|
|
|
|
|
 |
|
From its beginnings as an invasive intra-arterial research tool in the 1960s, ambulatory BP monitoring (ABPM) has slowly become established as a useful adjunct to the management of patients suspected to have, and with, raised BP. This was facilitated by the development of small, cheaper, portable, more accurate and reliable equipment, provided with more powerful software analysis packages, in the 1980s. In this presentation I will try to prove why ABPM may be particularly important in nephrology. I will focus especially on the consequences of abnormal diurnal BP variability.
Hypertension is very common at all stages of renal disease, especially so in patients on renal replacement therapy and after renal transplantation. However, measuring BP accurately can be problematic; cuff-size, digit-preference, observer bias, “office hypertension” and lability of BP being a few of the more obvious difficulties (Pickering TG. Blood pressure measurement and detection of hypertension. Lancet, 1994, 344, 31-35). Even if BP is taken using the best equipment and with appropriate time, the use of a single BP reading taken over a few seconds in a hospital setting to characterise the long-term behaviour of a continuously variable phenomenon is unrealistic – cf a single blood sugar reading compared to glycated haemoglobin to characterise diabetic control. Moreover, there are several clear clinical indications and advantages that this rather simple method offers.
Compared to the normal circadian pattern of variation, three situations are easy to recognize: 1) a transient increase in BP levels during the office visit – i.e. “white coat” hypertension, 2) a sustained increase in BP levels, but with a normal “physiological” decline during the night-time sleep hours – i.e. ABPM hypertensive and dipper, and 3) a sustained increase in BP levels with loss of the nocturnal decline in BP – i.e. non-dipper profile. These theoretical profiles are exemplified by an intra-arterial BP measurement (hypertensive, dipper) and a classical ABPM recording (hypertensive, non-dipper; of note the dichotomy between BP and HR circadian profiles).
For a correct use of ABPM in the daily nephrological practice we need to report to normal standards, usually derived from large populational studies. There have now been several large analyses of normotensive and hypertensive populations, which, when meta-analysed, provide a useful normal range for ABPM (Staessen JA, et al. What is normal blood pressure on ambulatory monitoring? NDT, 1996, 11, 241-245).
Predictably, there is a often good correlation between office-derived and ABPM-derived BP levels. For the same patient (Lebel M, et al. A.J.Hypertension 1995; 8:545-551) it is difficult to substitute one measure with the other. Many studies have tried to find and to use a single office BP reading that most closely approximates to ABPM data. This has been done mainly in the context of haemodialysis (HD), and the practice of using pre-HD, post-HD, or averaged pre- and post-HD BP values. Moreover, one of the main problems of assessing dialysis patients BP’s is the fact that if one relies on single BP values taken around the time of an outpatient dialysis session, there is a very poor correlation between these peri-dialysis values, and others taken away from hospital during the inter-dialytic interval. Zoccali et al showed that it took the average of 12 pre-dialysis BP values to have the same predictive power for LV mass as one single ABPM session in 64 non-diabetic haemodialysis patients without heart failure.
A complete description of the inter-relations between casual, office BP derived and true, ABPM-derived levels was offered by the group of Agarwal and colleagues who compared pre- and post- HD oscillometric BP readings with 44-hour interdialytic ambulatory BP and "home BP" in chronic US haemodialysis patients.
There are many studies that support the notion
of blunted diurnal rhythm with renal disease; those that oppose this consensus
are small and unconvincing.
When reviewing the existing literature it is clear that a reduction in diurnal BP fall with sleep is observed consistently and frequently in chronic renal diseases. In dialysis cohorts, the prevalence of abnormal diurnal BP rhythm reported in the literature varies from 22.2% (Korzets et al. NDT, 1994, 9, 274-276) to 80-100% (Ertuk S, et al. NDT, 1996, 11(6), 2050-4). One of the largest studies reported, from Spain, reported a 50% incidence of non-dipping in 414 CAPD patients. We calculated in a meta-analysis, the “average” (bearing in mind the different definitions in current use) of >80 studies and 2000 patients, which is about 59%. A rise in BP at night over the average daytime value (complete reversal of diurnal rhythm) is seen in less than 10 to 30% of dialysis patients.
