OpenMFM Clinician Education

One-Step vs. Two-Step
Diagnosing Gestational Diabetes

Two accepted strategies, two different diagnostic philosophies. ACOG favors one; the ADA favors the other. Both are correct — and that is exactly the problem worth understanding.

ACOG vs. ADAHAPO-derived thresholdsProvider & APP education

Why this isn't settled

No professional consensus exists — and both major U.S. societies say so

"There is no scientific consensus on how best to diagnose gestational diabetes. Expert professional organizations acknowledge two acceptable options: the IADPSG one-step screening approach (currently preferred by the American Diabetes Association) and the two-step Carpenter–Coustan screening approach (recommended by the American College of Obstetricians and Gynecologists)." Hillier TA, et al. A Pragmatic, Randomized Clinical Trial of Gestational Diabetes Screening. N Engl J Med. 2021.

This deck lays out the mechanics, thresholds, comparative outcomes, and practical tradeoffs of each strategy so the choice made at your institution is an informed one, not a default.

Two-step strategy — ACOG preferred

A nonfasting screen filters who needs the diagnostic test

Step 1: ScreenNonfasting 50g GCT at 24–28 weeks

Step 2: DiagnoseIf positive: fasting 3-hr 100g OGTT

130/135/140

1-hr GCT cutoff (mg/dL)

ACOG accepts any of the three; the choice is institutional.

~15–20%

Proceed to the OGTT

Only those who screen positive need the longer, fasting test.

2 visits

For a positive screen

A separate fasting visit is required to complete the diagnosis.

Two-step diagnostic thresholds

Carpenter-Coustan is preferred over the older NDDG criteria

Timepoint	Carpenter–Coustan	NDDG
Fasting	≥ 95 mg/dL	≥ 105 mg/dL
1 hour	≥ 180 mg/dL	≥ 190 mg/dL
2 hour	≥ 155 mg/dL	≥ 165 mg/dL
3 hour	≥ 140 mg/dL	≥ 145 mg/dL

Diagnosis: ≥ 2 of 4 values met or exceeded. The ADA recommends the lower Carpenter–Coustan thresholds if the two-step approach is used.

ACOG notes a single elevated value may be used for diagnosis instead of two — but this would significantly increase GDM incidence and is not the default.

One-step strategy — ADA preferred

A single fasting test completes screening and diagnosis at once

IADPSG criteria

Single fasting 75g OGTT

Performed once at 24–28 weeks. Any one of the following meets or exceeds threshold:

Fasting ≥ 92 mg/dL
1-hour ≥ 180 mg/dL
2-hour ≥ 153 mg/dL

Where the numbers come from

HAPO study thresholds

Set at the glucose level where the odds of adverse outcomes — LGA, neonatal hyperinsulinemia, excess adiposity — reached 1.75× estimated baseline odds.

These are the only GDM thresholds derived directly from pregnancy outcomes rather than prediction of later maternal diabetes.

All thresholds, one table

Glucose thresholds for diagnosing GDM after 24 weeks

Timepoint	Carpenter–Coustan (two-step)	NDDG (two-step)	IADPSG (one-step)
Fasting	≥ 95 mg/dL	≥ 105 mg/dL	≥ 92 mg/dL
1 hour	≥ 180 mg/dL	≥ 190 mg/dL	≥ 180 mg/dL
2 hour	≥ 155 mg/dL	≥ 165 mg/dL	≥ 153 mg/dL
3 hour	≥ 140 mg/dL	≥ 145 mg/dL	—
# values required	≥ 2 of 4	≥ 2 of 4	≥ 1 of 3

Sacks DB, et al. Guidelines and Recommendations for Laboratory Analysis in the Diagnosis and Management of Diabetes Mellitus. Diabetes Care. 2023. Table adapted per Will JS, Crellin H. Am Fam Physician. 2023.

Head-to-head evidence

The Hillier 2021 pragmatic RCT: nearly double the diagnoses, no measured benefit

23,792

women randomized

16.5%

GDM with one-step

8.5%

GDM with two-step

LGA infants

8.9% vs. 9.2% — not significant

Perinatal composite

3.1% vs. 3.0% — not significant

Gestational HTN / preeclampsia

13.6% vs. 13.5% — not significant

Primary cesarean

24.0% vs. 24.6% — not significant

One-step screening was associated with increased neonatal hypoglycemia — possibly reflecting heightened surveillance of infants born to mothers newly labeled with GDM, rather than a true biological difference.

