International Society for Pediatric and Adolescent Diabetes
ISPAD 2014 | 40th Anniversary |
255
P167
Racial disparity between African-Americans and Caucasians in HbA1c variability and
microalbuminuria risk among children with type 1 diabetes mellitus (T1D)
S. Raman
1,2,3
, H. Dai
4
, D.D. Williams
4
, S.A. DeLurgio
4
, M.A. Clements
1,2,3
1
Children's Mercy Hospital, Pediatric Endocrinology, Kansas City, United States,
2
University of
Missouri- Kansas City, Pediatrics, Kansas City, United States,
3
University of Kansas Medical Center,
Pediatrics, Kansas City, United States,
4
Children's Mercy Hospital, Kansas City, United States
Background:
African-Americans (AA) with T1D have higher risk for complications compared to
Caucasians (CC). HbA1c variability, as measured by Standard Deviation (SDA1c), influences T1D
nephropathy. Whether AA children with T1D have higher SDA1c and/or higher risk for
microalbuminuria (MAU) is unknown.
Objective:
To assess the differences in SDA1c and MAU risk between AA and CC children with T1D.
Methods:
Using retrospective chart review (1993-2009), children with T1D wh
o had ≥4 HbA1c values,
≥2 HbA1c⁄yr, and ≥1 year of T1D were identified. MAU was defined as albumin excretion rate
≥20mcg⁄min or urine albumin⁄creatinine ≥30mg⁄gm in 2 of 3 consecutive samples.
Results:
Among 1198 patients included, 1059(88%) were CC and 91(8%) AA (Table1A). AA had
higher meanA1c & SDA1c (p< .0001). In addition, 74% AA vs. 47% CC had SDA1c above the cohort
median of 1.47 (p< .0001). When stratified by both meanA1c and SDA1c, 64.8% of AA vs. only 27.6%
of CC fell in the HighMean/HighSD group (p< .0001). Although MAU incidence was not different, time
to MAU was significantly longer in AA. In Cox Proportional Hazard model (Table1B), SDA1c was
associated with MAU (HR 1.52, p=.0017) and AA had lower risk compared to CC (HR 0.28, p=.01).
1A) Summary of Patient Characteristics
1B) Cox Proportional Hazard (CPH)
Model for MAU
Variable
Entire
Cohort
(All,
n=1198)
Caucasians
(CC,
n=1059)
African-
Americans
(AA, n=91)
p-value
AA Vs.
CC
Variable
Hazard
Ratio
95%CI
p-value
Gender
(% Male) 53%
53%
48%
0.49
meanA1c 1.05
0.95 -
1.17
0.34
Age at
T1D
diagnosis
(yrs)
9.1 ± 3.9 9.1 ± 3.9
9.21 ± 3.4 0.88
SDA1c 1.52
1.17 -
1.97
0.0017
Duration
of T1D
(yrs)
5.7 ± 3.5 5.7 ± 3.5
6.1 ± 3.7
0.27
Age at
T1D
diagnosis
1.15
1.09 -
1.22
<.0001
MeanA1c
(%)
8.8 ± 1.3 8.7 ± 1.1
10.5 ± 2
<.0001
Gender
(M vs. F) 1.15
0.81-
1.64
0.44
SDA1c
(%)
1.7 ± 0.9 1.6 ± 0.9
2.2 ± 1.1
<.0001
Post-hoc analysis for Race
MAU
Incidence
14.4%
(n=172)
15%
(n=156)
9% (n=8) 0.12
AA vs.
CC
0.28
0.1-0.78
0.02
Time to
MAU(yrs)
4.6 ± 3.1
(n=172)
Min:1.0
Max:
15.2
Median:
4.0
4.6 ± 2.9
(n=156)
Min:1.0
Max: 15.2
Median: 4.0
7.5 ± 4.6
(n=8)
Min:1.1
Max: 13.3
Median:
7.5
0.0081
Non-AA
vs. CC
1.29
0.56-
2.98
0.55
Table 1: Summary of patient characteristics (1A) and results of the CPH model for MAU (1B).
Data in 1A are presented as Mean ± SD. Results in 1B are shown as hazard ratios for every 1-unit
increase in each variable. The CPH model was designed using annually updated meanA1c series &
SDA1c series and adjusted for gender, race, & age at T1D diagnosis; duration of T1D was used as
the time variable. Analyses were done using SAS 9.2 and SPSS 20.
[Table 1]
