Determinants of raised C-reactive protein concentration in type 1 diabetes

Q J Med 2000; 93:231–236
Determinants of raised C-reactive protein concentration in
type 1 diabetes
E.S. KILPATRICK, B.G. KEEVIL1, C. JAGGER2, R.J. SPOONER3 and M. SMALL4
From the Department of Clinical Biochemistry, Hull Royal Infirmary, Hull, 1Department of
Clinical Biochemistry, Wythenshawe Hospital, Manchester, 2Department of Clinical
Biochemistry, Royal Preston Hospital, Preston, and 3Department of Clinical Biochemistry, and
4The Diabetes Unit, Gartnavel General Hospital, Glasgow, UK
Received 19 November 1999 and in revised form 18 January 2000
Summary
subjects without overt CHD, multivariate analysis
showed increases in subject age ( p=0.0025), BMI
( p=0.001) and HbA ( p=0.012) to be associated
1
with a higher CRP concentration, as was female sex
( p=0.026) and a history of CHD in a first-degree
relative ( p=0.018, n=57). The duration of diabetes, current smoking status, presence of microvascular complications, lipid status and presence of
hypertension were unrelated. This study suggests
that some of the risk factors associated with CHD
in type 1 patients are also independently predictive
of high CRP concentrations. The reasons for this,
and whether intervention would prove useful,
require further investigation.
Introduction
Cardiovascular disease is the main cause of morbidity
and mortality in patients with diabetes. The reason
for this increase in atherosclerosis is multifactorial,
but is in excess of that which would be expected
when account is taken of traditional coronary heart
disease (CHD) risk factors such a smoking, hyperlipidaemia and hypertension.1–4
Atherosclerosis has been described as an inflammatory disease.5 Measurement of the acute-phase
reactant C-reactive protein (CRP) has been routinely
used to detect and monitor inflammatory changes in
patients with sepsis or connective tissue diseases.
Recently, the use of highly sensitive assays has
indicated that variations of CRP within the ‘normal
range’ i.e. <5 mg/l, are a potent risk indicator for
coronary heart disease in non-diabetic subjects.6–10
Intervention may also be successful in reducing CRP
levels10 and coronary events.7,9
This study aimed to determine factors which
predispose to high CRP concentrations in patients
with type 1 diabetes.
Methods
Patients with type 1 diabetes (n=167, 93M, 74F,
median age 30 years, range 13–67) attending the
Diabetes Unit, Gartnavel General Hospital, Glasgow,
participated in the study following written informed
consent and approval by the local Ethical Committee.
Address correspondence to Dr E.S. Kilpatrick, Department of Clinical Biochemistry, Hull Royal Infirmary, Anlaby Road,
Hull HU3 2JZ. e-mail: [email protected]
© Association of Physicians 2000
Downloaded from by guest on November 24, 2014
As a marker of systemic inflammation, raised
C-reactive protein (CRP) concentrations which are
still within the normal range have been associated
with an increased incidence of coronary heart disease (CHD) in non-diabetic subjects. This study
aimed to establish potential determinants of raised
CRP concentrations in type 1 diabetic patients. We
used a sensitive assay to measure ‘low-level’ CRP
concentrations in 167 type 1 patients (93M, 74F,
median age 30 years, range 13–67). Stepwise multivariate analysis was used to relate these CRP levels
to known cardiovascular risk factors and demographic data. Only six patients had established CHD
(median CRP 3.34 mg/l vs. 0.83 mg/l, p=0.032). In
232
E.S. Kilpatrick et al.
Results
Table 1 lists the characteristics of the diabetic patients
who participated in the study. Figure 1 shows the
log-normal distribution of CRP values found in
these patients.
Only six patients had overt coronary heart disease
(Table 1). In these patients CRP concentrations were
higher than in unaffected patients (median CRP
3.34 mg/l vs. 0.83 mg/l, p=0.032 using MannWhitney).
Table 2 shows the univariate correlations between
all the variables in the study, while Table 3 shows
those which remain significantly associated with
log CRP when the data is subjected to a forward
10
stepwise multivariate regression.
