Hypertension and Cardiovascular Risk Factors in

Hemodialyzed Diabetic Patients

EBERHARD Rrrz, CORNELIA STRUMPF, FRIEDER KATZ, ANTONY J. WING,
AND EDUARD QUELLHORST

SUMMARY In a retrospective study, the cause of death and the cardiovascular risk conferred by
hypertension and other risk factors were analyzed in 200 diabetic and 200 nondiabetic patients who
were matched for age, sex, year of admission, and center of treatment. Total and cardiovascular
mortality were considerably higher in diabetics, cardiovascular mortality being 4.8 times higher in
patients with type I and 3.0 times higher in those with type II diabetes compared to matched controls.
Cardiovascular mortality progressively increased with age and had not improved in recent years. In
both types I and II diabetes, the rate (58%) and proportion (38%) of deaths from cardiovascular
causes were significantly higher in diabetics than in matched controls. Myocardial infarction (13%)
and stroke (7%) accounted only for a minority of cardiovascular mortality, the majority (80%) being
due to "sudden death of unknown cause." Autopsy was carried out in 33% of patients with sudden
death. A documented history of long-standing hypertension increased cardiovascular death in diabetic
more than in nondiabetic patients. Diabetic retinopathy (an index of microangiopathy) and absence of
peripheral pulses, amputation, or history of myocardial infarction, stroke, or transient ischemic
attacks (as evidence of macroangiopathy) caused surprisingly little increase in relative risk for cardio-
vascular death. In diabetics but not hi nondiabetics, cardiomegaly, particularly in association with
electrocardiographic abnormalities, was a strong predictor of cardiovascular death.
(Hypertension 7 [Suppl II]: 11-118-0-124, 1985)