The abnormal BP circadian pattern (i.e. “non-dipping”)
is apparent even in patients with different renal conditions and normal renal
function. Patients with IgA nephropathy have increased BP levels during the
night compared to age, gender and BMI-matched controls (Stefanski et al. KI
1996; 50: 1321-1326). This was seen in other single-etiology study populations
(e.g. – ADPKD, Valero et al. JASN, 1999, 10, 1020-1026), and confirmed
by the largest cohort to date (Farmer et al. NDT, 1997:12, 2301-7): 55% of
patients with diverse renal conditions and a plasma creatinine of < 110
umol/l had reduced BP fall with sleep, compared to 33% of a BP-, age-, plasma
creatinine and gender-matched group of essential hypertensives.
In general there is agreement that the diurnal
BP rhythm is already abnormal at normal or near-normal renal function and
progressively is further blunted as renal function declines. Farmer and Goldsmith
retrospectively studied 480 ambulatory blood pressure recordings in 380 patients
with essential hypertension, secondary (renal) hypertension, and on renal
replacement therapy. Abnormal blood pressure diurnal rhythm (non-dipping)
was significantly more prevalent in patients with underlying renal disease
and normal excretory renal function than in matched controls with essential
hypertension. In patients with renal disease the prevalence of non-dipping
rose with worsening renal function, reaching statistical significance once
plasma creatinine was greater than 400 umol/l. There was a direct correlation
between plasma creatinine and the percent decline in blood pressure at night
for both systolic and diastolic blood pressure in patients with underlying
renal disease and impaired excretory renal function.
There is a clear impact of the renal replacement method used: in a cross-sectional study we found the highest prevalence of non-dipping in CAPD patients compared to standard HD, long-hours home- HD or transplantation. This descriptive data did not correlate differences in “non-dipping” prevalence with differences in BP control, hydration status or dialysis efficiency, between RRT methods. Once transplanted some of the abnormalities are reversed, irrespective of the length of the dialysis period pre-transplantation.
There is more to the diurnal variation in BP
than the expected nocturnal dip. A rapid increase in BP have been recorded
just before the HD session; post-HD the lowest BP value is recorded 3-4 hors
after the HD session. Controversies still exist related to changes in BP levels
during the two interdialytic days: an interesting possibility was suggested
by two independent groups (Chazot et al. Interdialysis blood pressure control
by long intradiaysis sessions. NDT, 1995, 10, 831-837 and Luik et al. Diurnal
blood pressure variations in hemodialysis and CAPD patients. NDT, 1994, 9,
1616-1621) that a progressive increase is seen only in hypertensive / less-well
controlled BP patients.
Caution is warranted when interpreting and comparing ABPM studies, since several methodological limitations / errors may be encountered. First, it is evident (from the graph above) that different observation periods are used to define day and night and therefore to calculate BP levels during true activity vs levels during recumbency / sleep. This methodological discrepancy makes comparisons between (often small) studies very difficult. Best ABPM practice should use activity diaries to derive awake and sleep periods.
Second, there is a plethora of definitions for
the “non-dipping”status, in the renal ABPM literature, very few
of them in accordance with modern best ABPM practice. Finally, there are highly
pertinent concerns about the dichotomous reporting and analysis (i.e. “dipper”
/ “non-dipper”) of a continuous variable – the diurnal BP
rhythm.
Extra problems for renal patients having ABPM
include the discomfort of repetitive readings of (typically) high systolic
BP; upper arm bruising (pressure, prednisolone, purpura); the presence of
a fistula (usually in the non-dominant arm, therefore mandating ABPM in the
dominant arm); increased likelihood of brachial artery calcification (especially
in diabetics) which may increase the possibility of “pseudo-hypertension”,
though in practice this is very rarely reported; reduced /disturbed night-sleep
from “uraemic-dialysis” problems such as joint or muscle pain,
myoclonus, akathisia, breathlessness, nocturia, pruritus and increased daytime
somnolence. Despite this, in our experience we achieve technically-successful
ABPM traces in over 90% of renal patients on the first attempt.
I will now turn to the general consequences of
an abnormal BP circadian variability pattern in renal patients. As a general
observation it is very hard, and possibly specious, to attempt to dissect
out the effects of diurnal BP rhythmicity from elevation of BP levels per
se, when so many renal patients have both elevated BP level and reduced BP
rhythmicity, and so many aetiological factors seem to be common to both. To
be truly relevant as a BP parameter, the (abnormal) circadian rhythm should
be associated with more important end-organ (kidney, heart, brain) damage,
and increased mortality.