Discussion
Independent studies have shown that the measurement of baseline concentrations of CRP as a marker
of systemic inflammation can predict the risk of
future myocardial infarction in non-diabetic subjects.6–9 In one such study, healthy subjects in the
upper quartile of the CRP distribution (2.11 mg/l)
who were not taking aspirin were four times as likely
to experience a myocardial infarction than those in
the lowest (∏0.55 mg/l).7 Of importance, this excess
risk was reduced substantially by aspirin treatment,
leading to the speculation that the cardioprotective
action of the drug may be due to its anti-inflammatory
action as well as its anti-platelet effect. The effect of
the HMGCoA reductase inhibitor, pravastatin, was
similar.9 It is thought that this theory of low-grade
inflammation and atherosclerosis could therefore
bring together, in one unifying hypothesis, the apparently disparate cardiovascular risk markers of fibrinogen, ferritin and white blood cell count, since they
are all acute-phase reactants.
Our study has sought to establish some of the
determinants of raised CRP concentrations in type 1
diabetic patients. Univariate analysis (Table 2) has
shown that some of the strongest established predictors of CHD in type 1 diabetic patients (age, BMI,
duration of diabetes, albumin excretion, serum triglyceride)13 are also the measurements which are
predictive of a raised CRP concentration. These
findings are in agreement with previous studies
involving only non-diabetic subjects.8,14 Additionally,
female sex, a family history of cardiovascular disease
and a high current glycated haemoglobin value are
also related to rises in CRP. However, when these
factors are adjusted for each other by stepwise
multiple regression, only the variables in Table 3
(age, sex, family history of CHD, BMI and current
HbA ) are found to be independently associated with
1
C-reactive protein concentrations.
The female/male difference may simply be associated with an increased prevalence of subclinical
urinary infection in women. However, compared to
diabetic men, diabetic women have a higher relative
risk of ischaemic heart disease than their nondiabetic counterparts.1,4 Only one previous study of
CRP in non-diabetic individuals has included or
commented on women, and no difference was
found.8
There remains considerable debate as to whether
poor glycaemic control can lead to an increase in
CHD in diabetic patients. In the Diabetes Control
and Complications Trial (DCCT), the cardiovascular
event rate was low because of the age of the type 1
patients recruited, but there was still an excess of
macrovascular events in the conventionally-treated
compared to the intensively-treated group (40 vs.
23), although this just failed to reach statistical
significance ( p=0.08).15 In the United Kingdom
Prospective Diabetes Study (UKPDS), the event rate
amongst the type 2 patients was higher, but the
HbA separation between the two groups was lower,
1c
and again the findings were statistically suggestive
but not conclusive ( p=0.052 for myocardial infarc-
Downloaded from by guest on November 24, 2014
None had overt evidence of infection or connective
tissue diseases, and all participants had negative
urinalysis for nitrite and leukocytes. All patients also
had a baseline ECG performed at the time of
diagnosis or referral if aged >35 years.
Each had low-level CRP measured using a modified latex-enhanced immuno-turbimetric assay (BioStat Diagnostics). Between-batch precision (CV) for
the assay was 5.6%, 6.9% and 8.2% at CRP concentrations of 5.3, 2.2 and 1.0 mg/l, respectively. The
locally-derived median reference value was
0.73 mg/l (0.95 interfractile reference interval
0.2–6.4 mg/l, n=77).11
Log CRP in the patients without known CHD
10
was related by univariate regression and by forward
stepwise multiple regression to the following variables: age, sex, duration of diabetes, presence of
microvascular complications (retinopathy defined as
early background diabetic retinopathy or worse when
examined by an experienced examiner using direct
ophthalmoscopy through dilated pupils; known microalbuminuria defined as a consistent overnight timed
urine albumin excretion rate >20 mg/min, neuropathy defined as peripheral neuropathy using DCCT
clinical criteria12), body mass index (kg/m2), current
HbA , current smoking status, current spot early1
morning urine microalbumin concentration (mg/l),
current non-fasting cholesterol/triglyceride/HDL-cholesterol, hypertension (on treatment or current blood
pressure >140/90) and first-degree family history of
cardiovascular disease.
