ARTICLE
Diabetes Mellitus Screening in Pediatric Primary Care
Shikha G. Anand, MD, MPHa, Supriya D. Mehta, PhD, MHSb, William G. Adams, MDa
aDivision of General Pediatrics andbDepartment of Emergency Medicine, Boston University School of Medicine, Boston, Massachusetts
The authors have indicated they have no financial relationships relevant to this article to disclose.
ABSTRACT
OBJECTIVE.The goal was to determine the rates of diabetes screening and the prev-alence of screening abnormalities in overweight and nonoverweight individuals in an urban primary care clinic.
METHODS.This study was a retrospective chart review conducted in a hospital-based urban primary care setting. Deidentified data for patients who were 10 to 19 years of age and hadⱖ1 BMI measurement between September 1, 2002, and September 1, 2004, were extracted from the hospital electronic health record.
RESULTS.A total of 7710 patients met the study criteria. Patients were 73.0% black or Hispanic and 47.0% female; 42.0% of children exceeded normal weight, with 18.2% at risk for overweight and 23.8% overweight. On the basis of BMI, family history, and race, 8.7% of patients met American Diabetes Association criteria for type 2 diabetes mellitus screening, and 2452 screening tests were performed for 1642 patients. Female gender, older age group, and family history of diabetes were associated with screening. Increasing BMI percentile was associated with screen-ing, exhibiting a dose-response relationship. Screening rates were significantly higher (45.4% vs 19.0%) for patients who met the American Diabetes Association criteria; however, less than one half of adolescents who should have been screened were screened. Abnormal glucose metabolism was seen for 9.2% of patients screened.
CONCLUSIONS.This study shows that, although pediatricians are screening for diabetes mellitus, screening is not being conducted according to the American Diabetes Association consensus statement. Point-of-care delivery of consensus recommen-dations could increase provider awareness of current recommenrecommen-dations, possibly improving rates of systematic screening and subsequent identification of children with laboratory evidence of abnormal glucose metabolism.
www.pediatrics.org/cgi/doi/10.1542/ peds.2006-0121
doi:10.1542/peds.2006-0121
Presented in part at the annual meeting of the Pediatric Academic Societies; May 15, 2006; Washington, DC.
Key Words
type 2 diabetes mellitus, screening, childhood obesity
Abbreviations
T2DM—type 2 diabetes mellitus ADA—American Diabetes Association FPG—fasting plasma glucose OGTT— oral glucose tolerance test RPG—random plasma glucose PPCC—pediatric primary care center EHR— electronic health record HbA1c— hemoglobin A1c CI— confidence interval
A
LTHOUGH PREVIOUSLY CONSIDEREDan adult disease, type 2 diabetes mellitus (T2DM) is increasing in US children1–3and children around the world.4,5T2DM nowaccounts for up to 45% of all newly diagnosed cases of diabetes in US children.6Black and Latino children show
the greatest rates of increase.7 The increasing
preva-lences of overweight8 and T2DM9 in minority youths
make the need for comprehensive T2DM screening an essential part of urban pediatric primary care.
The early detection of pediatric T2DM has become a national priority.6,10In 2000, the American Diabetes
As-sociation (ADA) issued a consensus statement on T2DM screening in children.6In that statement, the ADA
rec-ommends that children who are at risk for overweight or overweight (BMIⱖ85th percentile for age and gender) and have 2 of 3 risk factors should be screened (Table 1). Biannual testing beginning at 10 years of age, or at the onset of puberty if that occurs sooner, is recommended. Discretionary testing is recommended for other high-risk patients, as assessed by their medical providers. Fasting plasma glucose (FPG) is recommended as the primary screening test for T2DM because of decreased cost and greater convenience, compared with the 2-hour oral glucose tolerance test (OGTT), which is recommended as a secondary test. No recommendation has been made regarding the use of random (nonfasting) plasma glucose (RPG) or hemoglobin A1c (HbA1c) measurements for T2DM screening in children.6
The purpose of our study was to describe how pedi-atricians screen for diabetes in an urban pediatric pri-mary care center (PPCC) and to describe the results of screening for those patients. For this report, all possible screening tests related to diabetes, including FPG, OGTT, RPG, and HbA1c, were assessed. Specific outcome mea-sures included the type of screen, factors associated with screening, and results of screening.