First to consider is the effect of diurnal BP
rhythm on renal function and proteinuria in patients with renal disease and
declining renal function. The best study, by Timio et al. prospectively matched
for daytime BP two groups with hypertensive nephrosclerosis – one group
had reduced sleep-related BP fall, and hence a greater 24-hours BP load, compared
to the other group. Urinary protein excretion increase was higher in the non-dipper
group than in the dipper group. Also, more importantly, renal function declined
significantly faster in the non-dipping group.
Farmer and Goldsmith assessed the contribution
of abnormal blood pressure diurnal rhythm to the progression of diabetic nephropathy.
They studied 26 diabetic patients with hypertension proteinuria and relentless
progressive impairment of renal function due to diabetic nephropathy between
1990 and 1996. Patients underwent ABPM and were classified as either 'dippers'
or 'non-dippers' according to their blood pressure diurnal rhythm. Weight,
glycated haemoglobin, serum creatinine and blood pressure were recorded on
a 3-monthly basis. In the 'dipper' group, the rate of decline of creatinine
clearance was -2.9 ml/min/year vs -7.9 ml/min/year in those with abnormal
blood pressure diurnal rhythm (P<0.05). There was no significant difference
in day-time mean blood pressures, glycated haemoglobin, age and numbers with
insulin-dependent diabetes mellitus.
There are several studies linking the “non-dipping”
status with a higher albumin excretion rate in diabetics and also with the
development of microvascular disease. Moreover, in the study of Farmer et
al (NDT 1998, 13: 635-39), clinical evidence of autonomic neuropathy was much
more prevalent in the non-dipping group. Thus a blunted BP fall during the
night-time may represent a surrogate marker for a worse renal outcome in diabetics.
Csiky selected 126 consecutive IgA nephropathy
patients for ABPM. 55 patients were normotensive and 71 were treated hypertensives
(ACE-I alone or in combination with a CCB). The mean night-time BP of normotensives
(108±9/67±6 mmHg) was significantly lower than their day-time
BP (125±8/82±7 mmHg, P<0.05). There was no significant difference
between the mean day-time and night-time BP in hypertensive patients (125±9/82±7
mmHg vs 128±10/85±9 mmHg). The circadian variation of BP was
preserved ('dippers') in 82% of the normotensive but only in 7% of the hypertensive
patients (P<0.001). There was no difference in mean day-time BP among normotensive
and treated hypertensive patients and no difference in serum creatinine levels
among the different groups at the time of ABPM. However, thirty-six±4.1
months after the ABPM, hypertensive patients (n=52) had higher serum creatinine
levels (124±32 umol/l) than at the time of the ABPM (101±28
umol/l). The serum creatinine of dipper normotensive patients did not change
during the following period. In contrast, “non-dipper” normotensives
had significantly higher serum creatinine levels at the end of the follow-up
period than at its beginning (106±17 umol/l vs 89±18 umol/l,
P<0.05).
Therefore, from these studies we can conclude that there seems to be a profound
effect of non-dipping upon the rate of decline of renal function in patients
with non-diabetic and diabetic nephropaties.
Second to consider, is the effect of ABPM-derived
BP parameters on heart structure and function. Many studies confirm the closer
relation of BP assessed by ABPM to LV mass than by isolated office BP readings.
In all categories of renal patients an increased
LV mass is associated with nocturnal hypertension (i.e. blunted BP fall during
night), but all of these and other studies are cross-sectional associations,
and suffer from the methodological and interpretative deficiencies expected.
Moreover they do not discriminate the type of LV hypertrophy: concentric vs
eccentric.
In a study including 35 long-hours (8 hrs.) home-haemodialysis
patients, with excellent BP control and survival on dialysis for at least
10 years, we demonstrated a relationship between more abnormal circadian variability
and larger internal LV diameters (normalized to body surface area); in the
same group no relationship was found with LV parietal thickness. Therefore,
a blunted fall in BP during sleep is associated with eccentric LV hypertrophy
and LV dilatation – conditions carrying a worse cardiac outcome in general
and renal subjects.
Although there is general agreement that non-dippers
have a higher LVM and larger hearts, it is not clear if this increased target-organ
damage is solely a reflection of a greater 24-hours BP load, or if the lack
of normal diurnal rhythm acts as a surrogate for other injurious cardiovascular
mechanisms, eg dysautonomia. The only truly prospective study to examine ABPM-derived
BP values and dipping status in a normotensive (treated) HD cohort was undertaken
by Covic and co-workers (Am J Kidney Dis 2000 35(4): 617-23).