KEY WORDS • diabetes mellitus • hemodialysis • cardiovascular death

macroangiopathy

I N all studies reported to date, survival of patients

with diabetic uremia who are receiving mainte-
nance hemodialysis has been markedly inferior
to that of those with nondiabetic uremia.1"5 One-year
survival rates reported by the European Dialysis and
Transplantation Association (EDTA) registry for dia-
betics treated solely by hemodialysis were 62% for
those age 0 to 40 years at start, 67% for those 40 to 60
years, and 59% for persons over age 60. High cardio-
vascular mortality is the single most important factor
for poor survival of diabetics receiving hemodialysis.
The percentage of death from cardiovascular causes in
diabetics varies from 23% 2 to 54.2%. 3
From the Department of Internal Medicine, University of Heidel-
berg, Heidelberg, West Germany (E. Ritz, C. Strumpf, and F.
Katz); EDTA Registry, St. Thomas' Hospital, London, United
Kingdom (A.J. Wing); and Nephrologisches Zentrum Niedersach-
sen, Hann.-Mflnden, West Germany (E. Quellhorst).
Address for reprints: Professor Dr. Eberhard Ritz, Medizinische
Univ.-Klinik, Sektion Nephrologie, Bergheimer Strasse 56 a,
D-6900 Heidelberg, West Germany.
11-118
Downloaded from http://hyper.ahajournals.org/ at UNIVERSITY OF ALBERTA LIBRARY on March 13, 2015
The overriding importance of hypertension in the
prognosis of dialyzed diabetic patients has been recog-
nized by many authors; however, quantitative infor-
mation on the relationship of cardiovascular mortality
in these patients to hypertension and other risk factors,
as well as risk estimates with respect to nondiabetic
persons, is currently not available. The present retro-
spective multicenter study was designed specifically to
answer these questions. A preliminary report of the
study was given elsewhere.6
Patients and Methods
Patient Population
All diabetic patients who entered uremia treatment
programs in 17 German dialysis centers between Janu-
ary 1, 1972 and December 31, 1983 and who were
recorded on the EDTA registry were entered into the
study. In all, 228 diabetic patients were identified. The
hospital records of 24 patients could not be traced and
4 were found to have been erroneously reported as
having diabetes, leaving a total of 200 patients for
 HYPERTENSION IN DIALYZED DIABETIC PATIENTS/fli/z et al.
analysis. Patients who subsequently either underwent
kidney transplantation (9 of 200, 8 of whom had type I
diabetes) or received peritoneal dialysis (9 of 200, 3 of
whom had type I diabetes) were excluded. Of the re-
maining 182 hemodialyzed diabetic patients, 2.7% re-
ceived limited care dialysis, 2.7% home dialysis, and
94.6% center hemodialysis. Diabetes was classified as
type I or type II according to the recommendations of
the National Diabetes Data Group7; classification was
based on case histories obtained from interviewing the
patients or their private physicians or analyzing pa-
tients' records.
Included in the final analysis were 58 patients with
type I diabetes (median age at entry 41.1 years, range
21-67 years; 33 men, 25 women; median duration of
diabetes 24 years, range 4-56 years); 111 patients with
type II diabetes (median age at entry into dialysis 61.2
years, range 37-67 years; 59 men, 52 women; median
known duration of diabetes 12.7 years, range 0.25-27
years); and 13 patients with nonclassified diabetes
(median age at entry 55.3 years, range 47-72 years; 6
men, 7 women). Case controls were obtained by a
manual hierarchical search of the EDTA files. Patients
were listed for each center according to dates of first
treatment. This list was searched alternatively on ei-
ther side of the diabetic patients until two patients were
found of the same sex and age (± 5 years). The search
was extended to only 1 year on either side. Thefirstof
the two patients whoserecordswereretrievedwas used
as the control. Of the 200 controls entering treatment
programs, 14 underwent transplantation and 3 re-
ceived peritoneal dialysis, leaving a total of 183 pa-
tients for analysis. In unbalanced pairs where either the
diabetic patient or matched control was lost for analy-
sis because of transplantation or transfer to peritoneal
dialysis, the remaining partner was used for further
analysis. Of the 183 control patients included in the
final analysis, 16 (8%) received limited care dialysis,
10 (6%) home dialysis, and 157 (86%) center hemo-
dialysis. In the following discussion, the appropriate
matched controls for patients with type I or II diabetes
or the combined total controls for all diabetic patients
are given as indicated in the tables.
Analysis of Cardiovascular Risk Factors
Taking the date of admission to the dialysis program
as thereferencepoint, information concerning the fol-
lowing items was obtained from available written pa-
tient records and/or interview of the physician in
charge. All data were not available for every patient;
the number of patients for whom the respective infor-
mation was lacking is given in parentheses.
1. Documented history of long-standing hypertension: This
was assumed to be present if at least 5 years prior to dialysis
blood pressure was documented to be above 160/95 mm Hg
on three separate occasions or if antihypertensive medica-
tion was given (information missing in 34 patients and 37
controls).
2. Systolic blood pressure (phase I) at time of admission into
dialysis: The mean of 5 measurements at the time of first
dialysis was taken (information missing in 5 patients and 6
controls).
Downloaded from http://hyper.ahajournals.org/ at UNIVERSITY OF ALBERTA LIBRARY on March 13, 2015
11-119
3. Electrocardiographic (ECG) evidence of left ventricular
hypertrophy (LVH): Information was obtained either from
records or from inspection of the electrocardiogram when
available; LVH was defined as positive Sokolow index
and/or at least 5 points on the Romhilt-Estes scale8 (infor-
mation missing in 19 patients and 11 controls).
4. Additional ECG abnormalities: Evidence of ECG abnor-
malities other than LVH, such as bundle branch block,
depression of ST segment, inverted T wave, and so on was
obtained from reports or electrocardiograms when avail-
able (information missing in 12 patients and 8 controls).
5. Documented cardiomegaly at the time of first dialysis, de-
fined as cardiothoracic ratio greater than 0.5 (information
missing in 12 patients and 15 controls).
6. Presence of diabetic retinopathy as diagnosed by a written
record of an ophthalmologist (record missing in 27 pa-
tients); records did not permit unequivocal distinction be-
tween background and proliferative retinopathy.
7. Documented absence of one or more peripheral pulses on
physical examination (information missing in 21 patients
and 44 controls).
8. Recorded amputation (information missing in 19 patients).
9. Documented history of myocardial infarction (ECG or en-
zyme changes), stroke, or transient ischemic attacks (TIA).
10. History of angina pectoris reported by the patient and docu-
mented in the patient records at the time of admission to
dialysis (no information in 16 patients and 19 controls).
11. Serum cholesterol below 260 mg/dl (SMA-12 autoanalyzer
technique) (no information in 20 patients and 22 controls).
End Points
The times and causes of death were taken from rec-
ords or consultations with family members or private
physicians. The following events were classified as
cardiovascular mortality: sudden death (sudden de-
fined as less than 1 hour, cardiac arrest, cause un-
known9), death after documented myocardial infarc-
tion (ECG changes and/or enzyme changes), or death
after stroke. Autopsy records were available for 23
(25%) of 93 deceased diabetics and 13 (29%) of 45
deceased controls. Autopsy was performed on 3 of 7
diabetic patients with the diagnosis of myocardial in-
farction, none of the 4 with stroke, and 14 (33%) of the
42 with sudden death. In controls, autopsy records
were available for 1 of 3 dying from myocardial infarc-
tion, none of 2 dying from stroke, and 4 of 12 dying
from sudden death.
Statistical Evaluation
Actuarial survival was calculated using standard
techniques.10 The relative risk was calculated as R =
altiycln^ from a contingency table, where a equals the
number of dead with risk factor; c equals the number of
dead without risk factor; n, equals the number of all
individuals with risk factor; and n2 equals the number
of all individuals without risk factor. Significances for
relative risks PA/PB w e r e calculated as R = dn^aln^
from contingency tables. The statistical significance of
relative risks was evaluated according to Sachs."
Comparison of contingency tables (between diabetic
and nondiabetic patients) was carried out according to
Le Roy and the differential interpretation according to
Steingruber.12 To avoid a Bonferroni type of error from
II-120 DIABETES AND HYPERTENSION SUPPL II HYPERTENSION, VOL 7, No 6, NOV-DEC 1985