Hillier TA, Pedula KL, Ogasawara KK, et al. A Pragmatic, Randomized Clinical Trial of Gestational Diabetes Screening. N Engl J Med. 2021.

Independent confirmation

The USPSTF systematic review found the same pattern

Prevalence

One-step: 11.5% · Two-step: 4.9%

Across the pooled evidence base, one-step screening consistently identifies roughly twice the number of GDM cases.

Outcomes

No demonstrated short-term benefit

Despite the higher diagnosis rate, the review found no clear evidence that one-step screening improves short-term maternal or neonatal health outcomes.

Two independent, high-quality evidence sources — one pragmatic RCT, one systematic review — converge on the same conclusion: more diagnoses does not automatically mean better measured outcomes at delivery.

Pillay J, Donovan L, Guitard S, et al. Screening for Gestational Diabetes. JAMA. 2021.

Two-step: risks & benefits

Fewer diagnostic tests, at the cost of a second visit

Benefits

Initial screen requires no fasting
Only ~15–20% need the 3-hour OGTT
Treating the higher-threshold hyperglycemia it identifies has proven benefit — less macrosomia, LGA, shoulder dystocia

Tradeoffs

Requires two visits for anyone who screens positive
Diagnostic test is longer (3 hours) with a higher glucose load (100g)
May miss milder hyperglycemia still linked to adverse outcomes

One-step: risks & benefits

One visit, broader detection — and a long-term signal worth knowing

Benefits

Screening and diagnosis complete in a single visit
Shorter test (2 hours), lower glucose load (75g)
Identifies a broader at-risk population
HAPO Follow-Up: this group had a 3.4× higher risk of prediabetes/type 2 diabetes 11 years later, plus higher offspring obesity

Tradeoffs

Requires fasting for every patient screened
Roughly doubles or triples the GDM diagnosis rate (to ~15–20%)
NIH consensus panel: concern for medicalization of pregnancy and increased healthcare utilization/cost without demonstrated short-term outcome improvement

Where the societies land

ACOG and ADA endorse different strategies for different reasons

ACOG

Two-step (Carpenter–Coustan)

Cites proven treatment benefit at the higher thresholds and concern about over-diagnosis and unproven benefit of treating milder hyperglycemia.

ADA

One-step (IADPSG)

Favors thresholds derived directly from HAPO pregnancy-outcome data, arguing these best reflect the glucose levels that actually matter for the baby.

Both organizations acknowledge the need for further evidence, and either approach is considered acceptable. There is no professional or medico-legal expectation to pick one over the other.

Alternatives to a formal OGTT

When a full OGTT isn't feasible, the alternatives are imperfect substitutes

HbA1c at 24–28 weeks

Does not perform as well as the glucose load test for GDM screening. Not recommended as a standalone screening test.

Self-monitored blood glucose

Fasting ≥ 95, 1-hr postprandial ≥ 140, 2-hr postprandial ≥ 120 mg/dL — suggested when an OGTT can't be performed, but no consensus on how many abnormal values are required.

Fasting plasma glucose alone

Used in some international settings (e.g., China: FPG ≥ 5.6 mmol/L as a first-trimester screen), but lacks the sensitivity of a full OGTT.

Early first-trimester screening

Recommended for women with risk factors, to identify pre-existing diabetes — managed as diabetes complicating pregnancy, not as GDM.

Choosing an approach in practice

The right choice depends on what your practice is optimizing for

Favor two-step if you value

Fewer diagnostic tests, higher specificity

Minimizing the number of patients labeled and treated, at higher, outcome-validated-by-treatment-trial thresholds.

Favor one-step if you value

Single-visit efficiency, maximal sensitivity

Capturing the broadest population linked to adverse pregnancy outcomes and later metabolic risk, in one encounter.

Practical rule: whichever strategy is adopted, apply it consistently, use the paired thresholds correctly (never mix Carpenter–Coustan cutoffs with IADPSG rules), and document the protocol in your practice's standing orders.

Clinical pearls

Ten things to remember about GDM screening strategy

Both strategies are guideline-endorsed: two-step by ACOG, one-step by ADA/IADPSG.
Thresholds are not interchangeable: never pair Carpenter–Coustan cutoffs with a one-value IADPSG rule or vice versa.
One-step roughly doubles the GDM diagnosis rate compared with two-step (11.5% vs. 4.9% in the USPSTF review).
The largest head-to-head RCT found no significant difference in LGA, perinatal composite outcomes, preeclampsia, or cesarean rate.
IADPSG thresholds are uniquely outcome-derived — set at a 1.75× odds ratio for adverse outcomes in the HAPO study.

