High cardiovascular risk in patients with Type 2 diabetic nephropathy: the predictive role of albuminuria and glomerular filtration rate. The NID-2 Prospective Cohort Study.

Abstract

In Type 2 diabetic patients, clinical diagnosis of diabetic nephropathy (DN) is generally based on the concomitant presence of abnormal albuminuria and severe retinopathy. In this high-risk population, cardiovascular (CV) outcome has never been evaluated.
A cohort of 742 Type 2 diabetic patients with DN from 17 national centres was selected by the presence of persistent albuminuria ≥ 30 mg/day and severe diabetic retinopathy and was followed prospectively. Time to CV event (CV death, non-fatal myocardial infarction, non-fatal stroke, revascularization, major amputation) was the primary composite end point and it was analysed by multivariable Cox's proportional hazards model. The interaction between albuminuria and glomerular filtration rate (GFR) was specifically investigated.
Median follow-up was 4.6 years. Overall 242 events (26% of which fatal) were observed in 202 patients. The proportion of CV events increased from 19 to 40% as GFR declined from the highest (≥ 90 mL/min/1.73 m(2)) to the lowest (<45 mL/min/1.73 m(2)) category and was equal to 25 and 33% in microalbuminuria and macroalbuminuria, respectively. In multivariable analysis, the interaction between albuminuria and GFR was statistically significant (P = 0.012). Albuminuria, indeed, had a remarkable prognostic effect in subjects with high GFR that virtually disappeared as GFR became <30 mL/min/1.73 m(2). Age, smoking habit, previous occurrence of myocardial infarction or stroke and proliferative retinopathy were all found to have a statistically significant prognostic effect on CV outcome.
A clinically based diagnosis of DN in Type 2 diabetes allows the identification of subjects with high CV risk. Albuminuria has a relevant prognostic effect on CV morbidity and mortality; its effect is especially pronounced when GFR is normal or near normal.

Full text

Page 1

Nephrol Dial Transplant (2012) 27: 2269–2274
doi: 10.1093/ndt/gfr644
Advance Access publication 16 November 2011
High cardiovascular risk in patients with Type 2 diabetic nephropathy:
the predictive role of albuminuria and glomerular filtration rate.
The NID-2 Prospective Cohort Study
Ferdinando C. Sasso1,2, Paolo Chiodini3, Ornella Carbonara1,2, Luca De Nicola4, Giuseppe Conte4,
Teresa Salvatore1,2, Rodolfo Nasti1,2, Raffaele Marfella2, Ciro Gallo3, Simona Signoriello3,
Roberto Torella1,2and Roberto Minutolo4; on behalf of the NID-2 (Nephropathy In Type 2 Diabetes)
Study Group
1Department of Internal and Experimental Medicine, Second University of Naples, Naples, Italy,2Excellence Centre for Cardiovascular
Disease, Second University of Naples, Naples, Italy,3Department of Biostatistics, Second University of Naples, Naples, Italy and
4Department of Nephrology, Second University of Naples, Naples, Italy
Correspondence and offprint requests to: Ferdinando C. Sasso; E-mail: ferdinando.sasso@unina2.it
Abstract
Background. In Type 2 diabetic patients, clinical diagno-
sis of diabetic nephropathy (DN) is generally based on the
concomitant presence of abnormal albuminuria and severe
retinopathy. In this high-risk population, cardiovascular
(CV) outcome has never been evaluated.
Methods. A cohort of 742 Type 2 diabetic patients
with DN from 17 national centres was selected by the
presence of persistent albuminuria ?30 mg/day and severe
diabetic retinopathy and was followed prospectively. Time
to CV event (CV death, non-fatal myocardial infarction,
non-fatal stroke, revascularization, major amputation) was
the primary composite end point and it was analysed by
multivariable Cox’s proportional hazards model. The inter-
action between albuminuria and glomerular filtration rate
(GFR) was specifically investigated.
Results. Median follow-up was 4.6 years. Overall 242
events (26% of which fatal) were observed in 202 pa-
tients. The proportion of CV events increased from 19
to 40% as GFR declined from the highest (?90 mL/
min/1.73m2) to the lowest (<45 mL/min/1.73m2) cate-
gory and was equal to 25 and 33% in microalbuminuria
and macroalbuminuria, respectively. In multivariable
analysis, the interaction between albuminuria and GFR
was statistically significant (P ¼ 0.012). Albuminuria, in-
deed, had a remarkable prognostic effect in subjects with
high GFR that virtually disappeared as GFR became <30
mL/min/1.73m2. Age, smoking habit, previous occur-
rence of myocardial infarction or stroke and proliferative
retinopathy were all found to have a statistically signifi-
cant prognostic effect on CV outcome.
Conclusions. A clinically based diagnosis of DN in Type 2
diabetes allows the identification of subjects with high CV
risk. Albuminuria has a relevant prognostic effect on CV
morbidity and mortality; its effect is especially pronounced
when GFR is normal or near normal.
2269
? The Author 2011. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
For Permissions, please e-mail: journals.permissions@oup.com
at HINARI Afghanistan Administrative Account on November 11, 2012
http://ndt.oxfordjournals.org/
Downloaded from

