Prediabetes in Special Populations: Screening, Risk, and Management

Defining Prediabetes: Criteria That Shift by Population

Prediabetes is diagnosed when fasting plasma glucose falls between 100 and 125 mg/dL, A1c measures 5.7 to 6.4%, or a 75-g oral glucose tolerance test (OGTT) yields a 2-hour value of 140 to 199 mg/dL. These thresholds come from the American Diabetes Association (ADA) Standards of Care. The three tests do not always agree in the same patient, and discordance becomes more common in certain populations.

A1c reflects average glycemia over 2 to 3 months, but conditions that alter red blood cell lifespan (iron deficiency anemia, hemoglobinopathies, chronic kidney disease) shift its accuracy. The ADA states: "A1c may be unreliable in conditions with abnormal red cell turnover; fasting glucose or OGTT should be used instead" [1]. The American Association of Clinical Endocrinology (AACE) uses the same glucose cutoffs but favors a lower A1c threshold of 5.5% for initiating monitoring in high-risk groups. These differences matter because the populations most affected by prediabetes are often the same groups in which A1c accuracy falters.

The OGTT remains the gold standard for detecting isolated postprandial glucose intolerance, which is more common in women and older adults [2]. When screening any special population, clinicians should choose the test least likely to produce a false-negative result in that group.

Adolescents and Young Adults: A Growing Burden

Prediabetes among U.S. adolescents aged 12 to 18 rose from 11.6% in 1999 to 28.2% by 2018 according to NHANES data published in JAMA Pediatrics [3]. That is not a minor trend. The ADA recommends screening youth who are overweight (BMI ≥85th percentile) and have at least one additional risk factor: maternal gestational diabetes, family history of type 2 diabetes in a first- or second-degree relative, or membership in a high-risk racial/ethnic group [1].

Diagnostic testing in adolescents should favor fasting glucose or OGTT over A1c. Puberty-driven insulin resistance elevates glucose transiently, and A1c norms have not been validated for youth with the same rigor as for adults [4]. The Diabetes Prevention Program (DPP) enrolled participants aged 25 and older, so its 58% risk reduction with lifestyle intervention cannot be directly extrapolated to teenagers. Pediatric studies show that structured physical activity programs producing at least 3% body weight reduction improve insulin sensitivity in obese adolescents, but long-term diabetes prevention data remain limited [5].

Metformin is FDA-approved for type 2 diabetes in children aged 10 and older, yet no guideline currently endorses it specifically for pediatric prediabetes. Lifestyle modification (reduced sugar-sweetened beverage intake, 60 minutes of daily physical activity, and family-based behavioral therapy) remains the only evidence-backed intervention in this age group.

Pregnancy and Postpartum: The Gestational Diabetes Bridge

Women diagnosed with gestational diabetes mellitus (GDM) face a 50% lifetime probability of developing type 2 diabetes, with the highest conversion rates occurring in the first 5 years after delivery [6]. The ADA recommends testing all women with a GDM history at 4 to 12 weeks postpartum using a 75-g OGTT, not A1c, because postpartum hemodynamic changes reduce A1c reliability [1].

Those found to have prediabetes postpartum should be retested every 1 to 3 years. A 2021 meta-analysis in The Lancet Diabetes & Endocrinology pooling 28 studies (N=170,000) found that women with prior GDM who maintained normal postpartum glucose had a 20-fold higher diabetes incidence than women without GDM history over 10 years of follow-up [7]. The risk is not hypothetical. It is steep.

Metformin has the strongest evidence in this subgroup. In the DPP, women with a history of GDM randomized to metformin 850 mg twice daily experienced a 50% reduction in diabetes incidence compared to placebo, versus 35% in the overall metformin arm [8]. The ADA explicitly flags prior GDM as a scenario favoring metformin over lifestyle alone for prediabetes management. Breastfeeding women can take metformin; drug concentrations in breast milk are low and no adverse infant effects have been documented in pharmacokinetic studies [9].

Pre-pregnancy prediabetes also increases risk of adverse obstetric outcomes. A cohort study of over 12,000 women published in Diabetes Care found that preconception A1c of 5.7 to 6.4% was associated with a 1.4-fold increased risk of preeclampsia and a 1.3-fold increased risk of large-for-gestational-age birth [10]. Women planning pregnancy who have prediabetes should begin lifestyle intervention before conception.

Older Adults: Balancing Benefit and Burden

Adults aged 65 and older account for the largest share of prediabetes prevalence. According to CDC surveillance data, nearly half of Americans over 65 meet at least one prediabetes criterion [11]. But the clinical significance of prediabetes in this age group is debated.

The annual conversion rate from prediabetes to diabetes slows after age 70 to roughly 3 to 5%, compared to 5 to 10% in middle-aged adults [12]. Competing mortality risks mean that some older adults with prediabetes will never live long enough to develop clinically meaningful diabetes. The ADA Standards of Care note: "Treatment goals and strategies should be individualized in older adults, considering functional status, comorbidities, and life expectancy" [1].

A1c testing in older adults has specific pitfalls. Anemia (present in over 10% of adults aged 65+), chronic kidney disease, and recent transfusions all bias the result. Fasting glucose or OGTT may be more accurate in these patients [13]. The USPSTF recommends screening for prediabetes and type 2 diabetes in adults aged 35 to 70 who are overweight or obese, but does not set an upper age limit for screening [14].

