Egg and saturated fat containing breakfasts have no acute effect on acute glycemic control in healthy adults: a randomized partial crossover trial

Eggs are nutrient-dense and often linked to satiety and potential weight-loss benefits, with some interventional evidence suggesting improved glycemic control among individuals with diabetes, metabolic syndrome, or obesity, especially under calorie restriction. However, observational studies have associated higher egg intake with increased diabetes risk and adverse cardiovascular outcomes in people with diabetes, leading to ambiguity about eggs’ impact on glycemic control. A plausible confounder is that eggs are frequently consumed with saturated fat–rich foods (e.g., bacon, sausage), and higher saturated fat intake is associated with insulin resistance. Establishing eggs’ acute effects, with and without saturated fat, on glucose disposal is a necessary step to clarify causality and inform dietary guidance where tight glycemic control is critical. The study’s hypothesis was that any observed association between egg intake and poor glycemic control is driven by saturated fat typically consumed with eggs, rather than eggs per se.

Prior acute feeding trials often matched energy but not macronutrients, sometimes finding lower postprandial glycemia with egg-based or high-protein breakfasts compared to high-carbohydrate controls (e.g., bagel or pancake breakfasts), suggesting macronutrient composition may drive observed differences. Some long-term interventions (≥12 weeks), often within carbohydrate-restricted diets, report improvements in glycemic control and insulin sensitivity with egg consumption in at-risk populations. Conversely, epidemiological data have linked higher egg consumption with increased diabetes risk and, in people with diabetes, higher CVD risk, potentially confounded by dietary patterns rich in saturated fat. Short-term saturated fat–rich diets can induce insulin resistance, although the effects may depend on the type/source of saturated fat (e.g., dairy fat such as butter may not adversely affect glucose homeostasis and might be associated with slightly lower diabetes risk).

Design: Randomized partial crossover clinical trial with two study visits one week apart. Neither participants nor investigators were blinded. Participants: 48 adults (18–65 years; BMI 20–60 kg/m²; fasting glucose <126 mg/dL; habitual egg intake ≥1/week). Exclusions included diabetes, gestational diabetes history, antidiabetic medications, pregnancy/lactation, certain medical and psychiatric conditions, relevant medications, and egg allergy or low egg intake. Interventions: Four isocaloric, macronutrient-matched breakfasts: EB (egg breakfast), SB (saturated fat breakfast), ES (egg + saturated fat breakfast), and CB (control cereal-based breakfast). Each participant consumed two of the four breakfasts, with balanced assignment across 12 sequence combinations. Breakfasts were prepared according to standardized protocols. Macronutrient profiles (approximate): energy ~437–454 kcal; protein 19.9–22.6 g; fat 19.5–25.9 g; saturated fat 6.8–12.9 g; carbohydrates 38.7–52.8 g; fiber 6.4–8.0 g; GI ~54–66; GL ~16.7–30.0. Primary outcomes: Postprandial glucose and insulin measured at baseline (fasting) and 30, 60, 90, 120, and 180 minutes after breakfast. AUCs for glucose and insulin were computed. Statistical analysis: Conducted in R (v3.5.3). Descriptive statistics summarized sample characteristics. Missing and implausible values were multiply imputed (mice; 20 datasets; 50 iterations). Analyses involving missing data were pooled via a random-effects approach. Type I error was controlled using Holm–Bonferroni. Mixed-effects models (lme4, lmerTest) regressed AUCs on breakfast assignment (EB, SB, ES vs CB) controlling for baseline (fasting) value, with random effects to account for within-subject dependencies. The study was part of a larger project; power calculations were performed for the observational component rather than this acute trial.

- Sample: n=48 (45.8% male), mean age 25.8 ± 7.7 years; BMI 25.7 ± 4.6 kg/m². Baseline fasting glucose 95.8 ± 9.0 mg/dL; fasting insulin 8.5 ± 8.8 µIU/dL; HbA1c 5.3 ± 0.4%. - Primary outcomes: Compared with control breakfast (CB), there were no statistically significant differences in AUC for glucose or insulin for EB, SB, or ES. • Glucose AUC (mixed-effects, controlling for baseline): EB vs CB β=104.93 (SE 305.83), p=0.732; SB vs CB β=94.49 (SE 305.19), p=0.758; ES vs CB β=414.63 (SE 433.22), p=0.341. Baseline fasting glucose was a significant covariate (β=143.38, p<0.001). Reported Cohen’s d for glucose AUC comparisons: EB 0.072; SB 0.060; ES 0.200 (all p>0.05). • Insulin AUC (mixed-effects, controlling for baseline): EB vs CB β=−397.34 (SE 766.05), p=0.606; SB vs CB β=−461.76 (SE 740.43), p=0.535; ES vs CB β=594.81 (SE 1102.15), p=0.592. Baseline fasting insulin was a significant covariate (β=223.13, p=0.006). Reported Cohen’s d for insulin AUC comparisons: EB 0.108; SB 0.130; ES 0.113 (all p≥0.05). - Overall: No acute adverse effect of egg, saturated fat, or their combination on glucose disposal when breakfasts were matched for energy and macronutrients.

The study addressed whether eggs per se or saturated fat typically consumed with eggs acutely impair glycemic control. By tightly matching breakfasts for total energy and macronutrient composition and employing a partial crossover design, the results indicate that neither eggs nor saturated fat (alone or combined) alters postprandial glucose or insulin AUC compared to a control breakfast. These findings contrast with earlier acute studies that observed differences when macronutrient profiles were not matched, suggesting prior effects were likely driven by carbohydrate/protein content rather than egg presence. The null findings also suggest that the commonly presumed detrimental acute effect of saturated fat in the context of an egg breakfast may not manifest when meals are otherwise comparable, though the specific saturated fat source (butter) could matter. Given that participants were habitual egg consumers and breakfasts had comparable GI/GL within a mixed-meal context, the study supports that eggs and saturated fat do not acutely impair glycemic control under matched conditions. These results inform dietary counseling and suggest that concerns regarding acute glycemic effects of eggs (with or without saturated fat) may be overstated when macronutrient composition is controlled, while highlighting the need for longer-term, rigorously matched trials.

In healthy adults, isocaloric, macronutrient-matched breakfasts containing eggs, saturated fat, or their combination do not acutely affect glucose disposal compared to a similar control breakfast. The hypothesis that poor glycemic control associated with egg consumption is driven by co-consumed saturated fat was not supported in this acute setting. These findings are reassuring for continued egg consumption and should guide the design of longer-term, iso-caloric and macronutrient-matched studies to determine chronic effects on glycemic control.

- Underpowered for detecting small effects in the acute setting; post hoc power for differences in acute glycemic response was low (EB vs CB 5.7%, SB vs CB 5.5%, ES vs CB 10.4%). - Partial crossover rather than full crossover design reduced statistical power. - Convenience sample predominantly of college students, limiting generalizability. - Physical activity not assessed or controlled; could influence acute glucose disposal. - Medication use (e.g., statins, antihypertensives) and hormone therapy not considered in eligibility or analyses. - Breakfasts, while closely matched for energy and macronutrients, were not exactly matched for fiber, glycemic load/index, or energy density; manufactured food components (sweeteners, soy, fiber types) could have minor effects. - Saturated fat source was primarily butter; results may not generalize to other saturated fat sources. - Measurements were every 30 minutes; potential to miss earlier peaks, though curves suggested peaks around 30 minutes. - Single-site, unblinded study; short-term, acute assessment may not reflect chronic effects.