METHODS
Study Design and Population
The study was a retrospective chart review of electronic health record (EHR) data from all encounters that
oc-curred during the period between September 1, 2002, and September 1, 2004, for children 10 through 19 years of age in the PPCC at Boston Medical Center. The PPCC serves a population of predominantly low-income black, Caribbean American, and Hispanic children and adoles-cents. Providers include pediatric residents, nurse prac-titioners, and general pediatric and adolescent attending physicians.
The PPCC has used an EHR (Centricity; GE Medical Systems, Waukesha, WI) since September 1, 2001. To al-low providers to adapt to the EHR, the review period was initiated 1 year after EHR implementation. Children were included if they hadⱖ1 visit to the PPCC during the study period and height and weight measured at a PPCC visit on the same day. Because of the effects of pregnancy on glucose metabolism and our inability to determine the precise period of pregnancies, patients who were pregnant or ever pregnant were excluded.
Although electronic laboratory data from the PPCC were available from June 1, 2001, onward, only labora-tory tests that occurred within 90 days before the study period were included in the analyses. This ensured that the laboratory tests included were temporally associated with available BMI measurements. Although it is possi-ble that some patients were excluded from analyses be-cause they had been screened before the beginning of the analytic observation period, the ADA recommends repeat screening biannually and most eligible patients should have been screened at least once during the 2-year study period.
Data Extraction
The following data were extracted from the EHR data-base: age, height, weight, family history, and HbA1c, RPG, FPG, and OGTT results. No medical record was extracted in its entirety. BMI was calculated automati-cally from each height/weight pair obtained, with the formula BMI ⫽ weight/height2 (with weight in kilo-grams and height in meters). BMI percentile was deter-mined by using Centers for Disease Control and Preven-tion reference data.11
To ensure that we captured all possible screening results, any laboratory test that was performed within 90 days before a BMI measurement or any time after a BMI measurement and within the 2-year study period was included in our analyses. If a patient had⬎1 BMI mea-surement associated with a screening test, then only the BMI measurement closest in time to the most abnormal laboratory result was used in the analyses. For patients who had no abnormal laboratory test results but⬎1 BMI measurement, we used the highest BMI, to ensure re-cording of all patients with elevated BMI measurements within the study period.
The ADA considers a family history of T2DM in a first-or second-degree relative to be a risk factfirst-or ffirst-or T2DM in a child and uses family history as a screening criterion TABLE 1 Testing Guidelines for T2DM in Children
Age of 10 years or at onset of puberty if puberty occurs at a younger age Overweight, defined by BMI of⬎85th percentile for age and gender plus any 2 of
the following risk factors:
Family history of T2DM in first- or second-degree relative
Race/ethnicity (American Indian, black, Hispanic, or Asian/Pacific Islander) Signs of insulin resistance or conditions associated with insulin resistance
(acanthosis nigricans, hypertension, dyslipidemia, or polycystic ovary syndrome)
Frequency of every 2 years FPG is preferred test
Clinical judgment should be used to test for T2DM in high-risk patients who do not meet these guidelines
(Table 1). At the PPCC, family histories are recorded by clinicians as unstructured free text in the EHR. Free-text fields were transferred into a spreadsheet, where the data were searched by using the string search terms “dm,” “di,” and “dai.” Family history fields containing any of those strings were reviewed by a physician (S.G.A.) for the presence of any family history of diabe-tes. Records with a family history of T2DM or an un-specified history of diabetes were coded as positive for a family history of T2DM. Patients with a family history of diabetes in distant relatives (not first or second degree) were coded as negative for a family history of T2DM.
Patients were considered to meet the ADA criteria for T2DM screening if they met the following criteria: family history of T2DM, minority race, and BMI ofⱖ85th per-centile for age and gender. Because of our inability to extract physical examination data, physical examination findings and the presence of comorbidities of diabetes (Table 1) were not included in our definition of patients meeting ADA screening criteria.
