Gestational diabetes mellitus alone in the absence of subsequent diabetes is associated with microalbuminuria: results from the Kidney Early Evaluation Program (KEEP).

Abstract

Women with gestational diabetes mellitus (GDM) maintain a higher risk for recurrent GDM and overt diabetes. Overt diabetes is a risk factor for development of chronic kidney disease (CKD), but GDM alone, without subsequent development of overt diabetes, may also pose a risk for CKD.
This cross-sectional analysis included Kidney Early Evaluation Program (KEEP) participants from 2000 to 2009. Patient characteristics and kidney function among three categories (GDM alone, overt diabetes, and no history of diabetes) were compared. The prevalence of microalbuminuria, macroalbuminuria, and CKD stages 1-2 and 3-5 was assessed using logistic regression.
Of 37,716 KEEP female participants, 571 (1.5%) had GDM alone and 12,100 (32.1%) had overt diabetes. Women with GDM had a higher rate of microalbuminuria but not macroalbuminuria than their nondiabetic peers (10.0 vs. 7.7%) that was substantially lower than the 13.6% prevalence in diabetic women. In multivariate analysis, women with GDM alone, compared with nondiabetic women, demonstrated increased odds of CKD stages 1-2 (multivariate odds ratio 1.54 [95% CI 1.16-2.05]) similar to the odds for women with overt diabetes (1.68 [1.55-1.82]). In stratified analyses, age, race, BMI, and hypertension modified the odds for CKD stages 1-2 but not CKD stages 3-5 among women with GDM.
Women with GDM alone have a higher prevalence of microalbuminuria than women without any history of diabetes, translating to higher rates of CKD stages 1-2. These results suggest that GDM, even in the absence of subsequent overt diabetes, may increase the risk for future cardiovascular and kidney disease.

Full text

Page 1

Gestational Diabetes Mellitus Alone in the
Absence of Subsequent Diabetes Is
Associated With Microalbuminuria
Results from the Kidney Early Evaluation Program (KEEP)
ANDREW S. BOMBACK, MD, MPH1
YELENA REKHTMAN, MD1
ADAM T. WHALEY-CONNELL, DO, MSPH2
ABHIJIT V. KSHIRSAGAR, MD, MPH3
JAMES R. SOWERS, MD2
SHU-CHENG CHEN, MS4
SUYING LI, PHD4
KAVITHA M. CHINNAIYAN, MD5
GEORGE L. BAKRIS, MD6
PETER A. MCCULLOUGH, MD, MPH5
OBJECTIVE — Women with gestational diabetes mellitus (GDM) maintain a higher risk for
recurrent GDM and overt diabetes. Overt diabetes is a risk factor for development of chronic
kidney disease (CKD), but GDM alone, without subsequent development of overt diabetes, may
also pose a risk for CKD.
RESEARCH DESIGN AND METHODS — This cross-sectional analysis included Kid-
ney Early Evaluation Program (KEEP) participants from 2000 to 2009. Patient characteristics
and kidney function among three categories (GDM alone, overt diabetes, and no history of
diabetes) were compared. The prevalence of microalbuminuria, macroalbuminuria, and CKD
stages 1–2 and 3–5 was assessed using logistic regression.
RESULTS — Of 37,716 KEEP female participants, 571 (1.5%) had GDM alone and 12,100
(32.1%) had overt diabetes. Women with GDM had a higher rate of microalbuminuria but not
macroalbuminuriathantheirnondiabeticpeers(10.0vs.7.7%)thatwassubstantiallylowerthan
the 13.6% prevalence in diabetic women. In multivariate analysis, women with GDM alone,
compared with nondiabetic women, demonstrated increased odds of CKD stages 1–2 (multi-
variate odds ratio 1.54 [95% CI 1.16–2.05]) similar to the odds for women with overt diabetes
(1.68[1.55–1.82]).Instratifiedanalyses,age,race,BMI,andhypertensionmodifiedtheoddsfor
CKD stages 1 –2 but not CKD stages 3–5 among women with GDM.
CONCLUSIONS — WomenwithGDMalonehaveahigherprevalenceofmicroalbuminuria
thanwomenwithoutanyhistoryofdiabetes,translatingtohigherratesofCKDstages1–2.These
resultssuggestthatGDM,evenintheabsenceofsubsequentovertdiabetes,mayincreasetherisk
for future cardiovascular and kidney disease.