Page 2

Keywords: albuminuria; cardiovascular; diabetic nephropathy; GFR;
proliferative retinopathy
Introduction
Diabetic nephropathy (DN) is defined by severe retinop-
athy and albuminuria ?30 mg/day and occurs in a minority
of Type 2 diabetic patients [1]. DN is different from renal
disease occurring in diabetic patients and shows a more
rapid decline of renal function than nephropathy due to
other causes [2]. However, studies on cardiovascular
(CV) outcome in Type 2 DN are still lacking.
Although both albuminuria and glomerular filtration rate
(GFR) are believed to be risk factors for CV events, there
are limited data as to whether these two factors are associ-
ated with adverse outcomes independent not only of other
known CV risk factors but also of each other in patients
with Type 2 diabetes. Recently, the ADVANCE study [3]
showed that high albuminuria and low GFR are independ-
ent risk factors for CV events among patients with Type 2
diabetes. However, the two main clinical features defining
DN, that is diabetic retinopathy and albuminuria, were de-
tected in only 7.1 and 29.3% of patients, respectively [4].
More important, 62% of patients with a GFR <60 mL/min
did not have concurrent albuminuria [3]. Overall, these data
suggest that the ADVANCE study does not provide infor-
mation on the outcome of ‘true’ DN.
The Nephropathy In Diabetes-Type 2 (NID-2) study [5]
was originally designed to investigate the prevalence of CV
risk factors, their management and the achievement of in-
ternational guideline targets in a large population of Type 2
diabetic patients with a clinical diagnosis of DN (concom-
itance of albuminuria and severe diabetic retinopathy), fol-
lowed up in the tertiary care setting. The cross-sectional
phase of the NID-2 study pointed out that patients with DN
are characterized by clusters of risk factors, not at target,
compatible with a high CV risk profile [5]. Moreover, the
NID-2 study showed that frank chronic kidney disease
(CKD) (GFR < 60 mL/min) was present in 38 and 62%
of micro- and macroalbuminuric DN patients, respectively.
The aim of this prospective study was to evaluate CV
prognosis and determinants in the NID-2 cohort of 742 Type
2 diabetic patients, selected by the presence of abnormal al-
buminuria and severe diabetic retinopathy. In particular, we
investigated how renal risk factors, albuminuria and GFR
affected CV outcome in this sub-group of diabetic patients.
Materials and methods
The cross-sectional phase of the NID-2 study evaluated diabetic patients at
17 national centres during a 6-month period from November 2002 to May
2003. Inclusion criteria were: Type 2 diabetes mellitus, age ?40 years,
therapy with diet and/or oral hypoglycaemic agents during the first year of
the diagnosis of diabetes, persistent albuminuria ?30 mg/day in at least
two of three recordings in the last 6 months, a severe diabetic retinopathy
as judged by means of fundus oculi and fluorangiography when necessary.
Severe DR was defined as a proliferative diabetic retinopathy or a severe
non-proliferative diabetic retinopathy, with vascular closure. This last cri-
terion was chosen as the clinical hallmark of DN [1, 2, 6], thus excluding
other possible causes for increased albuminuria since renal biopsies are not
indicated for diagnostic purposes in microalbuminuric diabetic patients.
Exclusion criteria were prior dialysis or renal transplant, diagnosis of
diabetes at <30 years of age, insulin therapy during the first year of
diagnosis of the disease (in order to exclude unknown autoimmune dia-
betes of adults), severe liver or heart failure and known neoplastic or
psychiatric disease. All the participant physicians declared that they ad-
hered to recommendations of the clinical practice guidelines issued by the
American Diabetes Association [7]. Ethical committees approved the pro-
spective phase of the study and all the patients signed informed consent.
Active follow-up, with control visits planned every 6 months, was
completed on 30 November 2009. Baseline information included past
medical history, with particular reference to major CV events (myocardial
infarction and stroke), blood pressure (BP) measurement (calculated as a
mean of three measurements taken in a sitting position after 10 min of
rest), height and body weight as well as laboratory and therapeutic fea-
tures. Laboratory tests were performed locally and included glycaemic,
lipidic and renal function assessment. GFR was calculated by the four-
variable Modification of Diet in Renal Disease equation and albuminuria
was measured on 24-h urine collection; microalbuminuria and macroalbu-
minuria were defined by values of 30–300 and >300 mg/day, respectively.
The primary composite end point was time to CV events, defined as time
frombasalvisittoCVdeath,non-fatalmyocardialinfarction,non-fatalstroke,
revascularization or major amputation, whichever occurred first. When a CV
event was suspected, hospital records were collected to make the diagnosis
according to European Society of Cardiology and American College of Car-
diology criteria [8, 9]. Death certificates and autopsy reports were used to
establish the underlying cause of death and to adjudicate CV deaths, through
the ninth revision of the International Classification of Diseases.
Statistical analysis
For descriptive purpose, patients were categorized into four groups accord-
ingtoGFRvalues(GFR? 60and<60mL/min/1.73m2)andthepresenceof
micro- or macroalbuminuria. A regression model was used to assess the
main effects of the two factors (GFR categories and micro- and macroalbu-
minuria) and their interaction effect. A multiple linear regression model was
used for continuous dependent variables, while a logistic regression model
was used for categorical ones. Median follow-up was estimated by inverse
Kaplan–Meier curve. Multivariable Cox proportional hazards model, strati-
fied by centre, was used to estimate hazard ratio (HR) and corresponding
95% confidence intervals (CIs).Covariatesincluded predefined baselinerisk
factors [age, gender, body mass index (BMI), smoking, previous cardiovas-
cular disease, HbA1c, cholesterol, systolic BP and proliferative retinopathy]
and interaction between albuminuria and GFR. Interaction between either
albuminuria or GFR and covariates included in the model was also tested.
For each variable, restricted cubic splines were used to take into account the
non-linear association with endpoint and were tested by means of likelihood
ratio test. All statistical tests were two tailed and P < 0.05 was considered
significant. Statistical analyses were performed using SPSS version 12.0
(SPSS Inc, Chicago, IL) and R version 2.9.2 (R Foundation for Statistical
Computing, Vienna, Austria) software packages.
Results
From the original cohort of 847 patients, previously de-
scribed [5], three centres following 72 patients refused to
participate in the follow-up study and 33 patients were lost
to follow-up. Therefore, 742 (95.7%) patients were in-
cluded in the present study. Demographic and clinical char-
acteristics of patients lost to follow-up did not differ from
the study subjects. Baseline characteristics are reported in
Table 1. One-third of the patients (n ¼ 247) was aged >70
years, 287 (38.7%) were obese (BMI > 30 kg/m2) and 323
(43.5%) were overweight (BMI 25–30 kg/m2). Diagnosis
of diabetic neuropathy was reported in 309/660 (46.8%)
patients and autonomic neuropathy was diagnosed in
128/634 (20.2%) patients. BP target (<130/80 mmHg)
was attained in 137 patients (18.5%) with a 2-fold greater
prevalence of diastolic BP target (55.0%) than systolic
target (28.0%). Oral hypoglycaemic agents alone and in-
sulin alone were used by 57.8 and 19.3% of patients,
2270F.C. Sasso et al.
at HINARI Afghanistan Administrative Account on November 11, 2012
http://ndt.oxfordjournals.org/
Downloaded from