multiple testing, the findings were considered signifi-
cant only when p < 0.01.
Results
Survival and Cardiovascular Death
Actuarial survival was lower in patients with type I
and type II diabetes than in their respective matched
controls. In the observation period, 55.9% of female
(median age 53.9 years) and 46.9% of male patients
(median age 54.9 years) died.
Figure 1 shows actuarial rate of loss of patients due
to cardiovascular death for those with types I and II
diabetes and controls. It is obvious that the number of
deaths from cardiovascular causes in diabetics of all
groups vastly exceeded that of controls. As shown in
Figure 2, cardiovascular mortality progressively in-
creased with age in diabetic patients who received he-
modialysis. The overall proportion of patients with
type I diabetes who died from cardiovascular causes
was 4.8 times higher than in matched controls (p <
0.001); for those with type II diabetes the respective
proportion was increased 3.0 times (p < 0.01).
100
Analysis of total mortality and cardiovascular mor-
tality according to year of admission to dialysis did not
reveal significant differences among diabetic patients;
that is, actuarial cardiovascular mortality at 4 years
prior to 1979 was 38% and after 1979 it was 48%.
Patients admitted after 1979, however, tended to be
older, and treatment modalities were not comparable
between the periods. Total mortality and cardiovascu-
lar mortality varied considerably among centers; for
example, 5-year actuarial cardiovascular mortality was
20 to 55%.
Causes of Death
Table 1 lists the causes of death for patients with
types I and II diabetes and for all diabetic patients and
their controls. It is obvious that the rate and number of
deaths from cardiovascular causes among all causes of
death was significantly higher in patients with both
types of diabetes than in controls. Further analysis
showed that in the first 18 months of dialysis, 61.4%
(35 of 57 deaths) were due to cardiac causes, whereas
in diabetic patients dialyzed for more than 18 months
the percentage decreased to 41.6% (15 of 36 deaths).
In nondiabetic controls, the respective percentages
were 28.0% in those dialyzed for less and 40% in those
dialyzed for more than 18 months. There were no sig-
nificant sex differences in either patients or controls.

Risk Factor Profile in Diabetic Patients

60 —<—' TYPE I Treatment of hypertension in diabetic patients ad-
mitted to hemodialysis was unsatisfactory. Of 200 dia-
betic patients, 184 (92%) had hypertension. Of these
184, 7 (3.8%) had no antihypertensive treatment and
CONTROL 10 TYPE I
only 24 (13%) patients achieved systolic pressure less
CONTROL 10 TYPE II
20 • than 160 mm Hg with antihypertensive therapy. Hy-
pertension tended to be treated better in nondiabetic
0J controls. Of the 200 controls, 156 (78%) had hyperten-
8 YEARS
sion. Of these 156 controls, 11 (7.1%) had no anti-
hypertensive medication and 101 (64.7%) had in-
sufficient antihypertensive treatment, but 44 (28.2%)
FIGURE 1. Cardiovascular mortality in patients with type I or
achieved a systolic pressure less than 160 mm Hg with
type II diabetes and their matched controls receiving mainte-
antihypertensive therapy. There was a strong relation-
nance hemodiafysis.
ship between age and clinical evidence of macroangi-
opathy (Table 2).
1. Hypertension was interrelated with other risk fac-
100 tors. A higher proportion of patients with angina pec-
toris was hypertensive (90.2%) than those without an-
80 -
gina pectoris (67.5%), but angina pectoris was
unrelated to hypercholesterolemia or overweight, ac-
60 -
cording to the Broca formula. Peripheral arterial dis-
>60 Y
ease was not significantly related to hypertension
41-50 Y (present in 83.3% of patients with and 78.4% without
41-60 Y
51-60 Y
21-40 Y
peripheral arterial disease). Hypertension was not re-
lated to overweight (present in 89.1% of patients with
20 -
and 78.9% without overweight).