Diabetes Care 33:2586–2591, 2010
M
moglycemic after delivery but still main-
tain a higher risk for recurrent GDM,
ost women who develop diabetes
during a pregnancy, gestational
diabetes mellitus (GDM), are nor-
impaired glucose tolerance, and overt di-
abetes. Indeed, the odds of developing
subsequent type 2 diabetes for women
with GDM is roughly 5 times higher than
that for women with normoglycemic
pregnancies in the first 5 years after deliv-
ery; the odds rise to more than 9 times
higher in the years afterward (1).
Althoughovertdiabetesisrecognized
as a potent risk factor for development of
chronic kidney disease (CKD), it is cur-
rently unclear whether GDM alone, with-
out subsequent development of overt
diabetes, also poses any risk to kidney
function. Because certain clinical factors
(e.g.,waistcircumference,BMI,andyears
postdelivery) have been shown to in-
crease the risk for development of overt
diabetes in women with GDM (2), these
factors could potentially also modify the
risk for development of CKD.
We hypothesized that GDM alone
would impart an increased risk for CKD
and, specifically, that women with GDM
would have a level of risk intermediate
between that of women without any his-
tory of glucose abnormalities and women
with overt diabetes. Using data from the
National Kidney Foundation’s Kidney
Early Evaluation Program (KEEP), a pro-
gram designed to screen participants at
higherriskforCKDthanthegeneralpop-
ulation, we examined in cross-sectional
analyses whether GDM, in the absence of
subsequent overt diabetes, increases the
odds of abnormal urinary albumin excre-
tion and impaired glomerular filtration
rate. In addition, we examined whether
age, race, BMI, or hypertension modifies
thisrelationshipbetweenGDMandCKD.
RESEARCH DESIGN AND
METHODS
KEEP and participants
KEEP is a free, community-based health
screening program that targets popula-
tions aged ?18 years at high risk for kid-
ney disease, defined as a history of
diabetes or hypertension or a first-order
relative with diabetes, hypertension, or
kidney disease (3). The screening in-
cludes informed consent, health screen-
ing questionnaire, diagnostic panel, and
physician consultation. The study popu-
lationincludedeligibleKEEPparticipants
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
Fromthe1DepartmentofMedicine,DivisionofNephrology,ColumbiaUniversityCollegeofPhysiciansand
Surgeons, New York, New York; the2University of Missouri-Columbia School of Medicine and Harry S.
Truman VA Medical Center, Columbia, Missouri; the3Department of Medicine, Division of Nephrology
and Hypertension, University of North Carolina School of Medicine, Chapel Hill, North Carolina; the
4Chronic Disease Research Group, Minneapolis Medical Research Foundation, Minneapolis, Minnesota;
the5Department of Medicine, Divisions of Cardiology, Nutrition, and Preventive Medicine, William
BeaumontHospital,RoyalOak,Michigan;andthe6DepartmentofMedicine,HypertensiveDiseasesUnit,
University of Chicago, Pritzker School of Medicine, Chicago, Illinois.
Corresponding author: Andrew S. Bomback, asb68@columbia.edu.
Received8June2010andaccepted23August2010.Publishedaheadofprintathttp://care.diabetesjournals.
org on 31 August 2010. DOI: 10.2337/dc10-1095.
© 2010 by the American Diabetes Association. Readers may use this article as long as the work is properly
cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.
org/licenses/by-nc-nd/3.0/ for details.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby
marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
E p i d e m i o l o g y / H e a l t h S e r v i c e s R e s e a r c h
O R I G I N A LA R T I C L E
2586
DIABETES CARE, VOLUME 33, NUMBER 12, DECEMBER 2010care.diabetesjournals.org

Page 2

from August 2000 to May 2009 from 47
NationalKidneyFoundationaffiliatesand
2,336screeningprogramsin50statesand
the District of Columbia. This study cohort
included only eligible KEEP female partici-
pantsdividedintothreegroups:withGDM
alone, with overt diabetes, and without di-
abetes. All participants were screened after
their pregnancies, and details of their preg-
nancies were self-reported.
Definitions
Diabetes was defined as a fasting blood
glucose level ?126 mg/dl, nonfasting
blood glucose level ?200 mg/dl, self-
reported history of diabetes, or taking
glucose-lowering medications. GDM
alone was defined as self-reported diabe-
tes only during pregnancy without subse-
quent development of overt diabetes.