The study was designed to examine the effect
on echocardiographically derived measurements of the left ventricle in 60
stable chronic hemodialysis patients of abnormal (reduced) diurnal BP variability,
measuring ambulatory BP on three occasions and performing echocardiography
twice over a 12-month period. Only 46 / 60 patients maintained the same dipping
profile over 12 months and three ABPM recordings: 36 “non-dippers”
and 10 “dippers” from the start of the study.
Those patients with persistently reduced diurnal
BP rhythm had a larger internal LV diameter and thicker LV walls, compared
to “dippers”. Most importantly, all echocardiographic parameters
remained stable / unchanged in the dippers group, while a significant tendency
towards a more dilated heart was recorded in the “non-dipper”
subgroup (EDD-N from 35.9 mm/m2 at baseline to 38.2 mm/m2 after 12 months,
p<0.05), although there was no difference in daytime BP, Hb levels, type
of antihypertensive therapy or dialysis quality between subgroups. These results
suggested that persistent abnormal BP circadian variability is an independent
risk factor for a (progressively) dilated heart in hemodialysis subjects,
independent of the BP level or of other known and/or relevant risk factors.
Last but not least, data on mortality / prognosis. There is little of note to report here, but one important study has suggested that ABPM derived diurnal BP rhythm carries some prognostic importance for haemodialysis patients. Fifty-seven treated hypertensive hemodialysis patients (56.9 years, 30 men) were prospectively studied. All patients initially underwent ABPM. The duration of follow-up was 34.4 +/- 20.4 months, during which 10 CV and 8 non-CV fatal events occurred. In the 10 patients who died from CV complications, age, previous CV events, ambulatory systolic BP, ambulatory pulse pressure (PP), and life-long smoking level were significantly higher, and the office diastolic BP was lower at the time of inclusion than in those who did not die from CV complications (N = 47). Based on Cox analysis and after adjustment for age, sex, and previous CV events, a low office diastolic BP [relative risk (RR) 0.49, 95% CI, 0.25 to 0.93, P = 0.03], an elevated 24-hour PP (RR 1.85, 95% CI, 1.28 to 2.65, P = 0.009), and an elevated nocturnal systolic BP (RR 1.41, 95% CI, 1.08 to 1.84, P = 0.01) were predictors of CV mortality (RR associated with a 10 mm Hg increase in BP).
The incidences of CV outcomes (events/deaths)
per 100 patient-years were compared between dippers and non-dippers. Non-dippers
showed an approximately 3.5 and 9 times higher rate of CV events and CV deaths
than dippers did, (30.2 vs 8.6% and 10.8 vs 1.2%, respectively) respectively.
During the follow-up period, 36 CV events (seven
dippers and 29 non-dippers: AMI, 1/0; AP, 4/13; CHF, 0/2; percutaneous transluminal
coronary angioplasty (PTCA), 1/7; coronary artery bypass graft (CABG), 0/2;
stroke, 1/5 for dipper/non-dipper), 16 CV-related deaths (one dipper and 15
non-dippers) and nine non-CV-related deaths (three dippers and six non-dippers)
occurred. The cumulative CV event-free survival and CV survival rates were
significantly worse in non-dippers than in dippers (P=0.019 and P=0.0054 by
the log-rank test, respectively)
Important though these observations are, are
these enough to justify the use of ABPM in a more consistent fashion in renal
patients? Even with the emergence of prospective ABPM studies with hard-end
cardiovascular and renal end-points involving renal subjects (vide infra)
there will still be many who will stare and see only the half-empty glass
(e.g. the difficulties time and money expended to achieve an ABPM service
) rather than the wine therein. In the absence of adequate information from
renal patients, as in so many areas, we nephrologists have to look across
to the world of hypertension research to see what might hold for renal patients.
I will critically review only the solid (i.e. prospective RCT trials) evidence
supporting the use of ABPM to derive BP levels and variability as a better
prospective assessment strategy.
Zweiker et al performed a study involving 116 treated hypertensive patients followed for an average of 31 months. The investigators found a significantly higher rate of CV complications in “non-dippers” (4 events, 29 subjects) compared to “dippers” (1 event in 87 subjects).
At the same time, the preliminary data from the
Progetto Ipertensione Umbria Moniroraggio Ambulatoriale (PIUMA) were reported.