Page 3

respectively, and in combination by 7.0% of patients. Diet
was the only treatment for 15.9% of the study subjects.
In our cohort, albuminuria was 100 mg/day [interquartile
range (IQR) 54–222]. Microalbuminuria (30–300 mg/day)
was found in 603 (81.3%) patients, (median value 80 mg/
day, IQR 48–127) and macroalbuminuria (>300 mg/day) in
18.7% (median value 385 mg/day, IQR 350–516); among
macroalbuminuric patients, 7.2% had an albuminuria
>1000 mg/day. Mean GFR was 66 ? 24 mL/min/1.73m2
with a prevalence of CKD from Stages 1 to 5 being 16.2,
43.5, 34.4, 5.1 and 0.8%, respectively.
Demographic and clinical characteristics of patients are
reported in Table 1. No difference among sub-groups was
detected for BMI and dyslipidaemia. Patients with GFR
<60 mL/min/1.73m2(N ¼ 299, 40.3%) were older, pre-
dominantly females and with more frequent history of CV
disease and proliferative retinopathy (Table 1). They also
had a slightly higher systolic BP and consequently, these
patients received more anti-hypertensive drugs. As ex-
pected, in patients with lower GFR, prescription of insulin
was higher and that of oral hypoglycaemic agents was
lower (Table 1). In comparison with microalbuminuric pa-
tients, those with macroalbuminuria were characterized by
a larger prevalence of male gender, greater renal impair-
ment (57.8 ? 26.2 versus 68.2 ? 22.8 mL/min/1.73m2, P <
0.0001) and higher systolic BP, with a consequent higher
number of anti-hypertensive drugs. HbA1c and haemoglo-
bin values were influenced by both GFR and albuminuria.
Nephrology consultation was requested in 137 patients
(18.5%). This occurred more frequently in the sub-group
with low GFR and in macroalbuminuric patients. Specifi-
cally, nephrology consultation rate was 9.5, 24.1, 19.2 and
45.2% in patients with microalbuminuria and GFR ?60
mL/min/1.73m2, microalbuminuria and GFR <60 mL/
min/1.73m2, macroalbuminuria and GFR ?60 mL/min/
1.73m2and macroalbuminuria and GFR <60 mL/min/
1.73m2, respectively.
Median follow-up was equal to 4.6 years (IQR 3.2–6.3).
Overall, 242 CV events occurred in 202 (27.2%) patients
(CV death n ¼ 64, non-fatal myocardial infarction n ¼ 80,
non-fatal stroke n ¼ 51, revascularization n ¼ 43, amputa-
tion n ¼ 4). Eight patients died of non-CV cause and seven
patients progressed to end-stage renal disease. Incidence
rate of CV events occurring during the study are reported
in Table 2. At multivariable Cox regression analysis, taking
microalbuminuria and normal GFR as the reference, HR
was 1.43 (95% CI 1.00–2.04) for microalbuminuria and
GFR <60 mL/min/1.73m2, 1.83 (95% CI 1.04–3.02) for
macroalbuminuria and GFR ?60 mL/min/1.73m2and 1.72
(95% CI 1.08–2.76) for macroalbuminuria and GFR <60
mL/min/1.73m2. The final multivariable Cox regression
analysis was built replacing the four groups with GFR
and albuminuria as continuous variables (Table 3 and Fig-