Estimated glomerular filtration rate
(eGFR) was calculated from serum creat-
inine measured at the KEEP screening
using the reexpressed four-variable Mod-
ification of Diet in Renal Disease Study
equation(serumcreatininewascalibrated
by the Cleveland Clinic Research Labora-
tory). Albumin-to-creatinine ratios
(ACRs) were calculated from spot urine
samples collected at screening and re-
corded as ?30, 30–300, or ?300 mg/g.
Microalbuminuria was defined as ACR
30–300 mg/g and macroalbuminuria as
ACR ?300 mg/g. CKD stages were de-
fined as follows: no CKD, eGFR ?60 ml/
min per 1.73 m2 and no proteinuria;
stage 1, eGFR ?90 ml/min per 1.73 m2
with ACR ?30 mg/g; stage 2, eGFR
60–89 ml/min per 1.73 m2with ACR ?
30mg/g;stage3,eGFR30–59ml/minper
1.73 m2; and stage 4–5, eGFR ?30 ml/
min per 1.73 m2.
Hypertension was defined as average
systolic blood pressure ?129 mmHg or
diastolicbloodpressure?84mmHg,self-
reported history of hypertension, or tak-
ing blood pressure–lowering medication.
Dyslipidemia was defined as triglycerides
?150 mg/dl or HDL level ?50 mg/dl.
Other measures, including tobacco and
alcohol use and family history of diseases
and hypertension, were self-reported.
Blood pressure, height, weight, and waist
circumference were taken from direct
measurements for all participants.
Statistical analysis
Patient characteristics and assessment of
kidney function among study groups
(GDM, overt diabetes, and no diabetes)
were compared and tested by ANOVA.
Associations of GDM and overt diabetes
with respect to microalbuminuria, mac-
roalbuminuria, CKD stages 1–2, and
CKD stages 3–5 among KEEP female par-
ticipantswereassessedusingunivariateand
multivariate logistic regressions, respec-
tively. Adjusted odds ratios (ORs) for CKD
stages 1–2 and CKD stages 3–5 among
KEEP female participants were stratified by
age,race,hypertensivestatus,andBMI.Co-
variates in multivariate models included
age, race, BMI, current smoking, alcohol
use,hypertension,dyslipidemia,andfamily
history of kidney disease.
RESULTS— A total of 37,716 KEEP
female participants provided information
on glycemic status and were therefore el-
igible for analysis, of which 571 (1.5%)
had GDM alone and 12,100 (32.1%) had
overt diabetes (Table 1). Women who re-
ported GDM alone were significantly
younger than those with overt diabetes
andthosewithoutanyhistoryofdiabetes.
Women with GDM and women without
diabetes had virtually identical BMI and
waist circumference measurements that
were significantly lower than those of
women with overt diabetes. Likewise,
hypertension and dyslipidemia were sub-
stantially more prevalent among women
with overt diabetes compared with both
women with GDM alone and women
without any history of diabetes. At the
time of the screening visit, fasting blood
glucose values were essentially equal for
women with GDM and women without
diabetes (96.4 ? 13.0 vs. 95.0 ? 13.4
mg/dl)andsignificantlylowerthanvalues
for women with overt diabetes (135.3 ?
60.3 mg/dl).
Table 1—Characteristics of KEEP female participants with GDM alone, with overt diabetes, and without diabetes
CharacteristicsGDM Overt diabetesNo diabetes
P value
n
Age (years)
Race (%)
White
African American
Other
BMI (kg/m2)
Waist circumference (inches)*
Hypertension (%)
Systolic blood pressure (mmHg)
Diastolic blood pressure (mmHg)
Dyslipidemia (%)
Total cholesterol (mg/dl)
Triglycerides (mg/dl)
Current tobacco use (%)
Alcohol use (%)
Family history of kidney disease (%)
Family history of diabetes (%)
Family history of hypertension (%)
Fasting blood glucose (mg/dl)†
Data are means ? SD or %. *Data on waist circumference are available from 2008 (n ? 7,070). †Calculation including only participants with fasting glucose data
available (n ? 9,071).
