary disease, and atrial fibrilla-tion or conduction disturbances on resting ECGwere excluded. The diagnosis of MVP was based

1872 November 1996, Part II PACE, Vol. 19

AUTONOMIC NERVOUS SYSTEM TONE IN MVP

on the presence of a nonejection click and late sys-tolic murmur [20 cases) or holosystolic murmur (8cases) and was confirmed by M-mode and two di-mensional echocardiography.

These patients were compared with an ageand sex matched control group of 28 subjects,none of whom had clinical or echocardiographicevidence of MVP. Both patients and controls wereeither taking no medication or had agreed to ab-stain from medication for a minimum of 3 daysprior to the study and throughout the study pe-riod. Informed consent was obtained from bothpatients and control subjects.

Analysis of HRV

All patients and control subjects underwent3-channel, 24-hour Holter ambulatory ECG moni-toring.

The spectral indexes of HRV were calculatedfrom the Holter recordings using Fourier analysisfor each 2-minute interval, using a Hanning win-dow to minimize spectral leakage. HRV over the24 hours was taken as the average of all 2-minuteintervals. The spectral power was evaluated quan-titatively and expressed as In (ms^/Hz), where Inis the natural logarithm of tbe quotient, after mea-surement of the areas of two frequency ranges: lowfrequency (LF: 0.06-0.15 Hz), which gives mainlya measure of sympathetic activity with some in-fluence from the parasympathetic nervous system,and high frequency (HF: 0.15-0.40 Hz), which re-flects solely parasympatbetic activity. The ratioLF/HF was calculated as a measure of the sympa-thovagal balance.^

The mean heart rate was derived from themean of all normal RR intervals.

Statistical Analysis

Data are expressed as mean ± SD. Separatesubanalyses were carried out to assess differencesbetween the controls and the MVP patients withand without symptoms. Spectral indexes of HRVwere compared for controls and MVP patients us-ing a two-sided (-test. One-way ANOVA was usedto compare mean values in the three groups.

To compare the time course of LF, HF, andLF/HF between the groups, repeated measuresmultivariate ANOVA with one grouping factorwas used. This comparison was performed inthree stages: after initially assessing differencesover the entire 24-hour period, differences duringdaytime (07:00-22:00) and nighttime (22:00-07:00) were examined. Normality and homogene-ity assumptions were tested by the Kolmogorov-Smirnov and Cochran's C-test, respectively. Pil-lais' multivariate test was used when assumptionsheld; Greenhouse-Geisser e-adjusted P values-were used if not. P < 0.05 was the criterion of sig-nificance in all tests.

Results

Twenty-Four-Hour Average HRV Indexes

As shown in Table I, the mean values of LFand HF were significantly less in MVP patientsthan in controls (P < 0.001 for both). LF/HF aver-aged 1.25 in MVP patients as opposed to 1.20 incontrol subjects (P = 0.03). one-way ANOVA

Table I.

Twenty-Four-Hour HRV Indexes and Mean NN in Patients with Mitral Valve Prolapse and NormalControls

Normal(mean ± SD)

MVP(mean ± SD)

Symptomatic(mean ± SD)

Asymptomatic(mean ± SD)

LFHFLF/HFMean NN

7.096.011.20808

± 0.66± 0.82± 0.11± 97

6.31 ± 0.81'5.06 ± 0.95*1.25 ± 0.13'794 ± 98

6.154.891.27768

± 0.73*± 0.88'± 0.10'± 103*

6.43 ± 0.86'5.19 ± i.or1.23 ± 0.15813 ± 94

'P significantly different from normals.HF = high frequency; LF = low frequency; MVP = mitral valve prolapse.

PACE. VoL 19 November 1996, Part II 1873

KOCHIADAKIS, ETAL.

showed that LF and HF values in both asymp-tomatic and symptomatic patients were signifi-cantly less than in control subjects, whereasLF/HF in symptomatic MVP patients was greaterthan in controls (P = 0.01). Mean NN in MVP pa-tients tended to be lower than in normals, while insymptomatic patients this difference reached sta-tistical significance (P = 0.008).

Time Course of Spectral Indexes of HRV

Significant changes occurred in LF and HFvalues over the 24-hour period of observation (P <0.001 for both) in normal subjects and patientswith MVP (Fig. 1). Although group effects failed toattain statistical significance for the entire 24-bourperiod [P = 0.1], during daytime (07.00-22.00) the

vno

850

S TO Ifl

As/

V

kair

; i ! 1 1

80O

1

400

3 S S 2 5 ?how

5 S g 1

no

800

560

BOO'

^ ^ \ V A / S

2 S 5 2 5 ^ ^ ^ ? 1 Shwir

—•—AS*

-•—SK

15

1 t

1 1

^ .- — . - . - w Hktar

—•—NL

»-MVP

Figure 1. Patterns of changes in heart rate variability parameters (low [LF] and high [HFj fre-quency spectral power and the ratio LF/HF) over 24 hours in patients with mitral valve prolapse(MVP) and normal controls (NL). The three left-hand panels (a. c. and e) show comparisons be-tween all MVP patients and controls. In the three right-band panels (b, d, andf), the MVP patientsare divided into symptomatic (Sx) and asymptomatic (ASx).

1874 November 1996, Part II PACE, Vol. 19

AUTONOMIC NERVOUS SYSTEM TONE IN MVP

HF and LF values were significantly lower in MVPpatients than in normals (P < 0.001 and P < 0.008,respectively).