ure 1). Age, history of CV disease, smoking habit, prolif-
erative retinopathy, independently increased the risk of CV
events. As testified by restricted cubic spline analysis,
Table 1. Demographic and clinical characteristics in patients stratified for albuminuria and GFR (expressed as mL/min/1.73m2)a
Overall
(n ¼ 742)
MicroalbuminuriaMacroalbuminuria P-valueb
GFR ? 60
(n ¼ 384)
GFR < 60
(n ¼ 219)
GFR ? 60
(n ¼ 59)
GFR < 60
(n ¼ 80)
GFR < 60
versus
GFR ? 60
Micro
versus
MacroInteractionc
Age (years)
Male gender (%)
BMI (kg/m2)
Smokers (%)
CVD (%)
Proliferative DR (%)
HbA1c (%)
Cholesterol (mg/dL)
HDL (mg/dL)
LDL (mg/dL)
Triglycerides (mg/dL)
Haemoglobin (g/dL)
GFR (mL/min/1.73m2)
UAlb (mg/day)
Systolic BP (mmHg)
Diastolic BP (mmHg)
Anti-hypertensive drugs
CEI and/or ARB (%)
OHA (%)
Insulin (%)
Statins (%)
Aspirin (%)
65.8 6 8.9
46.6
29.3 6 4.9
23.3
23.3
70.1
7.53 6 1.24
197 6 40
48 6 11
119 6 35
152 6 79
13.2 6 1.4
66 6 24
100 (54–222)
136 6 13
78 6 7
1.8 6 1.1
73.5
64.8
26.3
32.9
44.9
64.2 6 9.0
52.1
29.2 6 4.6
27.9
17.9
75.5
7.58 6 1.26
196 6 37
49 6 11
119 6 33
145 6 75
13.5 6 1.2
82 6 16
71 (42–113)
135 6 13
78 6 7
1.5 6 1.0
68
71.6
19.8
27.9
40.9
68.2 6 8.1
34.2
29.4 6 5.2
14.6
29.2
64.4
7.33 6 1.13
195 6 39
47 6 11
117 6 35
156 6 74
12.9 6 1.3
45 6 11
98 (58-163)
138 6 13
78 6 7
2.1 6 1.1
81.3
58.4
30.6
39.7
48.9
63.9 6 9.3
62.7
30.0 6 5.2
28.8
18.6
69.5
7.90 6 1.34
201 6 44
47 6 14
122 6 41
167 6 112
13.6 6 1.6
84 6 15
384 (348–535)
138 6 13
80 6 7
1.7 6 1.1
72.9
72.9
27.1
30.5
52.5
67.5 6 8.8
42.5
29.1 6 5.4
21.3
40
60
7.51 6 1.30
201 6 48
48 6 11
121 6 38
165 6 75
12.3 6 1.7
39 6 12
391 (350–500)
140 6 14
78 6 8
2.3 6 1.2
78.8
43.8
45
40
47.5
<0.0001
<0.0001
0.437
0.007
<0.0001
0.019
0.007
0.907
0.552
0.627
0.584
<0.0001
0.543
0.045
0.613
0.270
0.201
0.229
0.036
0.137
0.744
0.238
0.041
0.038
0.828
0.826
0.253
0.369
0.400
0.775
0.537
0.772
0.272
0.955
0.390
0.011
0.044
0.068
<0.0001
0.021
<0.0001
0.001
0.019
0.752
0.010
0.104
0.008
0.862
0.197
0.014
0.731
0.278
0.654
0.049
0.856
0.381
0.110
0.619
0.774
0.171
aValues are mean 6 SD, median (IQR) or percent. CVD, history of myocardial infarctionor stroke; GFR, estimated glomerularfiltration rate; HDL, high-
density lipoprotein; LDL, low-density lipoprotein; DR, diabetic retinopathy; UAlb, urinary albumin excretion; CEI, converting enzyme inhibitors; ARB,
angiotensin II receptor blockers; OHA, oral hypoglycaemic agents.
bP-values were calculated from multiple linear regression model and logistic regression model for continuous and categorical-dependent variables,
respectively.
cGFR < 60 versus GFR ? 60, macroalbuminuria versus microalbuminuria and their interaction were used as covariates in each model.
CV outcome in DN2271
at HINARI Afghanistan Administrative Account on November 11, 2012
http://ndt.oxfordjournals.org/
Downloaded from