Relationship Between Cardiovascular Risk Factors on

S YEARS ( ON DIALYSIS) Admission for Dialysis and Subsequent Cardiovascular
Death
FIGURE 2. Cardiovascular mortality in diabetic patients un- Table 3 gives the relative risk for cardiovascular
dergoing hemodiafysis according to age. death with respect to patients' clinical status on admis-

Downloaded from http://hyper.ahajournals.org/ at UNIVERSITY OF ALBERTA LIBRARY on March 13, 2015

 HYPERTENSION IN DIALYZED DIABETIC PATIENTS/fl/rz et al. 11-121

TABLE 1. Causes of Death in Diabetic Patients and Controls

Diabetic patients Controls
Non- Total Non- Total
Cause Typel Type II classified diabetics Typel Type II classified controls
Sudden death 17/34 21/50 5/9 (43) 2/11 7/27 3/7 (12)
Myocardial infarction 0/34 5/50 2/9 (7) 1/11 2/27 0/7 (3)
Stroke 2/34 2/50 0/9 (4) 1/11 1/27 0/7 (2)
Total cardiovascular deaths 19/34 28/50 7/9 54 (58%) 4/11 10/27 3/7 17 (38%)
Infection 2/34 7/50 1/9 10 (11%) 0/11 1/27 0/7 1 (2%)
Other 7/34 2/50 0/9 9 (10%) 4/11 5/27 in 11 (24%)
Unknown causes 6734 13/50 1/9 20 (21%) 3/11 11/27 111 16 (36%)
Total dead 34 50 9 93 (100%) 11 27 1 45 (100%)
Total observed 58 111 13 182 58 HI 14 183

sion to dialysis. A similar analysis for total (as opposed LVH had significant predictive value for subsequent
to cardiovascular) mortality showed that none of the cardiovascular death. Such abnormalities were present
initial risk factors had higher predictive value for total in 29.5% of diabetic patients and 18.1% of controls.
as opposed to cardiovascular mortality (data not giv- Their presence increased the proportion dying from
en). The overall rate of death in diabetic patients ana- cardiovascular causes considerably (3.04 times) in
lyzed in Table 3 was approximately three times higher controls and to a lesser extent (1.66 times) in diabetic
than in controls (33.8% vs 11.2%). The prevalence of patients.
the respective risk factors (in parentheses) was gener- Subsequent cardiovascular death was strongly relat-
ally higher in diabetics than in controls. Although sys- ed to cardiomegaly at admission (Table 6) in diabetic
tolic blood pressure greater than 160 mm Hg signifi-
cantly increased the relative risk for cardiovascular
death, relative risk did not increase with increasing
systolic pressure. TABLE 3. Relative Risk for Cardiovascular Death
The relationship of clinical status at entry to dialysis
Total Total
and subsequent cardiovascular death is analyzed in diabetics controls
more detail in Tables 4, 5, and 6. Table 4 shows the n = 151 n = 152
relationship of cardiovascular death to history of hy- History of hyper-
pertension (more than 5 years of BP ^ 160/95 mm tension > 5 years 2.3 (89.7%) 1.3 (80.4%)
Hg). The percentage of patients with a history of hy- BP at start of dialysis
pertension was not significantly higher in diabetics < 160 mm Hg 0.4* (21.9%) 0.4 (35.1%)
(89.7%) than in controls (80.4%). Even in patients BP at start of dialysis
with no history of hypertension, the number dying > 160 mm Hg 2.5* (78.1%) 2.5 (64.9%)
from cardiovascular causes was higher in diabetics. In Combination of 1 and 3 2.8t (72.0%) 2.7 (52.6%)
patients with a history of hypertension, the proportion
ECG evidence of LVH 1.5 (52.1%) 1.9 (30.7%)
of diabetic individuals dying from cardiovascular
causes significantly (p < 0.001) exceeded that of non- Additional ECG abnor-
malities 1.7* (29.5%) 3.0* (18.1%)
diabetic controls.
As shown in Table 5, ECG abnormalities other than Cardiomegaly 1.7* (63.5%) 0.8 (51.2%)
Combination of 6 and 7 2.3t (23.5%)
Diabetic retinopathy 2.0 (82.7%)
TABLE 2. Macroangiopathy and Age in Diabetic Patients (Types
I and II) Admitted to Dialysis Absence of lower extrem-
ity pulses 1.4 (37.1%) 4.4t (25.4%)
Peripheral
Amputation 1.7 (18.4%)
vascular
disease History of History of myocardial in-
(absent myocardial farction, stroke, or TIA 1.6 (20.1%) 2.9* (23.0%)
pulses infarction, Combination of 10, 11,
or stroke, Angina and 12 0.9 (15.2%) 4.0* (14.3%)
Age (yr) n amputation) or TLA pectoris
Angina pectoris 1.8* (38.3%) 5.2t (30.1%)
21-40 33 9 (27.3%) 4 (12.1%) 8 (24.2%)
Hypercholesterolemia 1.2 (31.9%) 1.3 (25.6%)
41-50 37 16 (43.2%) 6 (16.2%) 10 (27.0%)
Significant increase or decrease of relative risk (in groups of
51-60 54 27 (50.0%) 12 (22.2%) 18 (33.3%) diabetics or controls respectively), Numbers in parentheses repre-
60 58 27 (46.4%) 19 (32.8%) 27 (46.6%) sent prevalence of risk factor.
•Nominal p < 0.05.
Totals 182 79 (43.4%) 41 (22.5%) 63 (34.6%)
tNominalp < 0.01.

