Dawn R. Albrecht, BA. Daniel C. Postellon, MD. James P. Gutai, MD. Objective: This study evaluated racial differences in the metabolic control of children and ...
Racial Differences in Metabolic Control of Children and Adolescents With Type I Diabetes Mellitus
Objective: This study evaluated racial differences in the metabolic control of children and adolescents with insulin-dependent (type I) diabetes mellitus and examined the interactive effects of race with age and sex. Research Design and Methods: Data on several demographic and clinical variables were obtained for 102 black and 108 white children, including the percentage of total HbA,, age, age at diagnosis, duration of diabetes, pubertal status, insulin dose (U * k g 1 • day 1 ), body mass index, number of clinic visits kept and missed, number of hospitalizations for diabetic ketoacidosis (DKA) for the year, and socioeconomic status (SES). Results: Black children had higher insulin dosages (P < 0.05) and lower SESs (P < 0.001) than white children. HbA, was higher in black than white children (P < 0.01) after statistically adjusting for the effects of insulin dose, diabetes duration, and SES. With HbA,based criteria, more black than white children were in poor and fewer in good metabolic control (P < 0.001). Older children (>13 yr) had higher HbA, levels than younger ( 0.12 for all variables); however, patients without SES data kept fewer (P < 0.004) and missed more (P < 0.002) clinic appointments. Because these patients generally did not have telephones, it can be inferred that they were of lower
TABLE 1 Demographic and clinical characteristics of black and white children and adolescents with insulin-dependent diabetes mellitus Variable Demographic Age (yr) n (M/F) Socioeconomic status Single parent (%) Clinical Age at diagnosis (yr) Duration of diabetes (yr) Tanner stage Body mass index Insulin dose (U • kg" 1 • day"1) HbA, (%) Previous HbA, (%) Clinic visits kept Clinic visits missed Diabetic ketoacidosis hospitalizations
Black (n = 102)
White (n = 106)
F
12.24 ± 4.27 59/47 35.50 ± 13.79 20
0.46
37/65 26.28 ± 15.02 60 8.32 4.47 2.98 21.35 0.93 11.54 11.18 2.10 0.59 0.60
7.32 dt 5.09 dt 2.73 dt 20.31 dt 0.83 dt 10.45 dt 10.10 ± 2.13 ± 0.26 ± 0.26 ±
± ± ± ± ± ± ± ± ± ±
4.30 3.52 1.52 4.67 0.38 2.57 2.47 0.96 0.81 1.31
3.85 3.55 1.58 6.88 0.29 2.16 2.34 0.91 0.54 0.70
X2
7.10* 18.35t 32.31 + 3.16 1.58 0.87 1.61 3.85* 11.01 + 7.21* 4.95 11.87* 8.34*
Data are expressed as means ± SD, except for sex and single parent. Sample size for socioeconomic status was n - 87 for black and n = 92 for white patients, previous HbA, was n = 67 for black and n = 76 for white patients, and Tanner stage was n = 63 for black and n = 62 for white patients. *P < 0.01, +P < 0.001, *P < 0.05. DIABETES CARE, VOL. 14, NO. 1, JANUARY 1991
21
RACIAL DIFFERENCES IN METABOLIC CONTROL
SES; an equal number of black and white patients were represented in this group. The primary measure of metabolic control tor this study was the percentage of total HbA, (i.e., HbA1a + HbA1b + HbAlc). The HbA, concentration was determined from a saline-incubated (37°C for 20 min) blood sample with an ion-exchange minicolumn method (Fast Hemoglobin Test System, Isolab, Akron, OH). The assay was performed at 23°C at the Clinical Chemistry Laboratory at Children's Hospital of Michigan. The normal nondiabetic mean for this assay was 7% (range 5.58.5%). The coefficient of variation for day-to-day reproducibility of the test was 3%. Because this assay is compromised in the presence of sickle-cell trait, all black children had previous hemoglobin electrophoresis to confirm the validity of this method for the determination of HbA,. The overall mean ± SD for the study sample (n = 208) was 10.98 ± 2.42% (range 5.0-18.2%, SE 0.17, skewness 0.53). Statistical analyses. This study was designed to determine whether there are racial differences in metabolic control of children and adolescents with type I diabetes mellitus and to examine possible interactive effects of race with age and sex. We first used analysis of variance or x2-test to evaluate racial and sex differences on demographic and clinical measures. Pearson product-moment correlational analyses were conducted to determine relationships of HbA, with relevant demographic and clinical measures; multiple regression analysis was then used to identify significant independent predictors of HbA,. The primary analysis of the study used analysis of covariance (ANCOVA) conducted on HbA,. The model included main effects of race, sex, age-group (children