Page 4

association of continuous variables with risk outcome was
always linear (data not shown). Interaction between GFR
and albuminuria was statistically significant (P ¼ 0.012),
suggesting that either effect varied as the other one was
changing. Conversely, no interaction was detected between
either GFR or albuminuria and other covariates included in
the model. The combined HRs from Cox regression model
and for specific values of albuminuria and GFR are re-
ported in Figure 1; reference values are 90 mL/min/
1.73m2for GFR and 30 mg/day for albuminuria. Albumi-
nuria had a remarkable prognostic effect in subjects with
high GFR that virtually disappeared as GFR became very
low. At the same time, a predictive role for GFR can be
detected only in microalbuminuric patients.
Discussion
This study originally evaluated the impact of risk factors on
CV outcome in a prospective cohort of 742 Type 2 diabetic
patients with persistent albuminuria and severe retinopathy.
In particular, we investigated the prognostic role of the two
main renal risk factors (albuminuria and GFR).
It is well established that diabetic subjects with associ-
ated nephropathy present an increased CV risk and that
most of them do not develop end-stage renal disease be-
cause of prior fatal CV events [10]. All the previous inves-
tigations, however, examined patients with ‘unspecific’
microalbuminuria or macroalbuminuria [3, 10, 11]. This
is a critical point of difference with our study because the
absence of retinopathy constitutes the strongest clinical
evidence to consider albuminuria as a non-diabetic lesion
by renal histological analysis, with very high sensitivity
(87%) and specificity (93%) [12]. On the other hand, in
clinical practice, kidney biopsy, which is the ideal diagnos-
tic tool to discern between true and false DN, is hardly
feasible for ethical reasons in patients without significant
proteinuria. Therefore, the NID-2 study investigated a spe-
cific sub-group of diabetic patients never studied before,
but easily identifiable.
The first result to be underlined is the high CV risk of the
population studied. Overall, 27.2% of the patients, during
the follow-up(4.6years onmedian),experiencedatleast one
CV event, fatal in about one-third of cases. The incidence
rates are definitelyhigherthanthose observedinstudies with
similar duration of follow-up conducted in ‘generic’ Type 2
diabetic populations [13]. Notably, CV event rate is even
greater than that observed in the diabetic microalbuminuric
group of HOPE Study and MICROHOPE sub-study [14].
Age, history of CV disease, smoking habit, proliferative
retinopathy, albuminuria and GFR independently increased
the risk of CV events. Among the non-renal risk factors, the
significant prognostic role of type of diabetic retinopathy is
of great interest, confirming recent observations [15, 16].
Indeed, in an 18-year follow-up study [15], proliferative
retinopathy predicted all-cause, CV and coronary death in
Type 2 diabetic subjects. These associations were inde-
pendent of current smoking, hypertension, total cholesterol,
high-density lipoprotein cholesterol, glycaemic control, du-
ration of diabetes and proteinuria. In a sub-analysis of the
VADT study [16], a relevant relationship was observed
between retinopathy and coronary atherosclerosis quanti-
fied by means of computed tomography-detectablecoronary
artery calcium. Individuals with proliferative retinopathy
were ~6-fold more likely to have high coronary artery
Table 2. Incidence rate of CV events occurring during the study overall and in sub-groups
Overall
(n ¼ 742)
MicroalbuminuriaMacroalbuminuria
GFR ? 60
(n ¼ 384)
GFR < 60
(n ¼ 219)
GFR ? 60
(n ¼ 59)
GFR < 60
(n ¼ 80)
CV death
Number of events
Event rate (100 patient-years)
Non-fatal CV event
Number of events
Event rate (100 patient-years)
Fatal/non-fatal events
Number of events
Event rate (100 patient-years)
64
1.98
22
1.28
26
2.69
4
1.71
12
3.84
150 71
4.51
46
5.26
12
6.00
21
7.615.13
202 88
5.59
68
7.77
15
7.50
31
11.23 6.90
Table 3. HR and 95% CI for fatal and non-fatal CV events by Cox
regression analysisa
b-Coefficient HR95% CIP
Age (years)0.02911.03
1.01–1.05
0.003
Male gender0.14691.16
0.85–1.58
0.352
BMI (kg/m2)
?0.0005
0.4276
1.00
0.97–1.03
0.976
Smokers (yes versus no)1.53
1.08–2.17
0.016
History of MI or stroke
(yes versus no)
Proliferative DR
(yes versus no)
HbA1c (%)
0.84552.33
1.70–3.19
<0.0001
0.45991.58
1.10–2.28
0.013
0.07801.08
0.96–1.22
0.212
Total cholesterol (mg/dL)
?0.0010
0.0067
1.00
1.00–1.00
0.612
Systolic BP (mmHg)1.01
1.00–1.02
0.274
GFR (5 mL/min/1.73m2)b
Albuminuria (g/day)b
Interaction
eGFR 3 albuminuriab
?0.0688
?1.0830
0.1799
0.002
0.165
0.012
aDR, diabetic retinopathy.
bFor HR, see Figure 1.
2272F.C. Sasso et al.
at HINARI Afghanistan Administrative Account on November 11, 2012
http://ndt.oxfordjournals.org/
Downloaded from