Downloaded from http://hyper.ahajournals.org/ at UNIVERSITY OF ALBERTA LIBRARY on March 13, 2015

11-122 DIABETES AND HYPERTENSION SUPPL II HYPERTENSION, VOL 7, No 6, NOV-DEC 1985

TABLE 4. Relationship of Cardiovascular Death to History of Hypertension

Diabetic patients Controls
Cardiovascular deaths Cardiovascular deaths
Yes No Totals Yes No Totals
History of hypertension 40(38.1%) 65 105 11 (12.8%)* 75 86
No history of hypertension 2(16.7%) 10 12 2 (9.5%) 19 21
The grand totals do not add up to the total number of diabetics or controls studied. In this and the following tables,
patients changing to transplantation or CAPD (including their matched controls) were excluded, as were patients with
missing information or unknown causes of death. The percentages refer to the proportion of patients with (upper line) or
without (lower line) the respective risk factor dying from cardiovascular causes.
Difference (*p < 0.05; t p < 0.01; %p < 0.001 respectively) of proportion dying from cardiovascular deaths for
diabetic versus nondiabetic patients with respective risk factors.

TABLE 5. Relationship of Cardiovascular Death to ECG Abnormalities Other than LVH

Diabetic patients Controls
Cardiovascular deaths Cardiovascular deaths
Yes No Totals Yes No Totals
ECG changes 18 (43.9%) 23 41 6* (23.1%)* 20 26
No ECG changes 26 (26.5%) 72 98 9 (7.6%)t 109 118
See notes for Table 4.

TABLE 6. Relationship of Cardiovascular Death to Cardiomegaty on X-ray at Admission to Dialysis

Diabetic patients Controls
Cardiovascular deaths Cardiovascular deaths
Yes No Totals Yes No Totals
Cardiomegaly 36 (38.3%)* 58 94 6 (9.0%)t 61 67
No cardiomegaly 12 (22.2%) 42 54 7 (10.9%) 57 64
See notes for Table 4.

patients but not in controls. Sixty-four percent of dia-
betic patients and 5 1 % of controls had cardiomegaly.
Cardiomegaly significantly increased the proportion of
deaths from cardiovascular causes in diabetics (1.73
times) but not in controls (0.83 times).
The presence of both peripheral arterial disease (ab-
sent pulses or amputation) and history of end organ
damage in the coronary or cerebral circulations (i.e.,
myocardial infarction, stroke, or TIA) increased the
relative risk of cardiovascular death 4.0 times in con-
trols but failed to change the relative risk significantly
(0.86 times) in diabetic patients.
Discussion
The present study documented that hypertension is
almost uniformly present when diabetic patients are in
terminal renal failure. This is in agreement with nu-
merous other reports. 3 4-13~16 Of particular note is the
finding that hypertension was much less effectively
controlled in diabetics than in matched controls ad-
mitted for hemodialysis. In a previous study,13 we not-
ed that of those receiving hemodialysis, a considerably
higher proportion of diabetic (50%) than nondiabetic
(27.7%) patients required antihypertensive medica-
tion. Such medication was mostly needed because pa-
tients did not comply with fluid restriction and/or be-
cause ultrafiltration was not well tolerated. Similar
problems of hypertension control prior to and during
dialysis may explain, at least in part, why a history of
hypertension tended to affect cardiovascular prognosis
more adversely in diabetic persons than in matched
controls. The finding that hypertension at entry into
dialysis predicted the risk of cardiovascular mortality
was in agreement with findings of Rostand et al.17
The high cardiovascular mortality noted in dialyzed
diabetics agreed with previous reports.2 ' 8 In our se-
ries, myocardial infarction accounted for only 7 of 54
cardiovascular deaths in diabetic patients, none occur-
ring in those with type I diabetes. The low proportion
of documented myocardial infarction agreed with re-
cent studies2 recording only 4 cases of myocardial
infarction among 35 deaths in insulin-dependent dia-
lyzed diabetics. In our study, "sudden death of un-
known cause" was the major cause of death. Although
admittedly, this category may include some patients
with unrecognized suicide, hyperkalemia (particularly
after omission of insulin injections), or unanticipated
death secondary to advanced autonomic polyneurop-