Laboratory Measures
Patients were classified as having undergone screening if they had a FPG measurement, RPG measurement, 2-hour OGTT, or HbA1c measurement during the study period. Patients with laboratory evidence of abnormal glucose metabolism (diabetes, impaired FPG, impaired glucose tolerance, or unspecified) were classified accord-ing to test type and level of impairment. Classifications were defined as follows: diabetes: FPG level of ⱖ126 mg/dL or OGTT result of ⱖ200 mg/dL6; impaired FPG:
FPG level of 100 to 125 mg/dL12; impaired glucose
tol-erance: OGTT result of 140 to 199 mg/dL12; unspecified:
RPG level ofⱖ200 mg/dL or HbA1c level ofⱖ6.0%. Although RPG and HbA1c measurements are not ADA-recommended screening tests for children, they have been suggested by experts as possible modes of testing and have been used previously for screening in settings other than pediatric primary care.13–17They were
included in our analyses because providers may use them for screening to avoid having patients return in the fasting state. Threshold values were determined by using ADA standards for adults with symptoms of diabetes for RPG and those for known diabetic patients for HbA1c.12
This study was approved by the institutional review boards of Boston University School of Medicine and the Boston Medical Center.
Data Analyses
Frequencies and 2 analyses of screening rates and re-sults were performed for the following variables: age, race, gender, family history of T2DM, BMI percentile, and presence of ADA criteria for screening. Age and BMI percentile were analyzed as categorical variables. Uni-variate analyses were performed for each of the afore-mentioned variables, and 95% confidence intervals (CIs)
were calculated. Logistic regression analyses were used to analyze the independent effects of each sociodemo-graphic characteristic on the outcomes of screening and laboratory evidence of abnormal glucose metabolism, by using backward likelihood ratio testing. The 17 Native American subjects in the sample (6 of whom were screened for T2DM) were categorized as other/unknown for logistic regression analyses. Variables that demon-strated significant associations at the level ofP⬍.05 in univariate analyses were included in the multivariate models. ADA risk was not included in the final regres-sion analyses because of its colinearity with other risk factors in the model. Data were analyzed by using Stata SE 8.0 for Windows software (Stata Corp, College Park, TX).
RESULTS
Patient Characteristics
The 7710 patients who met the study criteria were dis-tributed nearly equally among all age categories (10 –12, 13–15, and 16 –19 years of age) (Table 2). The majority of children were black (57.6%) and female (53.0%). A large proportion (42.0%) of children exceeded normal weight, with 18.2% at risk for overweight (BMI of 85th to 94th percentile for age and gender) and 23.8% over-weight (BMI ofⱖ95th percentile for age and gender). A high percentage (17.7%) had BMI values of⬎97th per-centile. Family history was positive for T2DM for 19.6% of patients. On the basis of BMI, family history, and race, 8.7% of patients met ADA criteria for screening (Table 2).6
Screening
Overall, 2452 screening tests were performed for 1642 patients (21.3%) in the cohort (Table 2). In bivariate analyses, screening was significantly more common for the 16- to 19-year-old age group, female patients, pa-tients with a family history of diabetes, and papa-tients who met ADA criteria for screening with our limited criteria. White, black, Hispanic, and Asian patients were all screened at similar rates (20.1%–24.4%) (Table 2). In-creasing BMI increased the likelihood of screening, with a dose-response relationship between increasing BMI and screening rates.
tests than were those who did not (22.6% vs 10.0%;P⬍ .001).
In logistic regression analyses, older children, female
patients, patients with higher BMI values, and patients with a family history of T2DM were more likely to be screened (Table 4). For example, 57.8% of female pa-TABLE 2 Patient Characteristics and Screening
Total No. in Class (% of Total Sample)
No. Screened (% of Class)
2
Age 169.1a
10–12 y 2760 (35.8) 438 (15.9)
13–15 y 2344 (30.4) 431 (18.4)
16–19 y 2606 (33.8) 773 (29.7)
Race 52.0a
White 522 (6.8) 105 (20.1)
Black 4441 (57.6) 1001 (22.5)
Hispanic 1186 (15.4) 289 (24.4)
Asian 380 (4.9) 85 (22.4)
Native American 17 (0.2) 6 (35.3)
Other 1164 (15.1) 156 (13.4)
Gender 69.4a
Male 3627 (47.0) 623 (17.2)
Female 4083 (53.0) 1019 (25.0)
BMI classification for age and gender
⬍85th percentile 4469 (58.0) 719 (16.1) 335.7a
85th to 89th percentile 620 (8.0) 115 (18.5)
90th to 94th percentile 784 (10.2) 162 (20.7)
95th to 97th percentile 472 (6.1) 112 (23.7)
⬎97th percentile 1365 (17.7) 534 (39.1)
Family history of T2DM 207.9a
No/not mentioned 6201 (80.4) 1115 (18.0)
Yes 1509 (19.6) 527 (34.9)
ADA criteria 254.9a
Negative 7038 (91.3) 1337 (19.0)
Positive 672 (8.7) 305 (45.4)
Total 7710 (100.0) 1642 (21.3)
aP⬍.001.