57112,100
59.0 ? 13.5
25,045
53.4 ? 15.451.5 ? 14.5
?0.0001
0.0003
46.8
29.9
23.3
51.3
32.1
16.6
50.6
32.9
16.5
30.0 ? 7.2
36.6 ? 5.4
64.6
128.7 ? 19.9
79.0 ? 11.8
39.2
203.0 ? 36.6
149.3 ? 99.9
11.0
56.2
15.6
51.5
86.0
96.4 ? 13.0
32.6 ? 7.5
41.3 ? 6.7
84.2
134.7 ? 19.8
78.3 ? 11.5
48.9
196.3 ? 42.7
172.3 ? 121.4
8.2
46.8
18.3
67.9
78.4
135.3 ? 60.3
29.5 ? 6.9
37.5 ? 6.1
67.9
130.0 ? 19.3
78.9 ? 11.3
34.1
204.2 ? 39.8
137.8 ? 91.6
9.3
57.5
18.0
49.7
85.6
95.0 ? 13.4
?0.0001
?0.0001
?0.0001
?0.0001
?0.0001
?0.0001
?0.0001
?0.0001
0.0004
?0.0001
0.2,487
?0.0001
?0.0001
?0.0001
Bomback and Associates
care.diabetesjournals.orgDIABETES CARE, VOLUME 33, NUMBER 12, DECEMBER 2010
2587

Page 3

Women who reported GDM alone
had lower serum creatinine and higher
eGFR measurements than women with or
without any history of diabetes (Table 2).
Less than 12% of women with GDM had
eGFR ?60 ml/min per 1.73 m2), com-
pared with ?14% of women without any
history of diabetes and 21% of women
with overt diabetes. These results trans-
lated to significantly lower prevalence
rates of CKD stages 3–5 for women with
GDM alone compared with women with
or without diabetes. Although women
with GDM and nondiabetic women had
similarly low rates of macroalbuminuria
(urinary ACR ?300 mg/g), women with
GDM had significantly higher rates of mi-
croalbuminuria (urinary ACR 30–300
mg/g) than their nondiabetic peers (10.0
vs, 7.7%) that were still substantially
lower than the 13.6% prevalence rate
seen in diabetic women.
Inmultivariatelogisticregression,the
OR for microalbuminuria among women
with GDM was 1.36 [95% CI 1.03–1.80]
and among women with overt diabetes
was 1.67 [1.55–1.80] using nondiabetic
women as the reference group (Table 3).
These results translated to increased odds
of CKD stages 1–2 for women with GDM
alone (multivariate OR 1.54 [95% CI
1.16–2.05])thatapproachedtheoddsfor
women with overt diabetes (1.68 [1.55–
1.82]). In univariate and multivariate
analyses, women with GDM alone and
women without any history of diabetes
had similarly lower odds for macroalbu-
minuria and CKD stages 3–5 than dia-
betic women.
In stratified analyses, age, race, BMI,
and the presence of hypertension modi-
fied the odds for CKD stages 1–2, but not
for CKD stages 3–5, among women with
GDM. These results imply that women
with GDM were at highest risk for abnor-
mal urinary albumin excretion if they
were younger, African American, obese,
and/or hypertensive (Table 4). For exam-
ple,whereaswhitewomenwithGDMdid
notappeartohaveincreasedoddsofCKD
stages 1–2 compared with white women
without any history of diabetes, African
American women with GDM had more
than twice the odds for CKD stages 1–2
(multivariate OR 2.32 [95% CI 1.50–
3.60]) than African American women
without diabetes. Furthermore, the pres-
ence of obesity (defined as BMI ?30
kg/m2)andhypertensionimpartedessen-
tially identical odds for CKD stages 1–2
among women with GDM and women
Table 2—Assessment of kidney function in KEEP female participants with GDM alone, with
overt diabetes, and without diabetes
CharacteristicsGDM
Overt
diabetes
No
diabetes
P value
n
Serum creatinine (mg/dl)
eGFR (ml/min per 1.73 m2)
eGFR ?60 ml/min per 1.73 m2(%)
ACR (%)
Normal (?30 mg/g)
Microalbuminuria (30–300 mg/g)
Macroalbuminuria (?300 mg/g)
CKD stages (%)
No CKD
Stage 1
Stage 2
Stage 3
Stage 4–5
Data are means ? SD or %.