Page 5

calcium than those with no proliferative retinopathy, even
after adjustment for the other CV risk factors. Our and the
previous data, therefore, suggest the presence of common
background pathways for diabeticmicro- and macrovascular
disease. In particular, abnormal albuminuria could be a link-
age between these diabetic vascular complications, being a
marker of an early inflammatory state in the atherosclerotic
disease process [17]. In this regard, a recent study has evi-
denced that the risk of developing retinopathy is greater in
the diabetic patients with GFR <60 and macroalbuminuria
with respect to diabetics with GFR <60 without macro-
albuminuria [18].
The main measures of renal damage markedly influenced
the CV outcome of these patients. In particular, proportion
of CV events increased from 19 to 40% as GFR declined
from the highest (?90 mL/min/1.73m2) to the lowest (<45
mL/min/1.73m2) category and was equal to 25 and 33% in
micro- and macroalbuminuria, respectively. In particular,
the most intriguing findings resulted from multivariable
analysis, showing an interaction between albuminuria and
GFR (Figure 1). Albuminuria, indeed, showed a remark-
able prognostic effect in subjects with high GFR that vir-
tually disappeared in the lower strata of GFR. These results
clashwiththeconclusionsofanotherstudy.Inasub-analysis
oftheADVANCEstudy[3],10 640Type2diabeticpatients
with levels of albuminuria and serum creatinine available at
baseline were studied, and it was observed that both in-
creased urinary albumin excretion and reduced GFR are
independently and continuously associated with the risk
for both CV and kidney outcomes in patients with Type 2
diabetes, but there was no evidence of any interaction be-
tween these risk factors.
On this matter, it is interesting to underline the great dif-
ference between NID-2 and ADVANCE studies. The two
studies were similar for age, HbA1c, BMI of diabetic pa-
tients at baseline and for duration of follow-up, but the ma-
jority of patients of ADVANCE was normoalbuminuric
(69%) and only 7% had retinopathy; therefore, it was a
generic diabetic population, with and without nephropathy.
Although patients with a previous CV event at baseline were
32% in ADVANCE versus 23% in NID study, and BP at
baseline was higher in ADVANCE versus NID (145/81
versus 136/78 mmHg) patients, thus suggesting a major
CV risk in ADVANCE group, the incidence rates of total
major CV events and of death from CV causes were, respec-
tively, 3 and 2.5-fold higher in NID group than in overall
ADVANCE diabetic population. Therefore, the different
role for GFR and albumin excretion rate (AER) on CV out-
come observed in the two studies could be the result of two
different diabetic populations.
The explanation for this interaction is not readily appa-
rent. We may speculate that in patients with normal or only
mildly reduced GFR, the prognostic role on CV outcome of
albuminuria is higher because under these conditions, al-
buminuria is a marker of endothelial dysfunction, therefore
heralding as such the increased CV risk [19]. On the other
hand, indirect evidence suggests that in the presence of
more impaired renal function, albuminuria is more related
to either glomerulosclerosis or advanced tubular damage
rather than endothelial dysfunction [20–23].
It is supposable that our results could offer a different,
less exacting, interpretation of the significant positive in-
teraction of GRF and AER. In fact, for the low prevalence
(<6%) of CKD at Stages 4 and 5, both measures could
simply provide significant prognostic information about
the risk of CV events, and their effects could be multipli-
cative. However, our intriguing interpretation of the find-
ings seems to be supported by the current knowledge of CV
risk in Type 2 diabetes.
The main limitation of this study is that the observational
nature precludes a proper cause-effect analysis; however,
results are hypothesis generating to design multifactorial
intervention trials aimed at verifying the possible improve-
ment of CV outcome in patients with true DN. In particular,
our findings suggest an intensive treatment in the early
phases of DN, when the increased albuminuria is coupled
with normal (or slightly decreased) GFR. On the basis of
our results, it is in fact intriguing to hypothesize that at this
stage of nephropathy, an intensive approach to abnormal
albuminuria may reduce CV risk. A potential area of im-
provement, to be verified by ad hoc randomized trials, may
be represented by the dual blockade of renin-angiotensin
system (RAS) in diabetics selected by abnormal albuminu-
ria [24]. This intervention was rare in the NID-2 patients
and in particular in the early phases of DN. In this regard, a
specific role of a dual RAS blockade on cardiorenal risk in
diabetic patients in the early phases of renal damage and
without established CV diseases should be tested.
As above stated, another limitation of the study is the
low prevalence (<6%) of GFR <30 mL/min.
Finally, the strict criteria used to select the population are
also a limitation of the study because its results cannot be
generalized to all diabetic patients. Nevertheless, the very
high CV risk observed in the NID-2 cohort strengthens the
prognostic usefulness of these selection criteria, independ-
ently of histological confirmation of DN. In fact, even if the
co-existence of albuminuria and retinopathy in Type 2 dia-
betes was recently reported to be not always histologically
Fig. 1. Combinedeffectsof albuminuriaand GFRlevels at baselineon the
risk for fatal and non-fatal CV events. The estimates were calculated at
specific values of GFR and albuminuria and adjusted for the covariates
included in the Cox model (Table 2). Reference category GFR 90 mL/min/
1.73m2and albuminuria 30 mg/day.
CV outcome in DN2273
at HINARI Afghanistan Administrative Account on November 11, 2012
http://ndt.oxfordjournals.org/
Downloaded from