Downloaded from http://hyper.ahajournals.org/ at UNIVERSITY OF ALBERTA LIBRARY on March 13, 2015

 HYPERTENSION IN DIALYZED DIABETIC PATIENTS/fl;7z et al.
athy, it is unlikely that many myocardial infarctions
were missed because 33% of the patients had a post-
mortem examination. Sudden death was equally fre-
quent in patients with insulin-dependent and non-
insulin-dependent diabetes, thus rendering unlikely
hyperkalemia secondary to withdrawal of insulin. Fur-
thermore, it would be difficult to explain why suicide
or hyperkalemia should be predicted by hypertension
and other risk factors.
Hypertension may cause cardiovascular death in
diabetics by one of several mechanisms. More severe
coronary atherosclerosis in diabetics has been shown
in studies using coronary angiography19 and in case
control autopsy studies using both coronary angiog-
raphy and histology.20 Diabetic persons with angina
pectoris have poorer prognosis, and those with myo-
cardial infarction have greater immediate and delayed
mortality.21 In addition, the frequency of myocardial
infarction is presumably higher,22 although this has
been questioned.23
Previous reports documented high prevalence and
poor prognosis of coronary lesions in diabetic patients
with uremia admitted to hemodialysis.24"26 Weinrauch
etal. 25 noted 2-year survival of 88% in 12 patients with
type I diabetes without coronary lesions and only 20%
in 9 dialyzed patients with type I diabetes with coro-
nary lesions. Kjellstrand et al. 4 found that cardiovascu-
lar mortality was much higher in initial months of
dialysis than later on, which was confirmed in the
present study. They suggested that a cohort of patients
with preexisting coronary lesions die soon after begin-
ning hemodialysis.
Several findings of our study, however, are com-
patible with the role of additional factors other than
coronary lesions. In a prospective study, using Dopp-
ler measurements of lower leg arterial pressure, 9.6
times higher cardiovascular mortality was found in
nonuremic diabetic patients with peripheral vascular
disease.27-28 No such increase was found in dialyzed
diabetic patients with evidence of macroangiopathy.
Furthermore, in dialyzed diabetics, a history of myo-
cardial infarction, stroke, TIA, or angina pectoris in-
creased the risk of subsequent cardiovascular death
much less than in nondiabetic persons, and in dialyzed
patients, cardiovascular mortality was high even in the
absence of documented risk factors (see Table 4).
Evidence for noncoronary cardiac disease in diabet-
ics comes from the demonstration of periodic acid-
Schiff (PAS)-positive hyaline material in extramural
and intramural coronary arteries29 and the interstitial
deposition of PAS-positive material and collagen in
the myocardium of diabetic patients.30-31 Further evi-
dence of cardiomyopathy was provided by Regan et
al.32 who found elevated left ventricular end-diastolic
pressure and diminished stroke volume index in dia-
betic patients even in the absence of coronary lesions,
ischemia, or myocardial dyskinesia. Cardiomyopathy
as a result of hypertension, diabetes, or both might
explain our finding that cardiomegaly on admission to
dialysis predicted subsequent cardiovascular death in
diabetic (but not in nondiabetic) patients. Overhydra-
Downloaded from http://hyper.ahajournals.org/ at UNIVERSITY OF ALBERTA LIBRARY on March 13, 2015
11-123
tion as an alternative cause of cardiomegaly is not
excluded, but it would be difficult to see why at admis-
sion to dialysis that should have had adverse conse-
quences only in dialyzed diabetics. The hypothesis of
an important role of noncoronary cardiomyopathic
mechanisms of cardiac death in dialyzed diabetic per-
sons would explain the rarity of myocardial infarction
and the predictive value of cardiomegaly; it would also
have interesting implications for treatment modalities.
Decreased left ventricular compliance from diabetic
cardiomyopathy, as demonstrated in experimental33
and clinical34 studies, would render dialyzed diabetics
more susceptible to left ventricular underfilling and.
hypotension during ultrafiltration. Hypotensive
episodes are more common in these patients33 and
might even cause more severe myocardial ischemia be-