57112,100
0.88 ? 0.31
77.9 ? 23.2
21.0
25,045
0.83 ? 0.23
82.5 ? 22.7
14.1
0.81 ? 0.20
85.8 ? 23.4
11.7
?0.0001
?0.0001
?0.0001
?0.0001
89.3
10.0
0.7
84.4
13.6
2.0
91.7
7.7
0.6
?0.0001
78.6
3.8
5.8
11.4
0.4
68.7
3.8
6.4
19.8
1.3
79.6
2.6
3.8
13.5
0.5
Table 3—Association of GDM and overt diabetes with prevalent microalbuminuria, macroalbuminuria, CKD stages 1–2, and CKD stages
3–5 among KEEP female participants
Condition GDM
P value Overt diabetes
P valueNo diabetes
Microalbuminuria vs. normal (n ? 37,305)
Univariate analysis
Multivariate analysis*
Macroalbuminuria vs. normal (n ? 34,101)
Univariate analysis
Multivariate analysis*
Macroalbuminuria vs. microalbuminuria (n ? 4,026)
Univariate analysis
Multivariate analysis*
CKD stages 1–2 vs. no CKD (n ? 31,581)
Univariate analysis
Multivariate analysis†
CKD stages 3–5 vs. no CKD (n ? 34,833)
Univariate analysis
Multivariate analysis†
CKD stages 3–5 vs. stages 1–2 (n ? 9,018)
Univariate analysis
Multivariate analysis†
Data are ORs (95% CI). *Adjusted for age, race, BMI, current smoking, alcohol use, hypertension, dyslipidemia, eGFR, and family history of kidney disease.
†Adjusted for age, race, BMI, current smoking, alcohol use, hypertension, dyslipidemia, and family history of kidney disease.
1.34 (1.02–1.77)
1.36 (1.03–1.80)
0.04
0.03
1.93 (1.80–2.07)
1.67 (1.55–1.80)
?0.001
?0.001
1.00 (referent)
1.00 (referent)
1.13 (0.42–3.07)
1.13 (0.41–3.09)
0.8
0.8
3.51 (2.87–4.29)
2.68 (2.16–3.31)
?0.001
?0.001
1.00 (referent)
1.00 (referent)
0.85 (0.30–2.36)
0.93 (0.33–2.64)
0.7
0.9
1.82 (1.47–2.24)
1.56 (1.25–1.96)
?0.001
?0.001
1.00 (referent)
1.00 (referent)
1.54 (1.16–2.05)
1.54 (1.16–2.05)
0.003
0.003
1.89 (1.75–2.04)
1.68 (1.55–1.82)
?0.001
?0.001
1.00 (referent)
1.00 (referent)
0.84 (0.65–1.09)
0.94 (0.71–1.25)
0.2
0.7
1.73 (1.64–1.83)
1.20 (1.12–1.28)
1.00 (referent)
1.00 (referent)
?0.001
?0.001
0.06
?0.001
0.55 (0.38–0.79)
0.70 (0.46–1.05)
0.001
0.08
0.92 (0.84–1.00)
0.75 (0.68–0.84)
1.00 (referent)
1.00 (referent)
GDM in KEEP
2588
DIABETES CARE, VOLUME 33, NUMBER 12, DECEMBER 2010care.diabetesjournals.org

Page 4

with overt diabetes, compared with those
for nondiabetic women.
CONCLUSIONS — In this cross-
sectional study of ?37,000 KEEP female
participants, we report that women with
GDM alone have higher prevalence rates
of microalbuminuria than women with-
out any history of diabetes, translating to
higher rates of CKD stages 1–2. The odds
for this early stage of CKD among women
with GDM alone fall roughly midway be-
tweentheoddsfordiseaseamongwomen
without any diabetes and women with
overt diabetes. In stratified analyses,
younger age, African American race, obe-
sity, and hypertension substantially in-
creased the odds for CKD stages 1–2
among women with GDM alone, virtually
matching the odds for disease among
women with overt diabetes.
GDM is clearly recognized as an im-
portant disease entity with health impli-
cations that persist well beyond the
gestational period. As is appropriate, the
mostattentionispaidtotheincreasedrisk
for recurrent GDM, impaired glucose tol-
erance, and overt diabetes in women with
prior episodes of GDM (1,4,5). However,
anepisode(orepisodes)ofGDMalsomay
impart increased risk for disease beyond
mere blood glucose abnormalities. Sev-
eral reports have demonstrated an in-
creased risk of metabolic syndrome in
women with GDM as early as 3 months
postpartumandasfaras11yearsafterthe
index pregnancy, independent of subse-
quent overt diabetes (5–7). Women with
GDM have also demonstrated a higher
risk for subsequent cardiovascular events
than women with euglycemic pregnan-
cies in two large population-based co-
horts, although this risk was somewhat
attenuated after adjustment for subse-
quent diabetes (8,9).
