Abstract
This survey study uses National Health and Nutrition Examination Survey data to assess trends in prediabetes among US youths from 1999 through 2018.
The US Preventive Services Task Force recently released a recommendation on screening for prediabetes and type 2 diabetes among adults,1 but no recommendation has been issued for youths to date. A recent study2 estimated that among youths aged 12 to 19 years, approximately 1 in 5 had prediabetes, with large variations across sociodemographic characteristics. However, trends in the prevalence of prediabetes among youths and associated disparities by population subgroups over the past 2 decades have not been reported to our knowledge, and such information is important for future diabetes prevention. In this study, we assessed trends in prediabetes among US youths from 1999 through 2018.
Methods
This survey study used data from 10 cycles of the National Health and Nutrition Examination Survey (NHANES) from 1999-2000 through 2017-2018 and combined every 2 consecutive cycles to obtain sufficient sample sizes. We included youths aged 12 to 19 years who completed the interview and examination. NHANES is a series of cross-sectional surveys using a complex, multistage probability design to sample the civilian, noninstitutionalized population. The NHANES protocol was approved by the US Centers for Disease Control and Prevention National Center for Health Statistics Ethics Review Board, and all participants provided written informed consent, or assent was obtained from participants or their guardians, respectively. This study was exempt from the Mount Sinai institutional review board review because it used publicly deidentified data sets. Details on NHANES survey methods and analytic methods are documented elsewhere.3 This study followed the American Association for Public Opinion Research (AAPOR) reporting guideline.
Sociodemographic variables included sex, age, race and ethnicity (Hispanic, non-Hispanic Asian, non-Hispanic Black, and non-Hispanic White), parental educational level (less than high school, high school, and some college and above), income level (family income to poverty ratio <1.3, 1.3 to <3.0, and ≥3.0; adjusted for household size and based on poverty guidelines specific to the survey year), household food-security status (full, marginal, low, and very low; categorized based on the US Household Food Security Survey Module developed by the US Department of Agriculture4), and body mass index category (underweight or normal weight, overweight, and obesity; age- and sex-specific body mass index z scores were calculated using US Centers for Disease Control and Prevention reference data5). Information on race and ethnicity was collected by trained NHANES interviewers according to the fixed categories provided by the National Center for Health Statistics. Blood samples were obtained by trained phlebotomists according to a standardized protocol, and data were recorded directly into a computerized database. Prediabetes was defined as no recorded diagnosis of diabetes but a hemoglobin A1c level of 5.7% to 6.4% (to convert to proportion of total hemoglobin, multiply by 0.01) or a fasting plasma glucose level of 100 mg/dL to 125 mg/dL (to convert to millimoles per liter, multiply by 0.02586).2
Survey analysis procedures were used to account for sampling weights, stratification, and clustering in the NHANES complex sampling design to derive nationally representative estimates. Logistic regression was used to estimate trends by treating the survey cycle as a continuous variable. A survey-weighted Wald F statistic was used to test for the interaction. All analyses were performed using Stata, version 14 (StataCorp LLC). Significance was set at 2-sided P = .05. Data were analyzed from August to September 2021.
Results
A total of 6598 youths (mean [SD] age, 15.5 [2.76] years; 3412 [51.2% weighted] male) were included in this analysis (Table 1). The mean response rate was 79.2% (range, 59.3%-86.4%). Overall, the prevalence of prediabetes among US youths increased significantly from 11.6% (95% CI, 9.49%-14.1%) in 1999-2002 to 28.2% (95% CI, 23.3%-33.6%) in 2015-2018 (Table 2). The increasing trend was observed across population subgroups. Disparities in prevalence of prediabetes remained stable and were most pronounced in subgroup analyses of sex and body mass index category. For example, from 1999-2002 to 2015-2018, the prevalence of prediabetes increased from 15.8% (95% CI, 12.3%-20.1%) to 36.4% (95% CI, 30.1%-43.1%) among male youths and from 7.1% (95% CI, 5.1%-9.9%) to 19.6% (95% CI, 14.7%-25.7%) among female youths (P < .001 for trend). During the same period, the prevalence increased from 9.41% (95% CI, 7.50%-11.8%) to 24.3% (95% CI, 18.9%-30.7%) among youths with underweight or normal weight (P < .001 for trend), from 15.3% (95% CI, 9.45%-23.8%) to 27.5% (95% CI, 19.7%-36.9%) among youths with overweight (P = .005 for trend), and from 18.2% (95% CI, 12.8%-25.2%) to 40.4% (95% CI, 30.2%-51.5%) among youths with obesity (P < .001 for trend).
Table 1. Characteristics of US Youths Included in the Study by NHANES Cycles From 1999 Through 2018.