TABLE 3 Types of Screening
No. (% of Patients Screened)
All Patients ADA Criteria-Positive Patients No. of different test types per person
1 1005 (61.2)
2 495 (30.1)
3 111 (6.8)
4 31 (1.9)
Screen typea
FPG 203 (12.4)
OGTT 245 (14.9)
RPG 1424 (86.7)
HbA1c 580 (35.3)
Recommended screening tests
Recommended tests only 76 (4.6)
Nonrecommended tests only 1263 (76.9)
Both recommended and nonrecommended tests 303 (18.5)
Abnormal testing resultsa
All tests 151 (100.0) 54 (100.0)
FPG 34 (22.5) 13 (24.0)
OGTT 12 (8.0) 5 (9.3)
RPG 96 (63.6) 34 (63.0)
HbA1c 9 (6.0) 2 (3.7)
Total screened 1642 (100.0)
tients who had a family history of T2DM and a BMI of ⬎97th percentile were screened, compared with 18.0% of female patients who had a negative family history and a BMI of⬍85th percentile.
Laboratory Evidence of Abnormal Glucose Metabolism
Among 1642 screened subjects, 9.2% (n ⫽ 151) had some laboratory evidence of abnormal glucose metabo-lism; 62.9% (n⫽95) were abnormal RPG results (Table 5). Abnormal glucose metabolism was more common for male patients and patients with a family history of dia-betes, BMI between the 90th percentile and 94th per-centile or of ⬎97th percentile, and ADA criteria for screening (Tables 5 and 6). Of all patients who showed laboratory evidence of abnormal glucose metabolism (n ⫽151), 11 (7.3%) met the criteria for T2DM with FPG results, 14 (9.3%) had impaired FPG, and 12 (0.1%) had impaired glucose tolerance. Abnormal glucose metabo-lism was demonstrated by HbA1c or RPG results for 71.3% of patients (n⫽107) with laboratory abnormal-ities.
Of 17 patients who met the criteria for T2DM with FPG results, 8 (47.1%) also had abnormal RPG results and 6 (35.3%) also had abnormal HbA1c results. Of 12 patients who had positive OGTT results, 1 (8.3%) also had abnormal RPG results and 1 (8.3%) also had abnor-mal HbA1c results. Comparisons could not be made among test sensitivities because tests for the same pa-tients might have been performed at different points during the observation period and thus at different stages in a subject’s disease trajectory.
DISCUSSION
In our sample of 7710 patients, 21.3% (n⫽1642) were screened for diabetes by using FPG, RPG, OGTT, HbA1c, or a combination of tests. Children were more likely to be screened if they were older, were female, were more overweight, or had a family history of diabetes. These attributes were cited previously as risk factors for diabe-tes.1,6,10,18
Unfortunately, many children who should have been screened were not. Of 672 children who met ADA cri-teria for screening in our sample, only 305 (45.4%) were screened. In addition, many of the tests performed were not those recommended by the ADA. Overall, only 4.6% of the patients screened had a FPG test or OGTT. The majority of patients screened (71.3%) had a RPG test. This might be attributable to the ease of performing the RPG test, in comparison with recommended tests. A RPG sample can be drawn at a routine visit, eliminating the difficulty (and the associated increase in nonadherence rates) of having the patient return in the fasting state. The frequency of nonrecommended testing in our clinic suggests that additional investigation into the use of RPG and HbA1c measurements as screening tests is war-ranted. Results of direct comparisons of test sensitivities in this setting, which have not been published to date, combined with cost and feasibility considerations, could render a role for pediatric T2DM testing that does not require a fasting state.