Here, we present data suggesting that
GDM is associated with microalbumin-
uria and, by extension, CKD stages 1–2.
Microalbuminuriaisamarkerofendothe-
lial dysfunction and an independent risk
factor for cardiovascular events (10,11).
Thus, the presence of microalbuminuria
clearly marks a patient for whom lifestyle
(e.g.,low-saltdiet)ormedical(e.g.,renin-
angiotensin system blockade) interven-
tion can be beneficial. The results
presented here suggest that women with
GDM may represent a group for whom
such an intervention may prove benefi-
cial. Indeed, we have previously shown
that the overlap between microalbumin-
uriaandreducedeGFRisverylowamong
younger individuals in KEEP (12); thus,
the finding of microalbuminuria offers
the best chance not only to detect kidney
damageassociatedwithGDMbutalsothe
potential to intervene long before signifi-
cant dysfunction ensues.
The finding of a greater risk for mi-
croalbuminuria, but not macroalbumin-
uria, also suggests that GDM may be a
transient manifestation of preexisting and
persistent metabolic dysfunction. The
later development of overt cardiovascular
disease among these women, indepen-
Table 4—Adjusted odds ratios for CKD stages 1–2 and CKD stages 3–5 among KEEP female participants, stratified by age, race, and BMI
Condition GDM
P valueOvert diabetes
P valueNo diabetes
CKD stages 1–2 vs. no CKD
Age*
?50 years (n ? 12,326)
?50 years (n ? 19,255)
Race†
White (n ? 15,066)
African American (n ? 10,945)
Other (n ? 5,570)
BMI‡
?30 kg/m2(n ? 16,964)
?30 kg/m2(n ? 14,617)
Hypertension§
No. (n ? 16,964)
Yes (n ? 14,617)
CKD stages 3–5 vs. no CKD
Age*
?50 years (n ? 11,796)
?50 years (n ? 23,037)
Race†
White (n ? 17,942)
African American (11,187)
Other (n ? 5,704)
BMI‡
?30 kg/m2(n ? 18,774)
?30 kg/m2(n ? 16,059)
Hypertension§
No (n ? 18,774)
Yes (n ? 16,059)
Data are ORs (95% CI). *Adjusted for race, BMI, current smoking, alcohol use, hypertension, dyslipidemia, and family history of kidney disease. †Adjusted for age,
BMI, current smoking, alcohol use, hypertension, dyslipidemia, and family history of kidney disease. ‡Adjusted for age, race, current smoking, alcohol use,
hypertension, dyslipidemia, and family history of kidney disease. §Adjusted for age, race, BMI, current smoking, alcohol use, dyslipidemia, and family history of
kidney disease.