However, when the spectral indexes were ex-amined separately for the three groups, a distinctdifference between normals and symptomatic pa-tients in HF and LF values was found (P = 0.021and P = 0.012, respectively). These differencespersisted throughout the 24-hour period.

No significant group effects could he detectedbetween normals and asymptomatic patients, al-though during daytime the values of LF and HF inasymptomatic patients were significantly lowerthan in normal suhjects (P = 0.03 for hoth).

Time Course of LF/HF

There was also a significant time dependenteffect on this ratio in hoth groups (P < 0.001). Nor-mals and MVP patients did not differ significantlywith respect to mean values of LF/HF, althoughthere was a tendency toward higher values duringdaytime (07.00-22.00) in MVP patients. Analyz-ing the three groups separately, the values ofLF/HF in symptomatic MVP patients were foundto be significantly higher than in the other groupsduring daytime (P < 0.05).

Discussion

Our results suggest that there is reduced vagaltone during daily activities in patients with MVP.This is indicated by the low values of HF and LFthat, over a 24-hour period, are hoth affected, ei-ther wholly or in part, by parasympathetic activ-ity. This finding applies equally to symptomaticand asymptomatic MVP patients, an observationconsistent with the findings of Gaffney et al.,^who, in a study using the diving reflex andphenylephrine infusion, reported reducedparasympathetic tone among their patients withMVP.

In contrast, the sympathetic predominanceshown by MVP patients compared to normals, es-pecially during daytime, appeared to be limited tothe symptomatic patients, since the asymptomaticMVP patients had LF/HF values that did not differsignificantly from those of the control group.

The fact that the asymptomatic patients hadno sympathetic predominance in spite of theparasympathetic withdrawal suggests that thesepatients have an equally reduced sympathetictone. The study of Weissman et al.," who reporteda reduced sympathetic tone in patients with MVP,most of whom were asymptomatic, supports thishypothesis. Given the reduced parasympathetictone in the symptomatic patients, the sympatheticpredominance in this group (increased LF/HF ra-tio) could be explained in terms of either in-creased or normal sympathetic tone.

This observation could explain the discrepan-cies between the findings of previous studies.While Boudoulas et al.^ and Pasternac et al.^"found an increase in urinary norepinephrine andepinephrine levels in symptomatic MVP patients,studies by Gbesler et al.^ and Lenders et al.^showed no difference in serum catecholamine lev-els between symptomatic patients with MVP andnormals. Our hypothesis is that the autonomic ab-normalities are not uniform in symptomatic MVPpatients. All these patients show a reduction inparasympathetic tone, while the sympathetic pre-dominance may also reflect increased sympa-thetic activity in some of these patients and nor-mal sympathetic tone in others.

The observation that the symptomatic pa-tients had reduced parasympathetic tone through-out the 24-hour period, while the asymptomaticpatients showed this reduction mainly duringdaytime, is interesting but remains unexplainedand requires further study.

Study Implications

The findings of this study support the viewthat the autonomic abnormalities in patients withMVP are not uniform. The reduction in parasym-pathetic tone, which is common to all MVP pa-tients, may be linked to the pathogenesis of MVP,while the symptom complex in patients thought tohave MVP syndrome could he attributable to asympathetic predominance in some suhjects.

It remains unknown whether the symptomcomplex is related to the valvular prolapse or rep-resents an independent manifestation in certainpatients with coincidental anatomical MVP.

PAGE, Vol. 19 November 1996, Part II 1875

KOCHIADAKIS, ET AL.

References

Wooley CF. The mitral valve prolapse syndrome.Hosp Pract 1983; 18:163-174.Chesler E, Weir EK, Broatz GA, et al. Normal cate-cholamine and hemodynamic responses to ortho-static tilt in subjects with valve prolapse. Correla-tion with psychological testing. Am J Med 1985:78:754-760.Lenders JWM. Fast JH, Blankers J. et al. Normalsympathetic neural activity in patients v̂ rith mitralvalve prolapse. Clin Cardiol 1986; 9:177-182.Caffney FA, Karlsson ES, Campbell W, et al. Auto-nomic dysfunction in women with mitral valveprolapse syndrome. Circulation 1979; 59:894-901.Caffney FA, Bastian BC, Lane LB. Abnormal car-diovascular regulation in the mitral valve prolapsesyndrome. Am J Cardiol 1983; 52:316-320.Davies AO, Mares A, Pool JL, et al. Mitral valveprolapse with symptoms of p-adrenergic hyper-sensitivity: [3-adrenergic receptor supercoupling

10,

with desensitization on isoproterenol exposure.Am J Med 1987; 82:193-201.Pagani M, Lombardi F, Guzzeti S, et al. Powerspectral analysis of heart rate and arterial pressurevariabilities as a marker of sympathovagal interac-tion in man and conscious dog. Circ Res 1986; 59:178-193.Weissman NJ, Shear MK, Kramer-Fox R, et al. Con-trasting patterns of autonomic dysfunction in pa-tients with mitral valve prolapse and panic at-tacks. Am J Med 1987; 82:880-887.Boudoulas H, Wooley CF. Mitral Valve Prolapseand the Mitral Valve Prolapse Syndrome. MountKisco, NY, Futura Publishing Company, Inc.,1988, pp. 479-510.Pastornac A, Tubau JF, Puddu PE, et al. Increasedplasma catecholamine levels in patients withsymptomatic mitral valve prolapse. Am J Med1982; 73:783-790.

1876 November 1996, Part II PACE, Vol. 19