Despite increased national emphasis on the early de-tection of T2DM, little is known about the actual screen-ing practices of clinicians in pediatric primary care set-tings. To our knowledge, only one published study is TABLE 4 Factors Associated With Being Screened for T2DM, in Univariate and Multivariate
Logistic-Regression Analyses
Univariate Odds Ratio (95% CI)
Adjusted Odds Ratio (95% CI) Age
10–12 y
13–15 y 1.19 (1.03–1.38) 1.21 (1.04–1.41)
16–19 y 2.25 (1.95–2.55) 2.44 (2.12–2.81)
Race White
Black 1.16 (0.92–1.45) NS
Hispanic 1.28 (0.99–1.65) NS
Asian 1.14 (0.83–1.58) NS
Native American/unknown/other 0.63 (0.48–0.83) 0.63 (0.52–0.75)
Female gender 1.60 (1.43–1.79) 1.48 (1.32–1.67)
Family history of T2DM 2.45 (2.16–2.77) 2.14 (1.88–2.44)
BMI category ⬍85th percentile
85th to 89th percentile 1.19 (0.96–1.47) NS
90th to 94th percentile 1.36 (1.12–1.64) 1.37 (1.13–1.66)
95th to 97th percentile 1.62 (1.29–2.04) 1.65 (1.31–2.08)
⬎97th percentile 3.35 (2.93–3.84) 3.43 (2.98–3.94)
available, a 2004 evaluation of the implementation of a screening protocol in a Chicago pediatric clinic.19In
com-parison with that study, our overall screening rate was
much higher (21.3% vs 7%), with similar rates of screening for ADA criteria-positive patients (45.4% vs 38%). Prevalences of T2DM (0.2% vs 0.0%), impaired TABLE 5 Laboratory Evidence of Abnormal Glucose Metabolism
No. Screened (% of Class)
Abnormal Glucose Metabolism
T2DM
n(%) 2 n(%) 2
Age 1.976 0.01
10–12 y 438 (15.9) 47 (10.7) 3 (0.7)
13–15 y 431 (18.4) 35 (8.1) 3 (0.7)
16–19 y 773 (29.7) 68 (8.8) 5 (0.6)
Race 10.344 4.4
White 105 (20.1) 9 (8.6) 0 (0.0)
Black 1001 (22.5) 109 (10.9) 6 (0.6)
Hispanic 289 (24.4) 16 (5.6) 2 (0.7)
Asian 85 (22.4) 6 (7.1) 2 (2.4)
Native American 6 (35.3) 0 (0.0) 0 (0.0)
Other 156 (13.4) 10 (6.4) 1 (0.6)
Gender 10.154 0.5
Male 623 (17.2) 75 (12.0) 3 (0.5)
Female 1019 (25.0) 75 (7.4) 8 (0.8)
BMI classification for age and gender 27.230 18.4a
⬍85th percentile 719 (16.1) 43 (6.0) 0 (0.0)
85th to 89th percentile 115 (18.5) 7 (6.1) 0 (0.0)
90th to 94th percentile 162 (20.7) 18 (11.1) 0 (0.0)
95th to 97th percentile 112 (23.7) 7 (6.3) 1 (0.9)
⬎97th percentile 534 (39.1) 75 (14.0) 10 (1.9)
Family history of T2DM 4.017 0.1
No/not mentioned 1115 (18.0) 44 (14.5) 7 (0.6)
Yes 527 (34.9) 106 (7.9) 4 (0.8)
ADA criteria 13.857 2.3
Negative 305 (45.4) 91 (8.2) 7 (0.5)
Positive 1337 (19.0) 59 (11.2) 4 (1.3)
Total 1642 (100) 150 (9.1) 11 (0.7)
aP⫽.001.
TABLE 6 Factors Associated With Having Abnormal Glucose Metabolism Among Patients Screened for T2DM, in Univariate and Multivariate Logistic Regression Analyses
Univariate Odds Ratio (95% CI)
Adjusted Odds Ratio (95% CI) Age
10–12 y
13–15 y 0.74 (0.46–1.16)
16–19 y 0.80 (0.54–1.19)
Race White
Black 1.30 (0.64–2.66)
Hispanic 0.63 (0.27–1.47)
Asian 0.81 (0.28–2.37)
Native American/unknown/other 0.70 (0.28–1.79)
Female gender 0.58 (0.41–0.81) 0.59 (0.42–0.84)
Family history of T2DM 1.42 (1.01–2.01) 1.29 (0.91–1.84)
BMI category ⬍85th percentile
85th to 89th percentile 1.03 (0.45–2.35) NS
90th to 94th percentile 1.97 (1.10–3.51) 1.98 (1.10–3.53)
95th to 97th percentile 1.05 (0.46–2.39) NS
⬎97th percentile 2.57 (1.73–3.81) 2.40 (1.61–3.58)
FPG (0.2% vs 0.0%), and impaired glucose tolerance (3.1% vs 3.0%) were similar in the 2 studies. Although the 2004 study included all of the screening variables in the ADA criteria, including the presence of acanthosis nigricans (a driving factor in provider screening), it was limited by its role in a quality improvement initiative and its failure to account for nonrecommended testing. Our larger study offered a view of current screening practices and also allowed us to determine more-precise estimates of diabetes risk in a high-risk urban popula-tion.