1.72 (1.17–2.53)
1.33 (0.87–2.04)
0.006
0.2
1.69 (1.46–1.95)
1.74 (1.57–1.93)
?0.0001
?0.0001
1.00 (referent)
1.00 (referent)
1.12 (0.68–1.84)
2.32 (1.50–3.60)
1.35 (0.75–2.42)
0.71.45 (1.28–1.65)
1.92 (1.68–2.19)
1.83 (1.52–2.21)
?0.0001
?0.0001
?0.0001
1.00 (referent)
1.00 (referent)
1.00 (referent)
0.0002
0.3
1.25 (0.82–1.89)
1.93 (1.30–2.86)
0.3
0.001
1.54 (1.36–1.74)
1.89 (1.69–2.11)
?0.0001
?0.0001
1.00 (referent)
1.00 (referent)
1.24 (0.82–1.88)
1.91 (1.29–2.84)
0.3
0.001
1.60 (1.41–1.81)
1.86 (1.66–2.08)
?0.0001
?0.0001
1.00 (referent)
1.00 (referent)
0.66 (0.31–1.42)
0.97 (0.72–1.30)
0.3
0.8
1.10 (0.90–1.36)
1.32 (1.24–1.41)
0.4 1.00 (referent)
1.00 (referent)
?0.0001
0.83 (0.58–1.19)
1.66 (0.96–2.88)
0.64 (0.29–1.43)
0.3
0.07
0.3
1.07 (0.98–1.16)
1.43 (1.26–1.62)
1.39 (1.16–1.66)
0.11.00 (referent)
1.00 (referent)
1.00 (referent)
?0.0001
0.0003
1.01 (0.70–1.44)
0.89 (0.56–1.39)
1.0
0.6
1.18 (1.07–1.29)
1.25 (1.14–1.36)
0.0006
?0.0001
1.00 (referent)
1.00 (referent)
0.99 (0.69–1.4)
0.87 (0.56–1.38)
0.9
0.6
1.19 (1.09–1.31)
1.22 (1.12–1.33)
0.0002
?0.0001
1.00 (referent)
1.00 (referent)
Bomback and Associates
care.diabetesjournals.orgDIABETES CARE, VOLUME 33, NUMBER 12, DECEMBER 2010
2589

Page 5

dent of overt metabolic syndrome or dia-
betes, lends support to the hypothesis
that women with previous GDM are at
risk for cardiovascular disease because of
subclinical inflammation and enduring,
generalized vascular dysfunction. Inflam-
matory markers that have been shown to
be independent predictors of future car-
diovascular events, including C-reactive
protein, interleukin-6, and plasminogen
activator inhibitor-1, are significantly ele-
vatedinwomenwithpreviousGDMcom-
pared with control women(13). Women
with previous episodes of GDM also ex-
hibit lower levels of plasma adiponectin,
an anti-inflammatory peptide (14), and
have higher central systolic pressure,
higher mean arterial pressure, higher pe-
ripheral resistance, lower mean stroke
volume, lower cardiac output, and higher
carotid intimal-medial thickness than
normoglycemic women without previous
GDM (15).
The strength of this study lies in the
large size and diversity of the cohort as-
sembled by KEEP screening activities.
Friedman et al. (16) previously reported
higher risk for microalbuminuria in
women with GDM compared with con-
trol women, but this observation was
based on data from only 72 individuals.
Goetal.(17),inanentirelyAfricanAmer-
ican cohort of 289 women with GDM as-
sessed at a median of 11 years after
delivery, reported an 11% prevalence of
microalbuminuria in normoglycemic
women versus 36% prevalence in women
who had developed overt diabetes after
GDM. Our large cohort of 571 women
with GDM alone allows us to find mean-
ingfulrelationshipswithinsubgroups.In-
deed, stratified analyses (Table 4) suggest
that GDM may be a different, and more
harmful, disease for African American
women and women with obesity com-
pared with white women and nonobese
women, respectively. As demonstrated in
recent reports from KEEP (18,19), these
patient groups may derive extra benefit
fromscreeningforalbuminuriaintheset-
ting of preserved GFR. Our data also sug-
gest that this screening is crucial when
hypertension, with or without concomi-
tant diabetes, has developed in the years
after delivery.
The study has limitations, however.
Because the analysis is cross-sectional,
microalbuminuria could have preceded
(and been responsible for) GDM in many
of the cases analyzed here rather than
have resulted from prior GDM. Previous
reports have suggested that microalbu-
minuria is associated with insulin resis-
tance in nondiabetic subjects and can
augur the arrival of subsequent diabetes
(20,21). We did not have data on the ex-
act date of pregnancy for these partici-
pants, and therefore inferences about the
effect of age may not be valid. Younger
women with GDM appeared to have a
higher risk for CKD stages 1–2 than older
women with GDM, but this may simply
represent closer proximity to the preg-
nancy. Older women may have pro-
gressed to overt diabetes by the time they
presented for screening activities and
been equally or more likely to have renal
dysfunction. In addition, although we
had data on BMI for all participants, we
only had limited data on waist circumfer-
ence. Abdominal adiposity has been
shown to predict the development of ab-
normal urinary albumin excretion and
may have better identified a subgroup at
risk than BMI in this cohort (22). Finally,
KEEP was designed to screen participants
at higher risk for CKD than the general
population, and therefore the results pre-
sented here may not be generalizable to
women with GDM and otherwise low-
risk profiles. Further studies into the link
betweenGDMandCKDideallyshouldbe
longitudinal,includewaistcircumference
as a covariate, and include women at var-
ious levels of risk for chronic disease.