All 1187 patients in the initial study underwent 24-hours ABPM, echocardiography
and metabolic and clinical assessments. Using BP data from similar large Italian
normotensive cohorts the investigators were able to classify the subjects
into normotensive, office-hypertensive, and ABPM-hypertensive. In a follow-up
period of 3.5 years the office-hypertension group had a cardiovascular event
rate similar to the normotensive group, while the event rate in the “non-dipping”
sub-group of the sustained hypertensive subjects was 3-times that of the “dipping”
sub-group.
Staessen undertook a substudy nested in the Syst-Eur
trial: 808 older (>60 years) patients whose untreated SBP (conventionally
measured) was 160 – 219 mm Hg and whose DBP was < 95 mm Hg were randomized
to nitrendipine and / or hydrochlorthiazide, or placebos. The outcome measures
were total and cardiovascular mortality, and all cardiovascular end-points
(fatal and non-fatal stroke and cardiac). Patients who lacked a nocturnal
decline in systolic BP had a greater incidence of stroke and myocardial infarction
than patients with normal diurnal BP variation.
Further evidence supporting a strong relationship
between stroke and abnormal circadian variability was provided by Yamamoto
et al who followed 105 patients with symptomatic lacunar infarcts with 24
hrs ABPM. Follow-up over 3 years showed that in the group with subsequent
further neurological events and silent lacunae, the day-to-night ABP reduction
was much less (1.3% SBP; 3.3% DBP) than in the group with no sequelae (7.2
% SBP; 10.4% DBP).
Very recently, Kario et al reported on the relationship between the extremes of dipping and non-dipping and stroke – studying prospectively stroke events in 575 older Japanese patients with sustained hypertension determined by ABPM, classified by their nocturnal systolic blood pressure fall: 97 extreme-dippers, with > 20% nocturnal systolic blood pressure fall; 230 dippers, with >/=10% but <20% fall; 185 "non-dippers", with >0% but <10% fall; and 63 reverse-dippers, with >0% fall). Baseline brain magnetic resonance imaging disclosed a prevalence of cerebral infarcts of 53% in extreme-dippers, 29% in dippers, 41% in "non-dippers", and 49% in reverse-dippers. There was a J-shaped relationship between dipping status and stroke incidence (extreme-dippers, 12%; dippers, 6.1%; "non-dipper", 7.6%; and reverse-dippers, 22%), and this remained significant in a Cox regression analysis after controlling for age, gender, body mass index, 24-hour systolic blood pressure, and antihypertensive medication. Intracranial haemorrhage was more common in reverse-dippers (29% of strokes) than in other subgroups (7.7% of strokes, P=0.04).
A study which may have special relevance to dialysis
patients was carried out by Nakano et al. who assessed the significance of
reversed (i.e. nocturnal BP higher than diurnal BP) circadian blood pressure
rhythms as a predictive factor of vascular events in NIDDM. 201 subjects had
a normal circadian BP rhythm (group N) while the remaining 87 had a reversed
one (group R). There was no difference in sex, HbA1c, prevalence of smokers,
serum lipids, or serum electrolytes between groups N and R at baseline, whereas
age, the prevalence of hypertension, serum creatinine, and diabetic complications
were more pronounced in group R than in group N. During the follow-up period
(52 months in group N and 36 months in group R), fatal and nonfatal vascular
events occurred in 20 subjects in group N and 56 in group R. The Cox proportional-hazards
model demonstrated that only circadian BP pattern and age exhibited significant
relative risks for fatal events, while diabetic nephropathy, postural hypotension,
and hypertension as well as circadian BP pattern exhibited significant relative
risks for various nonfatal vascular events.
However, any pathologically significant, cause-effect
type of relationship between abnormal circadian variability and definitive
end-points, is based on the presumption that BP diurnal rhythm behaviour is
consistent over time. Unfortunately, this presumption has rarely been investigated
(as most studies have been cross-sectional in nature), and what data exist
in the general population and in essential hypertensive subjects (Palatini
Hypertension 1994; 23: 211-6), tend to cast doubt on the assertion that either
the amplitude in nocturnal BP fall or the "dipper" vs "non-dipper"
status, is consistent over time.