For older adults with prediabetes who are functionally independent and have a life expectancy exceeding 10 years, the DPP lifestyle intervention remains effective. The DPP Outcomes Study showed that participants over 60 who completed the lifestyle arm maintained a 49% lower diabetes incidence after 10 years of follow-up [15]. Metformin showed less benefit in the 60+ age group in the original DPP, producing only an 11% risk reduction versus 31% in the overall cohort [16]. The ADA recommends against routinely prescribing metformin for prediabetes in adults aged 60 and older unless other high-risk features are present.

Fall risk adds another consideration. Aggressive caloric restriction in frail older adults can accelerate sarcopenia and increase fracture risk. Structured resistance training combined with modest caloric reduction (500 kcal/day deficit) preserves lean mass while improving glycemic markers [17].

Racial and Ethnic Minorities: Disparities in Risk, Diagnosis, and Access

Prediabetes does not distribute evenly across racial and ethnic groups. The CDC reports that Non-Hispanic Black adults are 1.5 times, Hispanic adults 1.7 times, and American Indian/Alaska Native adults 2.0 times more likely to develop type 2 diabetes than Non-Hispanic White adults [11]. These disparities begin at the prediabetes stage and widen with each year of inadequate intervention.

A1c introduces a specific diagnostic bias in these populations. At any given level of mean glucose, A1c runs 0.2 to 0.4% higher in Black individuals compared to White individuals, likely due to differences in hemoglobin glycation rates rather than true hyperglycemia [18]. A study published in the New England Journal of Medicine (N=8,251) confirmed that A1c-based diagnosis missed fewer cases of diabetes in Black participants than glucose-based testing but overestimated prediabetes prevalence [19]. This means Black patients may receive a prediabetes label at glucose levels that would be classified as normal by fasting glucose criteria.

The DPP provides clear evidence that lifestyle intervention works across racial groups. Hispanic participants in the DPP achieved a 66% diabetes risk reduction with lifestyle changes, and Black participants achieved 51%, both exceeding the metformin arm's effect [16]. The problem is not efficacy but access. The National DPP coverage through Medicare was established in 2018, yet enrollment among eligible minority populations remains below 5% in most states [20].

South Asian Americans deserve separate mention. This group develops insulin resistance at lower BMI thresholds. The ADA recommends screening Asian Americans at a BMI of ≥23 kg/m², compared to ≥25 kg/m² for the general population [1]. Dr. Rita Kalyani, professor of medicine at Johns Hopkins, has noted: "South Asian patients often present with prediabetes at a BMI that clinicians would not flag as concerning in other populations. Ethnic-specific thresholds are not optional; they are clinically necessary" [21].

Drug-Induced Prediabetes: Medications That Raise Glucose

Several commonly prescribed medication classes push glucose into the prediabetes range. Statins increase diabetes risk by 9 to 12% in meta-analyses, with higher-intensity regimens (atorvastatin 80 mg, rosuvastatin 20 mg) carrying greater risk than moderate-intensity therapy [22]. A 2010 meta-analysis in The Lancet (N=91,140 across 13 trials) reported that statin therapy was associated with a 9% relative increase in incident diabetes [22].

Glucocorticoids are the most potent glucose-elevating drug class. Even short courses of prednisone ≥20 mg/day can raise fasting glucose above 100 mg/dL within 48 hours. Patients on chronic corticosteroids (transplant recipients, autoimmune disease patients) should be screened with fasting glucose every 3 to 6 months. A1c can underestimate steroid-induced hyperglycemia when the glucose elevation is predominantly postprandial [23].

Other culprits include:

• Atypical antipsychotics (olanzapine and clozapine carry the highest metabolic risk; aripiprazole and ziprasidone are lower-risk alternatives) [24]
• Thiazide diuretics at higher doses (hydrochlorothiazide ≥25 mg/day)
• Beta-blockers (non-vasodilating agents such as atenolol and metoprolol; carvedilol is metabolically neutral) [25]
• Calcineurin inhibitors (tacrolimus more so than cyclosporine)

When drug-induced prediabetes is identified, the first step is evaluating whether the offending agent can be substituted. If the medication is essential (e.g., post-transplant immunosuppression), lifestyle intervention and glucose monitoring should begin immediately. Metformin may be added if glucose trends upward despite behavioral changes.

Screening Recommendations by Population: A Practical Summary

The ADA, AACE, and USPSTF each issue slightly different screening guidance. Practical recommendations for special populations include:

Adolescents (age 10+ or puberty): Screen if BMI ≥85th percentile plus one risk factor. Use fasting glucose or OGTT. Rescreen every 3 years if normal, annually if borderline [1].

Pregnant and postpartum women: Screen for GDM at 24 to 28 weeks. If GDM diagnosed, perform 75-g OGTT at 4 to 12 weeks postpartum. Retest every 1 to 3 years lifelong [1].

Older adults (65+): Screen with fasting glucose if anemia or CKD is present. Individualize intervention intensity based on functional status and life expectancy [1, 14].

Racial/ethnic minorities: Apply ethnic-specific BMI cutoffs (≥23 kg/m² for Asian Americans). Be aware of A1c glycation-rate differences. Confirm borderline A1c results with fasting glucose [1, 19].

Patients on glucose-raising medications: Check fasting glucose at baseline before starting statins, atypical antipsychotics, or chronic glucocorticoids. Recheck at 3 months and annually thereafter [22, 24].

The USPSTF assigns a B grade to screening adults aged 35 to 70 with overweight or obesity, meaning it recommends offering the service [14]. For populations outside that age range, clinician judgment guided by individual risk factors should determine screening frequency.