Characteristic | 1999-2002 (n = 1862) | 2003-2006 (n = 1788) | 2007-2010 (n = 998) | 2011-2014 (n = 1064) | 2015-2018 (n = 886) | |||||
---|---|---|---|---|---|---|---|---|---|---|
No. | % (95% CI)a | No. | % (95% CI)a | No. | % (95% CI)a | No. | % (95% CI)a | No. | % (95% CI)a | |
Age, y | ||||||||||
12-15 | 938 | 49.9 (46.7-53.1) | 882 | 48.7 (44.5-52.8) | 485 | 50.5 (46.4-54.6) | 544 | 50.5 (46.5-54.5) | 440 | 48.8 (44.4-53.3) |
16-19 | 924 | 50.1 (46.9-53.3) | 906 | 51.3 (47.2-55.5) | 513 | 49.5 (45.4-53.6) | 520 | 49.5 (45.5-53.5) | 446 | 51.2 (46.7-55.6) |
Sex | ||||||||||
Female | 903 | 48.8 (45.4-52.2) | 843 | 48.4 (44.8-52.0) | 450 | 49.0 (44.2-53.7) | 539 | 48.8 (44.9-52.8) | 451 | 49.2 (44.9-53.4) |
Male | 959 | 51.2 (47.8-54.6) | 945 | 51.6 (48.0-55.2) | 548 | 51.0 (46.3-55.8) | 525 | 51.2 (47.2-55.1) | 435 | 50.9 (46.6-55.1) |
Race and ethnicity | ||||||||||
Hispanic | 774 | 16.9 (13.0-21.8) | 623 | 15.7 (12.3-19.7) | 397 | 18.7 (14.0-24.4) | 325 | 21.9 (17.4-27.1) | 276 | 23.5 (17.9-30.2) |
Non-Hispanic | ||||||||||
Asian | NA | NA | NA | NA | NA | NA | 129 | 4.70 (3.23-6.78) | 93 | 5.01 (3.32-7.50) |
Black | 515 | 13.8 (10.5-18.1) | 628 | 14.8 (11.5-18.8) | 215 | 14.8 (11.9-18.4) | 295 | 15.1 (11.5-19.7) | 194 | 13.4 (10.1-17.7) |
White | 490 | 61.4 (56.4-66.2) | 461 | 63.8 (57.8-69.4) | 331 | 60.2 (53.4-66.7) | 269 | 54.4 (47.7-60.9) | 268 | 52.3 (45.7-58.9) |
Otherb | 83 | NC | 76 | NC | 55 | NC | 46 | NC | 55 | NC |
Parental educational levelc | ||||||||||
<High school | 706 | 21.3 (18.3-24.8) | 566 | 18.0 (15.3-21.0) | 287 | 18.2 (14.9-21.9) | 284 | 21.7 (18.2-25.8) | 186 | 18.4 (13.6-24.6) |
High school | 824 | 53.9 (48.4-59.3) | 918 | 59.3 (54.8-63.7) | 506 | 51.0 (46.1-55.8) | 556 | 53.1 (47.7-58.4) | 509 | 55.4 (48.6-62.0) |
≥Some college | 236 | 20.9 (16.5-26.3) | 225 | 18.5 (14.7-23.2) | 179 | 27.7 (21.1-35.5) | 200 | 23.2 (18.7-28.3) | 139 | 20.6 (15.1-27.5) |
Ratio of family income to poverty leveld | ||||||||||
<1.3 | 700 | 28.2 (23.9-33.0) | 698 | 28.5 (24.4-33.0) | 385 | 26.9 (23.4-30.8) | 478 | 34.4 (28.4-41.0) | 306 | 24.8 (20.4-29.8) |
1.3 to <3.0 | 518 | 27.7 (24.6-30.9) | 538 | 27.8 (24.5-31.0) | 284 | 27.2 (22.4-32.7) | 275 | 28.2 (22.8-34.3) | 289 | 32.0 (26.6-37.9) |
≥3.0 | 472 | 37.6 (33.2-42.1) | 466 | 39.5 (34.9-44.3) | 251 | 40.2 (34.1-46.7) | 234 | 31.4 (26.0-37.3) | 195 | 34.9 (29.7-40.3) |
Food securitye | ||||||||||
Very low | 142 | 5.87 (4.20-8.15) | 147 | 6.17 (4.69-8.08) | 76 | 4.70 (3.41-6.43) | 88 | 6.72 (4.51-9.90) | 95 | 8.80 (6.73-11.4) |
Low | 285 | 9.81 (7.88-12.1) | 273 | 9.66 (8.13-11.4) | 177 | 12.5 (10.3-15.1) | 179 | 13.9 (11.1-17.3) | 161 | 12.9 (10.5-15.7) |
Marginal | 153 | 5.82 (4.22-7.97) | 169 | 6.49 (4.87-8.58) | 136 | 9.08 (6.68-12.2) | 187 | 15.3 (12.4-18.6) | 118 | 11.0 (8.50-14.1) |
Full | 1207 | 74.6 (70.8-78.1) | 1143 | 74.4 (71.0-77.5) | 601 | 72.8 (67.9-77.3) | 604 | 62.8 (57.7-67.6) | 481 | 64.7 (59.7-69.4) |
BMI categoryf | ||||||||||
Underweight or normal weight | 1202 | 70.3 (67.5-73.0) | 1134 | 66.4 (62.4-70.2) | 614 | 65.5 (62.3-68.5) | 650 | 64.8 (60.2-69.0) | 512 | 60.1 (56.3-63.9) |