cause of impaired autoregulation in diabetes.34 We
have been struck by anecdotal observations of wide-
spread myocardial fibrosis in dialyzed diabetic pa-
tients who had patent coronary arteries (unpublished
observations, 1984).
As reported elsewhere,35 actuarial 5-year survival of
diabetic patients was 70% for those receiving hemofil-
tration, 56% CAPD, and only 34% hemodialysis, but
patients were not randomly allotted to treatment mo-
dalities. Sudden death of unknown cause accounted for
4 of 11 fatalities of subjects undergoing hemodialysis,
but only 1 of 7 for hemofiltration. It has been proposed
that electrical instability might explain adverse cardiac
effects of hemodialysis in diabetics, since in the same
patients, episodes of cardiac arrhythmia (Holter moni-
toring) were more prevalent after hemodialysis, de-
spite control of serum potassium, than after hemofil-
tration.36 Posttreatment cardiac arrhythmia was related
to hypotensive episodes during treatment irrespective
of treatment modality. Cardiomyopathy in dialyzed
diabetics may predispose to intradialytic hypotension,
posthypotensive arrhythmia, and sudden death. Better
survival of patients receiving hemofiltration may be
due to fewer hypotensive episodes. This interpretation
is clearly hypothetical and must be tested in controlled
prospective studies.
Acknowledgments
The present study would have been impossible without the in-
valuable help of numerous colleagues in 17 German hemodialysis
centers, whose permission to study their patient's reports is grate-
fully acknowledged. We thank Dr. Richard Moore for selecting
patients from the EDTA registry and Ms. Stelz for secretariaJ help.
References
1. Jacobs C, Broyer M, Brunner FP, et al. Combined report on
regular dialysis and transplantation in Europe, XI, 1980. Proc
Eur Dial Transplant Assoc 1981;18:2-58
2. Legrain M, Rottembourg J, Bentchikou A, et al. Dialysis treat-
ment of insulin-dependent diabetic patients: ten years' experi-
ence. Clin Nephrol 1984;21:72-«1
3. Shapiro FL, Comty CM. Hemodialysis in diabetics — 1981
update. In Friedman EA, L'Esperance FA, eds. Diabetic renal-
retinal syndrome. 2nd ed. New York: Grune & Stratton,
1983:333-345
4. Kjellstrand CM, Whitley K, Comty CM, Shapiro FL. Dialysis
in patients with diabetes mellitus. Diabetic Nephropathy
1983;1:5-17
5. Jacobs C, Brunner F, Brynger H, et al. The first five thousand
11-124 DIABETES AND HYPERTENSION
diabetics treated by dialysis and transplantation in Europe.
Diabetic Nephropathy 1983;2:12-16
6. Stmmpf C, Katz F, Wing AJ, Ritz E. Cardiovascular risk
factors and cardiovascular death in hemodialysed diabetics.
Presented at the 21 st Congress EDTA-European Renal Associ-
ation, Florence, Italy, September 23-26, 1984
7. National Diabetes Data Group. Classification and diagnosis of
diabetes mellitus and other categories of glucose intolerance.
Diabetes 1979;28:1039-1057
8. Romhilt DW, Greenfield JC Jr. Estes EH Jr. Vector cardio-
graphic diagnosis of left ventricular hypertrophy. Circulation
1968;37:15-28
9. Wing AJ, Brunner FP, Brynger H, et al. Cardiovascular-relat-
ed causes of death and the fate of patients with renovascular
disease. Contrib Nephrol 1983;41:306-311
10. Merrell M, Shulman LE. Determination of prognosis in chron-
ic disease. J Chron Dis 1955;1:12-32
11. Sachs L. Statistische Methoden. Berlin: Springer-Verlag,
1982:70-72
12. Lienert GA. Verteilungsfreie Methoden in der Biostatistik.
Meisenheim/Glan: Anton Hain Verlag, 1973:571-574
13. Hasslacher CH, Rambausek M, Ritz E. Genesis and manage-
ment of hypertension in diabetic nephropathy. In: Keen H,
Legrain M, eds. Prevention and treatment of diabetic nephrop-
athy. Lancaster, England: MTP Press, 1983:177-189
14. Ritz E, Bommer J, Andrassy K. Behandlung des Diabetikers
mit terminaler Niereninsuffizienz. MQnch Med Wochenschr
1983;125:516-520
15. Hasslacher CH, Stech W, Wahl P, Ritz E. Blood pressure and
metabolic control as risk factors for nephropathy in type I
(insulin-dependent) diabetes. Diabetologia 1985;28:6-11
16. Goldstein DA, Massry SG. Early dialysis in diabetic patients
with chronic renal failure. Nephron 1978;20:286-296
17. Rostand SG, Kirk KA, Rutsky EA. Relationship of coronary
risk factors to hemodialysis-associated ischemic heart disease.