We undertook a prospective study to describe,
for the first time, in a single renal condition 30 subjects with ADPKD,
mild chronic renal failure and normal office BP levels on standardised anti-hypertensive
treatment the reproducibility of the circadian profile, across more
than two consecutive ABPM determinations and over different time intervals
between consecutive measurements. Furthermore, to view circadian variability
in the context of having a potentially independent (from BP levels) impact
on target organ damage, we specifically chose to analyse the consistency of
a particular circadian BP profile over a nine month period, a time-frame entirely
consistent with potentially significant cardiac and vascular remodelling secondary
to BP, rather than to compare only two sets of measurements separated by short(er)
period of times.
A
positive linear correlation was seen between the nocturnal dip (expressed
as % of the awake levels) in BP from the first and second ABPM determination:
y = 0.42x + 6.14, P < 0.001, see previous slide - Figure 3. However a Bland
and Altman analysis for two consecutive ABPM demonstrates that the numerical
differences between mean BP levels for these consecutive ABPM readings can
be high and therefore have high clinical significance. This is explained by
the fact that across the 9 months period and 3 ABPM recordings only 37% of
the patients maintained the initial dipping category (defined by quartiles
of the diurnal BP distribution on the first ABPM).
And
another word of caution regarding the claimed ABPM superiority over office
BP measurements.
Figure shows the variance of echocardiographic parameters explained by the
multivariate models, as well as the additional variance explained by the corresponding
24-h ABPM measurements. Twenty-four-hour diastolic ABPM did add significant
information to the prediction of LVEDD, i.e. it increased by 9% (P < 0.01)
the variance already explained by pre-dialysis diastolic BP and other significant
covariates. Forcing 24-h ABPM into the models of PWT, IVST and LVMI did not
add any significant and independent prediction power. The additional power
of 24-h ABPM over pre-dialysis BP and signifcant covariates for these echocardiographic
parameters was indeed very slight, ranging from ±0.6 to 3.9 (average
1.1%).
I have presented solid evidence to support the importance of abnormal circadian patterns in renal disease. I will next review some of the data investigating mechanisms of this abnormality. Raised BP in end-stage renal failure has many aetiologies. It would be naïve to expect there to be but one explanation of abnormal diurnal BP rhythm. It would be as naïve to expect that the same reason or reasons was as important in the pre-dialysis clinical setting as the dialysis or post-transplantation scenario.
A circadian pattern is present in the normal subject for numerous other factors that may influence BP during the day and night periods: heart rate, electrolyte excretion, cortisol, plasma adrenalin and renin, GFR. These are profoundly altered in the CRF patient.
Diurnal BP changes are probably not an artefact of poor sleep in uraemia, and are not related to the underlying CRF aethiology (with the notable exception of DM).
Furthermore, diurnal BP changes are probably
not related to BP levels since the same abnormal circadian rhythm is recorded
in normotensive and hypertensive renal subjects and at least 50% of the Tassin
(Chazot NDT 1995) and Manchester (Covic et al, Contrib. Neprol 1996) - dialytic
populations (recognized for their exceptional BP control) are “non-dippers”.
The diurnal BP fall with sleep is progressively
reduced with increasing plasma creatinine pre-dialysis CRF (data from a large
cohort of CRF patients). Furthermore, a general influence of the “uremic
milieu” is supported by the observations that during 48 hrs. ABPM recordings
(1st and 2nd day of the interdialytic interval) the night-time / day-time
ABPM-1 ratio < night-time / day-time ABPM-2 ratio, although there is no
difference in BP levels during day-time and night-time ABPM-1 (dialytic day)
and ABPM-2 (next interdialytic day).
One of the culprits frequently investigated as a potential cause of abnormal diurnal BP rhythm is the volume status (expansion) - and related hormonal changes (atrial natriuretic peptide, renin-angiotensin-aldosterone axis) – but the existing information is conflicting. The main reason for these discrepancies is related to how volume status / expansion is to be measured and defined - dry weight, IVC-diameter, concentrations of brain natriuretic peptide, electrical bioimpedance, isotope dilution.
However,
one of the clearest pieces of evidence against volume expansion is the study
on the effect of daily HD (versus thrice weekly HD) on BP, LV mass and ECW
content (using bioimpedance) performed in 12 patients by Fagugli et al.: Short
daily hemodialysis: blood pressure control and left ventricular mass reduction
in hypertensive hemodialysis patients. AJKD 2001; 38: 371-376. The effect
of daily HD on BP, anti-hypertensive use, LV mass and ECW content was spectacular.
The effect on diurnal BP rhythm was negligible.