Children meeting ADA criteria might not have been screened for a number of reasons. First, familiarity with ADA criteria among pediatric providers is limited, with less than one half of pediatricians screening routinely for diabetes.20 Therefore, providers may not be identifying
accurately patients who should be screened. Second, providers may be less apt to recommend diabetes screen-ing for overweight children because providers lack re-ferral resources for nutrition and exercise counseling. Third, it is possible that providers recommend screening for patients who ultimately do not receive screening because of nonadherence. Despite these barriers to ap-propriate screening in our cohort of providers, we think that the ADA consensus guidelines, if used appropri-ately, would allow identification of many additional pa-tients at especially high risk for T2DM and thus should be used systematically in the pediatric primary care set-ting.
Abnormal glucose metabolism was seen for 1.9% (n ⫽ 151) of the overall sample and for 9.2% of patients screened; 13 of those patients (0.8% of the overall sam-ple) met the criteria for diabetes. Although good popu-lation data are lacking in the literature, our prevalence seems to be higher than that in the Third National Health and Nutrition Examination Survey, in which type 1 diabetes mellitus and T2DM combined to account for 4.1 cases per 1000 children.10This is likely attributable to the
increased prevalences of overweight and minority status in our population, compared with national data.
This study has some important limitations. Like other retrospective chart reviews, our data were limited to information recorded in the chart. We were unable to extract information such as the presence of acanthosis nigricans from the physical examination results; there-fore, the number of children who actually met the ADA criteria was likely greater than indicated by our data. The limited observation period might have led to the exclu-sion of patients who were screened or diagnosed before or after our study period. Patients might have been screened for diabetes in other settings that were not recorded in our analyses. In addition, although we lim-ited our study to primary care visits, it is possible that tests included in our analyses might not have been re-lated to diabetes or might have been used for monitor-ing, as opposed to initial screening.
Population-based prevalence studies for T2DM in children were recommended by the International Dia-betes Federation Consensus Workshop,10particularly for
high-risk populations. Our data represent results for screened patients in such a population. Because screen-ing was not universal, our estimates of prevalence might not reflect the true prevalence of T2DM in this popula-tion. A better understanding of the prevalence will re-quire assessment of screening results for a high-risk pop-ulation after implementation of a comprehensive screening protocol or a true population-based preva-lence study, as recommended by expert committees.
In this study of urban children, screening rates for children at highest risk were substantially lower than recommended and many children who did not meet ADA criteria were screened. The EHR has the potential to improve screening by gathering more-comprehensive risk assessment data and providing decision support and prompting when indicated. Additional research regard-ing how to design and to implement EHR functionality effectively in these areas is needed.
The results of this study confirm the concerning prob-lem of overweight in urban children. Although few screened children had results consistent with T2DM, a considerable number had evidence of abnormal glucose metabolism. Although longitudinal population-based studies are needed to determine outcomes with cer-tainty, these findings may be an early warning sign for children at increased risk for developing T2DM as young adults. We think that better provider education regard-ing consensus recommendations, screenregard-ing strategies, and guidelines for interpretation of screening results are needed to allow for earlier provision of effective lifestyle interventions for children at especially high risk for T2DM.
ACKNOWLEDGMENTS
Financial support was provided by National Institutes of Health grants 1-K24-HD0424891-A2 and 2-T32-HP10014-10.
We thank Howard Bauchner, MD, for his mentorship and thoughtful review of the manuscript.
REFERENCES
1. Pinhas-Hamiel O, Dolan LM, Daniels SR, Standiford D, Khoury PR, Zeitler P. Increased incidence of non-insulin-dependent diabetes mellitus among adolescents. J Pediatr. 1996;128: 608 – 615
2. Fagot-Campagna A, Pettitt DJ, Engelgau MM, et al. Type 2 diabetes among North American children and adolescents: an epidemiologic review and a public health perspective.J Pediatr.