233
C-reactive protein in type 1 diabetes
Table 1 Patient characteristics
Characteristic
Patients without known
ischaemic heart disease
Patients with ischaemic
heart disease
Sex (male5female)
Age (median (IQR))
Duration of diabetes (median (IQR))
Body mass index (mean (SD))
Presence of microvascular complications
(n (%))
Current smoker (n (%))
Family history of cardiovascular disease
(n (%))
Presence of hypertension (n (%))
HbA (median (IQR))
1
Cholesterol (mean (SD))
Triglyceride (geometric mean (±SD))
C-Reactive protein (median (IQR))
89572
30 years (21–37)
12 years (5–20)
24.7 kg/m2 (3.4)
52 (32%)
452
43 years (36–54)
28 years (15–32)
27.8 kg/m2 (2.2)
1 (17%)
41 (25%)
52 (32%)
1 (17%)
5 (83%)
7 (4%)
9.6% (8.7–11.1)
4.95 mmol/l (1.05)
0.92 mmol/l (0.56–1.50)
0.83 mg/l (0.45–1.80)
3 (50%)
9.8% (8.2–10.3)
5.27 mmol/l (0.38)
1.00 mmol/l (0.63–1.58)
3.34 mg/l (1.80–4.87)
tion).16 However, recent analysis has shown that
when the whole range of UKPDS patient HbA
1c
concentrations is taken into account, there is a highly
significant relationship between HbA and coronary
1c
heart disease risk in these patients.17 It is thus of
interest that this study has found that glycated
haemoglobin is an independent predictor of CRP
levels, even though it is only the patients’ current
value.
Our data on smoking is consistent with previous
studies, in finding no relationship between current
smoking and CRP.6,14,18 However, two of these
studies did find a relationship with pack-years
smoked14,18 (an index not collected by ourselves),
indicating that the effect of smoking on CRP may be
long-lasting.
The lack of association between CRP and hyper-
Downloaded from by guest on November 24, 2014
Figure 1. Log -transformed distribution of C-reactive pro10
tein concentrations in study patients.
tension may reflect the low prevalence of raised or
treated blood pressure (6% of patients) found in this
study, presumably because of the young mean age
(30 years) of participants. Nevertheless, in studies
with non-diabetic subjects, no blood pressure/CRP
associations have been noted either.6,14 However, it
must be regarded as surprising that although the
cholesterol, triglyceride and HDL cholesterol of our
patients were related to several factors such as HbA ,
1
duration of diabetes and microvascular complications, none of these lipid parameters were independently related to CRP levels.
Given all the associations found with CRP in this
study, the reasons why inflammation is present at all
remains speculative. Some authors believe CRP
values may reflect the intrinsic inflammation and
tissue damage within arterial lesions themselves, with
severe atheroma resulting in both raised CRP levels
and an increased risk of coronary occlusion.8 Others
believe high levels may be consequent on chronic
infection with potentially atherogenic organisms such
as C. pneumoniae or H. pylori.19,20
An attractive alternative possibility is related to
the fact that CRP production by hepatocytes is
stimulated by inflammatory cytokines.21 One such
cytokine, tumour necrosis factor a (TNFa) has also
been implicated in the pathogenesis of obesityassociated insulin resistance.22 Plasma TNFa levels
correlate positively with percentage body fat and
body mass index (BMI).23 Since BMI is known to
relate to increasing subject age,24 it may help to
explain some of the associations found here. Also,
women have a larger percentage body fat than men
for a given BMI,25 so they may produce relatively
more TNFa, thereby explaining their higher CRP
values. Low-level CRP measurements in this situation
may thus partly be a surrogate marker for circulating
TNFa concentrations.
234
Table 2 Univariate relationships in patients without overt CHD (n=161)
Age
Sex
Body mass index
Microalbumin
concentration
Cholesterol
Log triglyceride
10
HDL cholesterol
Hypertension
Current smoker
Sex
r=0.02
p=0.83
r=0.18
p=0.020
r=0.78
p<0.0001
r=0.50
p<0.0001
r=−0.02
p=0.76
r=0.10
p=0.19
r=0.25
p=0.0014
r=0.15
p=0.060
r=0.19
p=0.016
r=0.09
p=0.2397
r=0.20
p=0.0095
r=−0.05
p=0.52
r=0.08
p=0.30
r=−0.11
p=0.15
r=0.10
p=0.18
r=0.08
p=0.29
r=0.12
p=0.12
r=−0.25
p=0.0009
r=0.32
p<0.0001
r=−0.07
p=0.3827
r=0.11
p=0.1847
Body mass
index
Duration of
diabetes
Microvascular
complications
r=0.23
p=0.0030
r=0.13
p=0.10
r=−0.04
p=0.59
r=0.13
p=0.11
r=0.24
p=0.0023
r=0.08
p=0.32
r=−0.06
p=0.48
r=0.05
p=0.5076
r=−0.13
p=0.1006
r=0.60
p<0.0001
r=0.03
p=0.73
r=0.16
p=0.045
r=0.32
p<0.0001
r=0.10
p=0.20
r=0.14
p=0.084
r=0.15
p=0.0539
r=0.11
p=0.1810
r=0.11
p=0.15
r=0.35
p<0.0001
r=0.20
p=0.0094
r=0.23
p=0.0033
r=0.10
p=0.19
r=0.11
p=0.1525
r=0.18
p=0.0259
HbA
1
r=0.17
p=0.027
r=0.36
p<0.0001
r=0.37
p<0.0001
r=0.09
p=0.25
r=−0.06
p=0.4730
r=0.16
p=0.0365
Microalbumin
concentration
Cholesterol
Log
triglyceride
HDL
cholesterol
r=−0.23
p=0.0037
r=−0.02
p=0.841
r=0.16
p=0.0464
r=0.10
p=0.2114
r<−0.01
p=0.9995
Hypertension
E.S. Kilpatrick et al.