The role of dysregulation of the autonomic nervous
system in abnormal diurnal BP rhythm is indisputable. Spinal injury patients,
patients with autonomic syndromes (e.g. Parkinson, Shy-Drager), and diabetic
neuropaths all have profound alteration in diurnal BP. Asymptomatic, infraclinic
autonomic dysfunction is frequent in CRF, characterized by a syndrome of vagal
denervation and sympathetic over-activity. Several studies have confirmed
that non-dippers have an abnormally high active sympathetic nervous system
in sleep and/or a decrease vagal activity compared to dippers.
25 / 50 patients studied by Zoccali et al. had
at least one episode of nocturnal hypoxaemia (median 13, interquartile range
4–31). The parasympathetic control of the heart was assessed by monitoring
the RR interval as the subjects breathed deeply for 100 s. A DB score <10
is considered an abnormal response to this test. The integrity of the overall
reflex arc and efferent sympathetic activity was assessed by the arterial
pressure response to standing. On univariate analysis the DB was strongly
related to the average nocturnal SaO2 (Figure a). Similarly, the postural
change in MAP was directly related to the average SaO2 (Figure b). In multivariable
models, besides age (ß=-0.44, P=0.001), average nocturnal SaO2 (ß=0.32,
P=0.01) was the only independent predictor of the DB score. Sex, duration
of dialysis treatment, 24-h MAP, BMI, haematocrit, serum albumin, and PTH
had no independent effect on the DB score.
In the last decade a very exciting potential
reason for this excess sympathetic drive during sleep has emerged - obstructive
sleep apnoea (OSA). This abnormality is commoner with age and obesity, and
may also have an important role on BP abnormalities, and sudden death, in
diabetes, heart failure and old age. In renal patients excellent studies by
the group of Zoccali have shown that OSA is common, associated with increased
sympathetic drive, nocturnal BP and also LV mass. As this condition can be
treated using continuous pressure airway support this opens up the possibility
of a successful intervention.
In dialysis patients the abnormalities in the diurnal rhythm increase in time and are not corrected by transplantation immediately nor completely, suggesting a potential role for some sort of structural dammage in the cardiovascular system.
These vascular changes are frequently seen in CRF patients and increase with age, advancing uraemia, time on dialysis.
Arterial stiffness is common in dialysis patients, reflected by increased PWV and AGI. PWV is related in dialysis patients with age, presence of diabetes and calcifications scores. Both AGI and PWV represent strong predictors of survival.
Several studies have linked abnormal BP variability
(nocturnal/diurnal BP ratio, SD of the 24 hrs. BP mean) with structural arterial
alterations and possibly increased arterial stiffness in essential hypertensives.
We investigated differences in arterial stiffness and endothelial vasomotor
function according to circadian variability profiles, in HD subjects. In HD
patients a profound abnormal circadian variability profile is associated with
stiffer arteries and with important abnormalities in NO-dependent vasodilatation,
supporting a central aetiologic role for the uraemic arteriosclerosis in blunted
BP diurnal rithmicity.
We (Covic and Goldsmith) hypothesize that the calcifications and mural thickening of large arteries will “splint” baroreceptors and play a role in the dysautonomia of uraemia (arteriosclerosis and autonomic dysfunction have been previously linked by London et al 1993). Dysautonomia and vessel structural alterations will not alter significantly / quickly after RTx thus explaining the lack of marked improvement in diurnal BP rhythmicity post-RTx. Our hypothesis will also explain the high short-term variability (expressed as SD of the mean BP) recorded in CRF patients. 24 hrs. SD values reported in uremia are much higher than those reported for normals: 12/10 (Covic NDT 1996), 21.3/13.7 (Chazot, NDT 1995) vs a maximum level of 11/7 (O’Brien J. Hypertension 1991); prevalence of abnormal SD: 72-82%!
A proposed integrative scheme linking abnormal
circadian variability with structural and functional vascular abnormalities,
autonomic dysfunction, sleep related hypoxemia, and uraemic cardiomyopathy.
The diabetic renal patient represents a special
case: the impaired function of the autonomous nerve system may be particularly
severe and this category of subjects have a higher extracellular volume (latent
overhydration) – as major contributors of an abnormal circadian rhythm.
Since the abnormal diurnal rhythm is so frequent in (renal) dialysis patients, and as discussed carries an important negative prognostic significance, a legitimate and central question from any nephrologist is if, today, we have effective therapeutic strategies. There are some answers, but clearly the only existing evidence is provided by small trials, without controls. ACE-I may have, compared to other antihypertensives, a beneficial effect.