Page 6

matched in DN [25], the possibility to clinically identify a
sub-group of patients at very high CV risk represents the
highlight of this study.
In conclusion, patients with Type 2 DN show a CV risk
independentlyrelatedtoage,historyofCVdisease,smoking
habit, proliferative retinopathy, albuminuria and GFR.Albu-
minuria has a relevant prognostic effect on CV morbidity
and mortality; its effect is especially pronounced when GFR
is normal or near normal. These findings strongly suggest an
intensive treatment since the early phases of DN, especially
for albuminuria and smoking cessation, but it has to be
confirmed by an interventional study.
Acknowledgements. This work was partially supported by an Italian Gov-
ernment grant from M.I.U.R. (Ministero della Istruzione, Universita `,
e Ricerca) Rome, Italy. The study sponsor had no role in the collection,
analysis and interpretation of data in the writing of the report and in the
decision to submit the paper for publication.
F.C.S. had full access to all of the data in the study and takes respon-
sibility for the integrity of the data and the accuracy of the data analysis.
Contribution statement. F.C.S. and R.M.: conception and design, inter-
pretation of data, drafting the article and final approval of the manuscript;
P.C., O.C., L.D.N., G.C., S.S. and C.G.: analysis and interpretation of
data, final approval of the version to be published; T.S., R.N., R.M.,
R.T.: revising the manuscript critically for important intellectual content,
final approval of the version to be published.
Conflict of interest statement. None declared.
Appendix
NID-2 Study Group: F.C. Sasso, O. Carbonara, R. Torella,
R. D’Urso, A. Lampitella Jr, N. Lascar, R. Nasti, A. Pagano,
E. Pisa, L. Zirpoli, F. Zibella, M. Corigliano, G. Conte, L. De
Nicola, R. Minutolo, P. Trucillo, V. Armentano, P. Calatola,
G. Corigliano, E. Del Vecchio, N. De Rosa, G. Di Giovanni,
A. De Matteo, O. Egione, A. Gatti, S. Gentile, L. Gesue `,
L. Improta, A. Lampitella, A. Lanzilli, S. Masi, P. Mattei,
V. Mastrilli, P. Memoli, E. Rossi, S. Sorrentino, R. Troise,
A.A. Turco, S. Turco.
References
1. Fioretto P, Mauer M, Brocco E et al. Patterns of renal injury in NIDDM
patients with microalbuminuria. Diabetologia 1996; 39: 1569–1576
2. Wong TY, Choi PC, Szeto CC et al. Renal outcome in type 2 diabetic
patients with or without coexisting nondiabetic nephropathies. Dia-
betes Care 2002; 25: 900–905
3. Ninomiya T, Perkovic V, de Galan BE et al. Albuminuria and kidney
function independently predict cardiovascular and renal outcomes in
diabetes. J Am Soc Nephrol 2009; 20: 1813–1821
4. The ADVANCE Collaborative Group. Intensive blood glucose con-
trol and vascular outcomes in patients with type 2 diabetes. N Engl
J Med 2008; 358: 2560–2572
5. Sasso FC, De Nicola L, Carbonara O et al. Cardiovascular risk factors
and disease management in type 2 diabetic nephropathy. Diabetes
Care 2006; 29: 498–503
6. Kramer HJ, Nguyen QD, Curhan G et al. Renal insufficiency in the
absence of albuminuria and retinopathy among adults with type 2
diabetes mellitus. JAMA 2003; 289: 3273–3277
7. Expert Committee on the Diagnosis and Classification of Diabetes
Mellitus.AmericanDiabetesAssociation: clinicalpractice recommen-
dations. Diabetes Care 2002; 25 (Suppl 1): S1–S147
8. American College of Cardiology Committee. Myocardial infarction
redefined-a consensus document of The Joint European Society of
Cardiology/American College of Cardiology for the redefinition of
myocardial infarction. J Am Coll Cardiol 2000; 36: 959–969
9. Luepker R, Apple FS, Christenson RH et al. Case definitions for acute