Duration of
diabetes
Microvascular
complications
HbA
1
r=0.28
p=0.0003
r=0.18
p=0.019
r=0.25
p=0.0015
r=0.24
p=0.0024
r=0.26
p=0.0010
r=0.19
p=0.016
r=0.23
p=0.0030
r=0.11
p=0.18
r=0.14
p=0.067
r=0.020
p=0.80
r=0.030
p=0.7503
r=0.07
p=0.3707
Age
Downloaded from by guest on November 24, 2014
Log CRP
10
r=0.09
p=0.25
r<0.01
p=0.97
r=0.11
p=0.18
r=0.12
p=0.1410
r=0.02
p=0.7679
r=0.41
p<0.0001
r=0.28
p=0.0003
r=0.13
p=0.104
r<0.01
p=0.9972
r=0.02
p=0.8483
235
C-reactive protein in type 1 diabetes
Table 3 Determinants of C-reactive protein concentration in patients without overt CHD analysed by stepwise multiple
regression (n=161)
Variable
Relative increase in C-reactive protein
per unit increase in explanatory variable
(95% CI)
p
Body mass index (kg/m2)
Age (per 10 years)
HbA (%)
1
Family history of cardiovascular disease (yes vs. no)
Female sex (vs. male)
1.08
1.18
1.10
1.45
1.38
0.0010
0.0025
0.012
0.018
0.026
9.
10.
11.
12.
13.
References
1. Kannel WB, McGee DL. Diabetes and cardiovascular
disease. The Framingham study. JAMA 1979; 241:2035–8.
2. Stamler J, Vaccaro O, Neaton JD, Wentworth D. Diabetes,
other risk factors, and 12-yr cardiovascular mortality for
men screened in the Multiple Risk Factor Intervention Trial.
Diabetes Care 1993; 16:434–44.
3. Adlerberth AM, Rosengren A, Wilhelmsen L. Diabetes and
long-term risk of mortality from coronary and other causes
in middle-aged Swedish men, a general population study.
Diabetes Care 1998; 21:539–45.
4. Yudkin JS, Blauth C, Drury P, Fuller J, Henley J, Lancaster T,
et al. Prevention and management of cardiovascular disease
in patients with diabetes mellitus: an evidence base.
Diabetic Med 1996; 13:S101–21.
5. Ross R. Atherosclerosis- an inflammatory disease. N Engl
J Med 1999; 340:115–26.
6. Kuller LH, Tracy RP, Shaten J, Meilahn EN. Relation of
C-reactive protein and coronary heart disease in the MRFIT
nested case-control study. Multiple Risk Factor Intervention
Trial. Am J Epidemiol 1996; 144:537–47.
7. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens
CH. Inflammation, aspirin, and the risk of cardiovascular
disease in apparently healthy men. N Engl J Med 1997;
336:973–9.
8. Haverkate F, Thompson SG, Pyke SD, Gallimore JR, Pepys
MB. Production of C-reactive protein and risk of coronary
events in stable and unstable angina. European Concerted
14.
15.
16.
17.
18.
19.
20.
Action on Thrombosis and Disabilities Angina Pectoris
Study Group. Lancet 1997; 349:462–6.
Ridker PM, Rifai N, Pfeffer MA, Sacks FM, Moye LA,
Goldman S, et al. Inflammation, pravastatin, and the risk of
coronary events after myocardial infarction in patients with
average cholesterol levels. Circulation 1998; 98:839–44.