Overweight | 291 | 13.3 (11.2-15.9) | 291 | 15.5 (13.5-17.7) | 173 | 16.1 (13.3-19.5) | 160 | 14.2 (11.2-17.8) | 148 | 18.4 (15.6-21.6) |
Obese | 354 | 16.4 (14.3-18.6) | 350 | 18.1 (14.5-22.5) | 193 | 18.4 (15.3-21.9) | 226 | 21.0 (17.2-25.4) | 211 | 21.5 (18.4-24.9) |
Abbreviations: BMI, body mass index; NA, not available; NC, not calculated; NHANES, National Health and Nutrition Examination Survey.
Percentages were adjusted for NHANES survey weight.
Owing to the small sample size, other racial and ethnic groups were not included in the analysis. “Other” included groups indicated as “other” in NHANES, including multiracial.
Data missing for 277 individuals.
Data missing for 509 individuals.
Data missing for 176 individuals.
Data missing for 89 individuals.
Table 2. Trends in Weighted Prevalence of Prediabetes Overall and by Population Subgroups Among Youths by NHANES Cycles From 1999 Through 2018.
Subgroup | Survey-weighted % (95% CI)a | P value | ||||||
---|---|---|---|---|---|---|---|---|
1999-2002 (n = 1862) | 2003-2006 (n = 1788) | 2007-2010 (n = 998) | 2011-2014 (n = 1064) | 2015-2018 (n = 886) | Trend | Interaction | ||
Overall | 11.6 (9.49-14.1) | 15.4 (12.7-18.5) | 23.4 (20.0-27.4) | 22.7 (19.7-25.9) | 28.2 (23.3-33.6) | <.001 | NA | |
Age, y | ||||||||
12-15 | 13.1 (10.2-16.8) | 17.5 (13.8-22.1) | 24.6 (19.2-30.9) | 23.3 (19.3-27.9) | 30.8 (24.9-37.4) | <.001 | .65 | |
16-19 | 10.0 (7.7-13.0) | 13.4 (10.3-17.2) | 22.2 (19.1-25.6) | 21.9 (17.3-27.4) | 25.6 (20.8-31.2) | <.001 | ||
Sex | ||||||||
Female | 7.1 (5.1-9.9) | 8.7 (6.6-11.4) | 16.7 (13.1-21.0) | 14.3 (11.5-17.8) | 19.6 (14.7-25.7) | <.001 | .84 | |
Male | 15.8 (12.3-20.1) | 21.7 (18.2-25.7) | 29.8 (25.5-34.5) | 30.6 (25.4-36.3) | 36.4 (30.1-43.1) | <.001 | ||
Race and ethnicity | ||||||||
Hispanic | 13.4 (10.6-16.9) | 16.0 (11.8-21.4) | 30.6 (24.2-37.8) | 26.7 (22.5-31.2) | 28.6 (23.0-34.9) | <.001 | .56b | |
Non-Hispanic | ||||||||
Asian | NA | NA | NA | 23.5 (15.8-33.6) | 26.0 (15.5-40.2) | .75 | ||
Black | 11.7 (9.1-14.8) | 14.7 (11.9-18.0) | 26.8 (21.2-33.3) | 23.2 (17.2-30.6) | 32.3 (24.9-40.8) | <.001 | ||
White | 11.5 (8.4-15.5) | 15.3 (11.6-20.1) | 19.8 (15.2-25.4) | 20.6 (16.2-25.8) | 26.8 (19.1-36.2) | <.001 | ||
Parental educational level | ||||||||
<High school | 15.6 (12.8-18.9) | 15.9 (11.7-21.4) | 24.2 (17.6-32.2) | 29.9 (23.6-37.1) | 29.1 (23.9-34.9) | <.001 | .28 | |
High school | 10.0 (8.0-12.4) | 14.7 (12.3-17.4) | 24.0 (19.8-28.6) | 20.5 (17.0-24.6) | 29.2 (23.7-35.4) | <.001 | ||
≥Some college | 11.1 (6.3-18.8) | 17.0 (10.9-25.7) | 19.4 (12.6-28.7) | 20.4 (13.8-29.0) | 26.8 (15.7-42.0) | .03 | ||
Ratio of family income to poverty level | ||||||||
<1.3 | 12.5 (8.5-17.9) | 12.7 (9.1-17.3) | 29.4 (23.9-35.6) | 26.6 (22.3-31.4) | 31.2 (25.6-37.3) | <.001 | .54 | |
1.3 to <3.0 | 9.3 (6.4-13.3) | 19.8 (15.2-25.4) | 23.0 (17.1-30.0) | 20.6 (15.6-26.7) | 30.0 (23.2-37.7) | <.001 | ||
≥3.0 | 11.8 (8.8-15.7) | 13.8 (9.7-19.4) | 18.5 (13.0-25.7) | 20.8 (14.9-28.3) | 24.5 (15.2-37.0) | <.006 | ||