Carpenter–Coustan/NDDG thresholds were derived to predict future maternal diabetes, not pregnancy outcomes directly.
The two-step 50g screen requires no fasting; the one-step 75g test requires universal fasting.
HAPO Follow-Up data links IADPSG-positive women to a 3.4-fold higher long-term T2DM risk — a metabolic signal, not a pregnancy-outcome one.
HbA1c alone is not an adequate GDM screening test.
Consistency matters more than which strategy you choose — document your practice's protocol and apply it uniformly.

Evidence & controversies

What's known, what's uncertain, and where experts still disagree

Known

One-step screening identifies roughly twice as many GDM cases as two-step, replicated across a pragmatic RCT and a systematic review.

Uncertain

Whether treating the additional, milder hyperglycemia identified only by one-step criteria improves clinically meaningful short-term outcomes.

Contested

Whether IADPSG's outcome-derived thresholds or Carpenter–Coustan's treatment-trial-validated thresholds are the more appropriate diagnostic anchor.

Research need

Long-term, prospective comparisons of maternal and offspring metabolic outcomes under each diagnostic strategy, beyond the pregnancy episode itself.

Both ADA and ACOG acknowledge the evidence gaps and consider either approach acceptable.

References

Selected references

American Diabetes Association Professional Practice Committee. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes—2026. Diabetes Care. 2026.
Hillier TA, Pedula KL, Ogasawara KK, et al. A Pragmatic, Randomized Clinical Trial of Gestational Diabetes Screening. N Engl J Med. 2021.
Sacks DB, Arnold M, Bakris GL, et al. Guidelines and Recommendations for Laboratory Analysis in the Diagnosis and Management of Diabetes Mellitus. Diabetes Care. 2023.
Committee on Practice Bulletins—Obstetrics. Gestational Diabetes Mellitus. ACOG Practice Bulletin No. 190. Obstet Gynecol. 2018.
Will JS, Crellin H. Gestational Diabetes Mellitus: Update on Screening, Diagnosis, and Management. Am Fam Physician. 2023.
Pillay J, Donovan L, Guitard S, et al. Screening for Gestational Diabetes. JAMA. 2021.
Hod M, Kapur A, McIntyre HD. Evidence in Support of the International Association of Diabetes in Pregnancy Study Groups' Criteria for Diagnosing Gestational Diabetes Mellitus Worldwide in 2019. Am J Obstet Gynecol. 2019.
Simmons D, Gupta Y, Hernandez TL, et al. Call to Action for a Life Course Approach. Lancet. 2024.

Educational material for physicians and advanced practice providers. Apply current institutional protocols and individualized clinical judgment.

One-Step vs. Two-StepDiagnosing Gestational Diabetes

No professional consensus exists — and both major U.S. societies say so

A nonfasting screen filters who needs the diagnostic test

1-hr GCT cutoff (mg/dL)

Proceed to the OGTT

For a positive screen

Carpenter-Coustan is preferred over the older NDDG criteria

A single fasting test completes screening and diagnosis at once

Single fasting 75g OGTT

HAPO study thresholds

Glucose thresholds for diagnosing GDM after 24 weeks

The Hillier 2021 pragmatic RCT: nearly double the diagnoses, no measured benefit

LGA infants

Perinatal composite

Gestational HTN / preeclampsia

Primary cesarean

The USPSTF systematic review found the same pattern

Fewer diagnostic tests, at the cost of a second visit

Benefits

Tradeoffs

One visit, broader detection — and a long-term signal worth knowing

Benefits

Tradeoffs

ACOG and ADA endorse different strategies for different reasons

Two-step (Carpenter–Coustan)

One-step (IADPSG)

When a full OGTT isn't feasible, the alternatives are imperfect substitutes

HbA1c at 24–28 weeks

Self-monitored blood glucose

Fasting plasma glucose alone

Early first-trimester screening

The right choice depends on what your practice is optimizing for

Ten things to remember about GDM screening strategy

What's known, what's uncertain, and where experts still disagree

Selected references

One-Step vs. Two-Step
Diagnosing Gestational Diabetes