Strandberg TE, Vanhanen H, Tikkanen MJ. Effect of statins in
C-reactive protein in patients with coronary artery disease.
Lancet 1999; 353:118–19.
Keevil BG, Nicholls SP, Kilpatrick ES. Evaluation of a latexenhanced immunoturbimetric assay for measuring low
concentrations of C-reactive protein. Ann Clin Biochem
1998; 35:671–3.
The DCCT Research Group. Factors in the development of
diabetic neuropathy. Baseline analysis of neuropathy in
feasibility phase of Diabetes Control and Complications
Trial. Diabetes 1988; 37:476–81.
Winocour PH, Durrington PN, Bhatnagar D, Mbewu AD,
Ishola M, Mackness M, Arrol S. A cross sectional evaluation
of cardiovascular risk factors in coronary heart disease
associated with type 1 (insulin-dependent) diabetes mellitus.
Diabetes Res Clin Pract 1992; 18:173–84.
Mendall MA, Patel P, Ballam L, Strachan D, Northfield TC.
C reactive protein and its relation to cardiovascular risk
factors: a population based cross sectional study. Br Med J
1996; 312:1061–5.
Effect of intensive diabetes management on macrovascular
events and risk factors in the Diabetes Control and
Complications Trial. Am J Cardiol 1995; 75:894–903.
Intensive blood-glucose control with sulphonylureas or
insulin compared with conventional treatment and risk of
complications in patients with type 2 diabetes (UKPDS 33).
UK Prospective Diabetes Study (UKPDS) Group. Lancet
1998; 352:837–53.
Adler AI, Levy J, Stevens R, Matthews D, Holman RR,
Turner RC. Association between insulin resistance at
diagnosis of diabetes and macrovascular complications—
the UKPDS. Diabetologia 1999; 42:A187.
Tracy RP, Psaty BM, Macy E, Bovill EG, Cushman M,
Cornell ES, Kuller LH. Lifetime smoking exposure affects the
association of C-reactive protein with cardiovascular disease
risk factors and subclinical disease in healthy elderly
subjects. Arterioscler Thromb Vasc Biol 1997; 17:2167–76.
Lip GYH, Beevers DG. Can we treat coronary artery disease
with antibiotics? Lancet 1997; 350:378–9.
Danesh J, Koreth J, Youngman L, Collins R, Arnold JR,
Balarajan Y, McGee J, Roskell D. Is helicobacter pylori a
factor in coronary atherosclerosis? J Clin Microbiol 1991;
37:1651.
Downloaded from by guest on November 24, 2014
Of clinical relevance to diabetic patients, studies
in non-diabetic subjects have shown that aspirin and
statin treatment specifically reduce the increased
cardiovascular risk associated with high CRP
values.7,9 If applied to the asymptomatic diabetic
patients in our study, it would suggest that it is the
older, female, poorly controlled, overweight patient
with a bad family history of heart disease who is
most likely to benefit from such treatment, since they
tend to have the highest CRP levels.
In summary, this study has found that factors
already known to be associated with cardiovascular
risk in type 1 diabetes are also indicative of raised
CRP concentrations. However, the mechanisms
involved in these relationships require further investigation.
(1.03–1.12)
(1.07–1.30)
(1.02–1.18)
(1.07–1.97)
(1.04–1.84)
236
E.S. Kilpatrick et al.
21. Sheldon J, Riches P, Gooding R, Soni N, Hobbs JR.
C-reactive protein and its cytokine mediators in intensivecare patients. Clin Chem 1993; 39:147–50.
22. Hotamisligil GS, Spiegelman BM. Tumor necrosis factor-a: a
key component of the obesity-diabetes link. Diabetes 1994;
43:1271–8.
23. Hotamisligil GS, Arner P, Caro JF, Atkinson RL, Spiegelman
BM. Increased adipose tissue expression of tumor necrosis
factor-a in human obesity and insulin resistance. J Clin
Invest 1995; 95:2409–15.
24. Lamon-Fava S, Wilson PW, Schaefer EJ. Impact of body
mass index on coronary heart disease risk factors in men
and women. The Framingham Offspring Study. Arterioscler
Thromb Vasc Biol 1996; 16:1509–15.
25. Gallagher D, Visser M, Sepulveda D, Pierson RN, Harris T,
Heymsfield SB. How useful is body mass index for
comparison of body fatness across age, sex, and ethnic
groups? Am J Epidemiol 1996; 143:228–39.
Downloaded from by guest on November 24, 2014