Kidney Int 1982;22:3O4-3O8
18. Shapiro FL, Umen A. Risk factors in hemodialysis patient
survival. ASAIO Journal 1983;6:176-184
19. Hamby RO, Sherman L, Mehta J, Aitablian A. Reappraisal of
the rote of the diabetic state in coronary artery disease. Chest
1976:70:251-257
20. Vigorita VJ, Moore GW, Hutchins GM. Absence of correla-
tion between coronary arterial atherosclerosis and severity or
duration of diabetes mellitus of adult onset. Am J Cardiol
1980:46:535-542
21. Harrower AD, Clarke BF. Experience of coronary care in
diabetes. Br Med J 1976;1:126—128
22. Herman JB, Medalie JH, Goldbourt U. Differences in cardio-
vascular morbidity and mortality between previously known
Downloaded from http://hyper.ahajournals.org/ at UNIVERSITY OF ALBERTA LIBRARY on March 13, 2015
SUPPL U HYPERTENSION, VOL 7, No 6, NOV-DEC 1985
and newly diagnosed adult diabetics. Diabetologia
1977;13:229-234
23. Kvetny J. Diabetes mellitus and acute myocardial infarction.
Acta Med Scand 1976;200:151-153
24. Bennett WM, Kloster F, Rosch J, Barry J, Porter GA. Natural
history of asymptomatic coronary arteriographic lesions in dia-
betic patients with end-stage renal disease. Am J Med
1978:65:779-784
25. Weinrauch LA, D'Elia JA, Healy RW, et al. Asymptomatic
coronary artery disease: angiography in diabetic patients before
renal transplantation. Ann Intern Med 1978;88:346-348
26. Braun WE, Phillips D, Vidt DG. Coronary arteriography and
coronary artery disease in 99 diabetic and non-diabetic patients
on chronic hemodialysis or renal transplantation programs.
Transplant Proc 1981; 13:128—135
27. Janka HU, Haupt E, Standl E. Gefasskrankheiten bei Diabetes
mellitus. In: Mehnert H, Schoffling K, eds. Diabetologie.
Stuttgart, New York: Georg Thieme, 1984:405-^29
28. Janka HU, Grunwald P, Waldmann G, Standl E, Mehnert H.
Five-year mortality of diabetic patients and associated cardio-
vascular risk factors: the Schwabing study on macroangiopathy
in diabetes (SSMD). Diabetologia 1982;23:176-177
29. Ledet T. Histological and histochemical changes in the coro-
nary arteries of old diabetic patients. Diabetologia 1968;
4:268-272
30. Ahmed SS, Jaferi GA, Narang RM, Regan TJ. Preclinical
abnormality of left ventricular function in diabetes mellitus.
Am Heart J 1975;89:153-158
31. Rubier S, Slugash J, Yuceoglo YZ, Kumral T, Branwood AW,
Grishman A. New type of cardiomyopathy associated with
diabetic glomerulosclerosis. Am Heart J 1972:30:595-602
32. Regan TJ, Lyons MM, Ahmed SS, et al. Evidence of cardio-
myopathy in familial diabetes mellitus. J Clin Invest
1977;60:885-899
33. Regan TJ, Ettinger PO, Khan I, et al. Altered myocardial
function and metabolism in chronic diabetes mellitus without
ischemia in dogs. Circ Res 1974;35:222-237
34. Degoulet O, Reach I, Di Giulio S, et al. Epidemiology of
dialysis-induced hypotension. Proc Eur Dial Transplant Assoc
1981:18:133-138
35. Bentsen N, Larsen B, Lassen NA. Chronically impaired auto-
regulation of cerebral blood flow in long-term diabetics. Stroke
1975;6:497-502
36. Quellhorst E. Herzrhythmusstorungen wahrend und nach Ha-
modialyse, Hamofiltration und Hamofiltration bei Patienten
mit chronischer Niereninsuffizienz — vergleichende Langzeit-
EKG-Untersuchungen. In: Braun J, Pilgrim R, Gessler, Sey-
bold D, eds. Die Behandlung von Herzrhythmusstorungen bei
Nierenkranken. Basle, Paris, London, New York, Tokyo,
Sydney: S. Karger Verlag, 1984:23-35
 Hypertension and cardiovascular risk factors in hemodialyzed diabetic patients.
E Ritz, C Strumpf, F Katz, A J Wing and E Quellhorst

Hypertension. 1985;7:II118
doi: 10.1161/01.HYP.7.6_Pt_2.II118
Hypertension is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1985 American Heart Association, Inc. All rights reserved.
Print ISSN: 0194-911X. Online ISSN: 1524-4563

The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://hyper.ahajournals.org/content/7/6_Pt_2/II118

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Hypertension can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at:

http://www.lww.com/reprints

Subscriptions: Information about subscribing to Hypertension is online at:

http://hyper.ahajournals.org//subscriptions/

Downloaded from http://hyper.ahajournals.org/ at UNIVERSITY OF ALBERTA LIBRARY on March 13, 2015