Food security | ||||||||
Very low | 18.4 (11.0-29.1) | 13.3 (6.83-24.2) | 40.7 (32.0-50.1) | 20.1 (12.1-31.7) | 26.3 (16.3-39.6) | .20 | .24 | |
Low | 12.0 (8.0-17.6) | 15.2 (10.5-21.6) | 26.0 (20.1-33.1) | 27.6 (19.4-37.7) | 29.6 (22.2-38.2) | <.001 | ||
Marginal | 11.4 (6.9-18.3) | 20.5 (13.9-29.3) | 24.9 (15.8-36.9) | 20.4 (13.8-29.1) | 45.5 (31.6-60.2) | <.001 | ||
Full | 10.8 (8.3-13.9) | 15.5 (12.2-19.4) | 21.7 (17.4-26.7) | 22.4 (18.2-27.3) | 25.6 (19.7-32.5) | <.001 | ||
BMI category | ||||||||
Underweight or normal weight | 9.41 (7.50-11.8) | 13.3 (10.3-17.0) | 23.0 (18.1-28.8) | 18.3 (15.5-21.4) | 24.3 (18.9-30.7) | <.001 | .94 | |
Overweight | 15.3 (9.45-23.8) | 14.4 (9.02-22.3) | 17.7 (12.4-24.6) | 26.7 (18.7-36.6) | 27.5 (19.7-36.9) | <.005 | ||
Obese | 18.2 (12.8-25.2) | 24.1 (16.0-34.6) | 29.5 (22.8-37.3) | 34.2 (24.5-45.4) | 40.4 (30.2-51.5) | <.001 |
Abbreviations: BMI, body mass index; NA, not applicable; NHANES, National Health and Nutrition Examination Survey.
Percentages were adjusted for NHANES survey weight.
The P value for interaction did not include data for Asian participants owing to inconsistent data across cycles.
Discussion
In this survey study, the prevalence of prediabetes increased significantly among US youths from 1999 to 2018. Several limitations should be noted. First, there was only 1 measure of blood biomarkers for prediabetes, and thus, seasonal variations were not accounted for in the analysis. Second, we did not use the oral glucose tolerance test to define prediabetes because the information was not available in some of the NHANES cycles; thus, our results may have underestimated the prevalence of prediabetes. Third, owing to the small sample sizes, the statistical power might not have been sufficient to detect an interaction.
References
- 1.Davidson KW, Barry MJ, Mangione CM, et al. ; US Preventive Services Task Force . Screening for prediabetes and type 2 diabetes: US Preventive Services Task Force recommendation statement. JAMA. 2021;326(8):736-743. doi: 10.1001/jama.2021.12531 [DOI] [PubMed] [Google Scholar]
- 2.Andes LJ, Cheng YJ, Rolka DB, Gregg EW, Imperatore G. Prevalence of prediabetes among adolescents and young adults in the United States, 2005-2016. JAMA Pediatr. 2020;174(2):e194498. doi: 10.1001/jamapediatrics.2019.4498 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.US Centers for Disease Control and Prevention; National Center for Health Statistics . NHANES survey methods and analytic guidelines. Accessed September 28, 2021. https://wwwn.cdc.gov/nchs/nhanes/analyticguidelines.aspx
- 4.Bickel G, Nord M, Price C, Hamilton W, Cook J. Guide to measuring household food security. Revised 2000. US Department of Agriculture, Food and Nutrition Service; March 2000. Accessed September 28, 2021. https://fns-prod.azureedge.net/sites/default/files/FSGuide.pdf [Google Scholar]
- 5.Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC growth charts for the United States: methods and development. Vital Health Stat 11. 2002;(246):1-190. [PubMed] [Google Scholar]