In summary, in a national cohort of
?37,000 KEEP female participants,
women with GDM alone demonstrated a
greater prevalence of microalbuminuria
than normoglycemic women. African
American and obese women with prior
GDM emerged as having the highest odds
of CKD stages 1–2, comparable to that of
women with overt diabetes. These results
suggest that GDM, even in the absence of
subsequent overt diabetes, is marked by
generalized endothelial dysfunction and
may place women at increased risk for
future cardiovascular and kidney disease.
Acknowledgments— KEEP is a program of
the National Kidney Foundation, Inc., and is
supported by Amgen, Abbott, Novartis, Sie-
mens,Genentech,Genzyme,Nephroceuticals,
Pfizer, LifeScan, and Suplena.
No other potential conflicts of interest rele-
vant to this article were reported.
A.S.B. conceived the study concept and
wrote the manuscript. Y.R. researched data
and reviewed/edited the manuscript.
A.T.W.-C., A.V.K., J.R.S., K.M.C., G.L.B.,
and P.A.M. reviewed/edited the manuscript.
S.-C.C. and S.L. performed the statistical
analyses.
References
1. Bellamy L, Casas JP, Hingorani AD, Wil-
liams D. Type 2 diabetes mellitus after
gestational diabetes: a systematic review
and meta-analysis. Lancet 2009;373:1773–
1779
2. Baptiste-Roberts K, Barone BB, Gary TL,
Golden SH, Wilson LM, Bass EB, Nichol-
son WK. Risk factors for type 2 diabetes
among women with gestational diabetes:
a systematic review. Am J Med 2009;122:
207–214 e204
3. Jurkovitz CT, Qiu Y, Wang C, Gilbertson
DT, Brown WW. The Kidney Early Eval-
uation Program (KEEP): program design
and demographic characteristics of the
population. Am J Kidney Dis 2008;
51:S3–S12
4. FeigDS,ZinmanB,WangX,HuxJE.Risk
of development of diabetes mellitus after
diagnosis of gestational diabetes. CMAJ
2008;179:229–234
5. RetnakaranR,QiY,ConnellyPW,Sermer
M, Zinman B, Hanley AJ. Glucose intoler-
ance in pregnancy and postpartum risk of
metabolic syndrome in young women.
J Clin Endocrinol Metab 2010;95:670–
677
6. Akinci B, Celtik A, Yener S, Yesil S. Pre-
diction of developing metabolic syn-
drome after gestational diabetes mellitus.
Fertil Steril 2010;93:1248–1254
7. Lauenborg J, Mathiesen E, Hansen T,
Glu ¨merC,JørgensenT,Borch-JohnsenK,
Hornnes P, Pedersen O, Damm P. The
prevalenceofthemetabolicsyndromeina
Danish population of women with previ-
ous gestational diabetes mellitus is three-
foldhigher than
population. J Clin Endocrinol Metab
2005;90:4004–4010
8. Shah BR, Retnakaran R, Booth GL. In-
creased risk of cardiovascular disease in
young women following gestational dia-
betes mellitus. Diabetes Care 2008;31:
1668–1669
9. Retnakaran R, Shah BR. Mild glucose in-
tolerance in pregnancy and risk of cardio-
vascular disease: a population-based
cohort study. CMAJ 2009;181:371–376
10. Giner V, Tormos C, Chaves FJ, Sa ´ez G,
Redo ´n J. Microalbuminuria and oxidative
stress in essential hypertension. J Intern
Med 2004;255:588–594
11. Kistorp C, Raymond I, Pedersen F,
Gustafsson F, Faber J, Hildebrandt P. N-
terminal pro-brain natriuretic peptide, C-
reactive protein, and urinary albumin
levels as predictors of mortality and car-
diovascular events in older adults JAMA
2005;293:1609–1616
12. McCullough PA, Vassalotti JA, Collins AJ,
Chen SC, Bakris GL. National Kidney Foun-
dation’s Kidney Early Evaluation Program
(KEEP) annual data report 2009: executive
summary.AmJKidneyDis2010;55:S1–S3
13. Heitritter SM, Solomon CG, Mitchell GF,
in thegeneral
GDM in KEEP
2590
DIABETES CARE, VOLUME 33, NUMBER 12, DECEMBER 2010care.diabetesjournals.org