2000;136:664 – 672
3. Dabelea D, Hanson RL, Bennett PH, Roumain J, Knowler WC, Pettitt DJ. Increasing prevalence of type II diabetes in American Indian children.Diabetologia.1998;41:904 –910
5. Kitagawa T, Owada M, Urakami T, Yamauchi K. Increased incidence of non-insulin dependent diabetes mellitus among Japanese schoolchildren correlates with an increased intake of animal protein and fat.Clin Pediatr (Phila).1998;37:111–115 6. American Diabetes Association. Type 2 diabetes in children and
adolescents.Diabetes Care.2000;23:381–389
7. Lipton R, Keenan H, Onyemere KU, Freels S. Incidence and onset features of diabetes in African-American and Latino chil-dren in Chicago, 1985–1994.Diabetes Metab Res Rev.2002;18: 135–142
8. Crawford PB, Story M, Wang MC, Ritchie LD, Sabry ZI. Ethnic issues in the epidemiology of childhood obesity. Pediatr Clin North Am.2001;48:855– 878
9. Dabelea D, Pettitt DJ, Jones KL, Arslanian SA. Type 2 diabetes mellitus in minority children and adolescents: an emerging problem.Endocrinol Metab Clin North Am.1999;28:709 –729 10. Alberti G, Zimmet P, Shaw J, et al. Type 2 diabetes in the
young: the evolving epidemic: the International Diabetes Fed-eration Consensus Workshop. Diabetes Care. 2004;27: 1798 –1811
11. Kuczmarksi RJ, Ogden CL, Grummer-Strawn LM, et al. CDC growth charts: United States.Adv Data.2000;(314):1–27 12. American Diabetes Association. Standards of medical care in
diabetes.Diabetes Care.2004;27(suppl 1):S15–S35
13. Wollitzer A, Giammattei J, Jovanovic L. HbA1c as a potential screening test for childhood type 2 diabetes mellitus.Diabetes.
2001;50(suppl 2):A227
14. Peters AL, Davidson MB, Schriger DL, Hasselblad V. A clinical approach for the diagnosis of diabetes mellitus: an analysis using glycosylated hemoglobin levels: Meta-analysis Research Group on the Diagnosis of Diabetes Using Glycated Hemoglo-bin Levels.JAMA.1996;276:1246 –1252
15. Pettitt DJ, Giammattei J, Wollitzer AO, Jovanovic L. Glycohe-moglobin A1c distribution in school children: results from a school-based screening program.Diabetes Res Clin Pract.2004; 65:45– 49
16. Martin DD, Shephard MD, Freeman H, et al. Point-of-care testing of HbA1c and blood glucose in a remote Aboriginal Australian community.Med J Aust.2005;182:524 –527 17. Expert Committee on the Diagnosis and Classification of
Dia-betes Mellitus. Report of the Expert Committee on the Diag-nosis and Classification of Diabetes Mellitus. Diabetes Care.
1997;20:1183–1197
18. Rosenbloom AL, Joe JR, Young RS, Winter WE. Emerging epidemic of type 2 diabetes in youth.Diabetes Care.1999;22: 345–354
19. Drobac S, Brickman W, Smith T, Binns HJ. Evaluation of a type 2 diabetes screening protocol in an urban pediatric clinic.
Pediatrics.2004;114:141–148
DOI: 10.1542/peds.2006-0121
2006;118;1888
Pediatrics
Shikha G. Anand, Supriya D. Mehta and William G. Adams
Diabetes Mellitus Screening in Pediatric Primary Care
Services
Updated Information &
http://pediatrics.aappublications.org/content/118/5/1888 including high resolution figures, can be found at:
References
http://pediatrics.aappublications.org/content/118/5/1888#BIBL This article cites 17 articles, 4 of which you can access for free at:
Subspecialty Collections
http://www.aappublications.org/cgi/collection/endocrinology_sub Endocrinology
following collection(s):
This article, along with others on similar topics, appears in the
Permissions & Licensing
http://www.aappublications.org/site/misc/Permissions.xhtml in its entirety can be found online at:
Information about reproducing this article in parts (figures, tables) or
Reprints
DOI: 10.1542/peds.2006-0121
2006;118;1888
Pediatrics
Shikha G. Anand, Supriya D. Mehta and William G. Adams
Diabetes Mellitus Screening in Pediatric Primary Care
http://pediatrics.aappublications.org/content/118/5/1888
located on the World Wide Web at:
The online version of this article, along with updated information and services, is
by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.