coronary heart disease in epidemiology and clinical research studies.
A statement from the AHA Council on Epidemiology and Prevention;
AHA Statistics Committee; World Heart Federation Council on Epi-
demiology and Prevention; the European Society of Cardiology
WorkingGroupon Epidemiology and Prevention; Centersfor Disease
Control and Prevention; and the National Heart, Lung, and Blood
Institute. Circulation 2003; 108: 2543–2549
10. Patel UD, Young EW, Ojo AO et al. CKD progression and mortality
amongolderpatientswithdiabetes.AmJKidneyDis2005;46:406–414
11. Gaede P, Vedel P, Larsen N et al. Multifactorial intervention and
cardiovascular disease in patients with type 2 diabetes. N Engl
J Med 2003; 348: 383–393
12. Tone A, Shikata K, Matsuda M et al. Clinical features of non-diabetic
renaldiseasesin patientswithtype2diabetes.DiabetesRes ClinPract
2005; 69: 237–242
13. Knopp RH, d’Emden M, Smilde JG et al. Efficacy and safety of
atorvastatin in the prevention of cardiovascular end points in subjects
with type 2 diabetes: the Atorvastatin Study for Prevention of Coro-
nary Heart Disease Endpoints in non-insulin-dependent diabetes mel-
litus (ASPEN). Diabetes Care 2006; 29: 1478–1485
14. Gerstein HC, Pogue J, Mann JF et al. for HOPE investigators. The
relationship between dysglycaemia and cardiovascular and renal
risk in diabetic and non-diabetic participants in the HOPE study:
a prospective epidemiological analysis. Diabetologia 2005; 48:
1749–1755
15. Juutilainen A, Lehto S, Ro ¨nnemaa T et al. Retinopathy predicts car-
diovascular mortality in type 2 diabetic men and women. Diabetes
Care 2007; 30: 292–299
16. ReavenPD,EmanueleN,MoritzTetal.Proliferativediabeticretinopathy
in type 2 diabetes is related to coronary artery calcium in the Veterans
Affairs Diabetes Trial (VADT). Diabetes Care 2008; 31: 952–957
17. Tsioufis C, Dimitriadis K, Taxiarchou E et al. Diverse associations of
microalbuminuria with C-reactive protein, interleukin-18 and soluble
CD 40 ligand in male essential hypertensive subjects. Am J Hypertens
2006; 19: 462–466
18. Agarwal R, Debella YT, Giduma HD et al. Long-term retinal, renal
and cardiovascular outcomes in diabetic chronic kidney disease with-
out proteinuria. Nephrol Dial Transplant 2012; 27: 310–317
19. Deckert T, Feldt-Rasmussen B, Borch-Johnsen K et al. Albuminuria
reflects widespread vascular damage. The Steno hypothesis. Diabeto-
logia 1989; 32: 219–226
20. Russo LM, Sandoval RM, Campos SB et al. Impaired tubular uptake
explains albuminuria in early diabetic nephropathy. J Am Soc Nephrol
2009; 20: 489–494
21. Rule AD, Amer H, Cornell LD et al. The association between age and
nephrosclerosis on renal biopsy among healthy adults. Ann Intern
Med 2010; 152: 561–567
22. Ahmed AK, Kamath NS, El Kossi M et al. The impact of stopping
inhibitors of the renin-angiotensin system in patients with advanced
chronic kidney disease. Nephrol Dial Transplant 2010; 25:
3977–3982
23. Comper WD, Hilliard LM, Nikolic-Paterson DJ et al. Disease-
dependent mechanisms of albuminuria. Am J Physiol Renal Physiol
2008; 295: F1589–F1600
24. Jennings DL, Kalus JS, Coleman CI et al. Combination therapy
with an ACE inhibitor and an angiotensin receptor blocker for
diabetic nephropathy: a meta-analysis. Diabet Med 2007; 24:
486–493
25. Lin YL, Peng SJ, Ferng SH et al. Clinical indicators which necessitate
renal biopsy in type 2 diabetes mellitus patients with renal disease. Int
J Clin Pract 2009; 63: 1167–1176
Received for publication: 11.7.11; Accepted in revised form: 3.10.11
2274 F.C. Sasso et al.
at HINARI Afghanistan Administrative Account on November 11, 2012
http://ndt.oxfordjournals.org/
Downloaded from