Results from the HALT study suggest that, in essential hypertension, diurnal BP rhythm can be successfully modulated using targeted alpha-blockade with doxazosin, and by extrapolation, this strategy may prove to be relevant for over sympathethc hyperactivity states: CRF, RTX.
Kario et al studied the effect of night-time
dosing of an alpha(1)-adrenergic blocker, doxazosin, on the BP dipping status
of 118 hypertensives, all of whom underwent 24-hour ABPM before and after
treatment. The mean night-time/daytime ratio of systolic BP was increased
(0.91 after therapy versus 0.89 at baseline, P<0.05). The patients were
initially divided into 4 groups on the basis of their dipping status at the
baseline assessment: 15% were extreme dippers, with a nighttime systolic BP
fall of at least 20% of daytime BP; 39% were dippers (fall between 10% and
20%); 41% were non-dippers (fall between 0% and 10%); and 5% had a nocturnal
increase of systolic BP. In this trial the effects of doxazosin on the mean
nocturnal systolic BP changes were an increase of 4.3 mm Hg in extreme dippers
and decreases of 0.7 mm Hg in dippers, 12 mm Hg in "non-dippers",
and 18 mm Hg in risers; the reduction was only significant in the latter 2
groups (both P<0.01). In contrast, there was a substantial drop in daytime
SBP after doxazosin therapy that did not vary by group.
Palatini et al performed ABPM in 18 patients
with hypertension to assess whether timing of administration can influence
the antihypertensive effect of an ACE-I. Quinapril, 20 mg, was given at 8
AM or 10 PM for 4 weeks in a double-blind crossover fashion. The 24-hour blood
pressure profiles showed a more sustained antihypertensive action with the
evening administration of quinapril compared with the morning administration;
as with the morning administration, a partial loss of effectiveness was observed
during night time hours. Measurement of ACE activity showed that evening administration
caused a less pronounced but a more sustained decline of plasma ACE.
In renal transplantation ABPM should be routinely performed since the few existing studies indicate large differences between casual and 24-hour BP measurements: only 63% of 27 renal transplant pediatric patients were found to be in the same BP category (i.e. either hypertensive or normotensive) by both methods [Lingens, Ped. Nephrol. 1997, 11: 23]. Furthermore, no systematic algorithm or regression equation can be determined from comparative studies, since ABPM is both underestimated and overestimated by CBP (Kooman, AJKD 2001, 37:1170).
Moreover, in renal transplant subjects, ABPM
parameters are stronger predictors of the renal function or LV mass, compared
with casual BP levels.
A non-dipping pattern is extremely common –
up to 90% in numerous studies.
However, again the non-dipping prevalence is variable according to the definition used, probably similar to dialysis populations (i.e. = 95% if more stringent definitions are used – see table). More importantly, one third of the patients are inverted dippers (higher nocturnal BP levels – for the particularly deleterious significance vide supra Nakano et al.
Indeed, Faria in 12 RTx patients on CyA found
that although non-dipping was ubiquitous 8-10 days after transplantation,
there was a very early tendency of improvement after 35-40 days, associated
with the decrease in fluid overload and the reduction of immunosuppressive
drugs. In a longer-term observation, Gatzka demonstrated that the prevalence
of dippers increased from 27% in the early post-transplant phase (< 7 months)
to 73% in the late phase (> 1 year), independently of the mean 24-hour
BP level and of the antihypertensive or immunosuppressive medication.
In those cases where the NDP persists, the graft function is often altered. In the small cohort studied by Lingens, a non-dipping pattern was found only in patients with renal pathology: renal artery stenosis, chronic rejection, recurrent FSGS and past acute rejection. Koomans et al studied only patients with CAN and found a significant relation between the nightly decline in mean BP and the creatinine clearance, regardless of the time after transplantation and type of immunosuppressive therapy.
CONCLUSIONS: ABPM should be routinely performed 1-2 times a year, since marked BP short-term variability, and abnormal (often reversed) diurnal BP variability are frequent in renal disease, with important proved consequences in terms of disease progression, CDV complications and survival, in all categories of renal subjects. ABPM reproducibility is somehow limited, but often “a blunt axe may fell a tree”. Uraemia-induced vascular structural changes and the autonomic dysfunction are possibly the most important aethiological factors of these BP variability abnormalities.
