Continuous Glucose Monitoring (CGM) & Glycemic Variability - References

Continuous Glucose Monitoring (CGM) & Glycemic Variability — Literature Dossier

Navigation

• Backed Topic Note: Continuous Glucose Monitoring (CGM) & Glycemic Variability
• References Hub: 03 - Diagnostics & Quantified Self References MOC
• Parent Silo MOC: 03 - Diagnostics & Quantified Self MOC

---

Landmark Peer-Reviewed Studies

1. Shah et al. (2019) — Continuous Glucose Monitoring Profiles in Healthy Nondiabetic Participants

• Citation: Shah, V. N., DuBose, S. N., Li, Z., Beck, R. W., Peters, A. L., Weinstock, R. S., … & Bergenstal, R. M. (2019). Continuous glucose monitoring profiles in healthy nondiabetic participants: a multicenter prospective study. The Journal of Clinical Endocrinology & Metabolism, 104(10), 4356–4364.
• Identifiers: DOI: 10.1210/jc.2018-02763 (DOI · SciDB · Scholar) | PMID: 31127824 (PubMed · Europe PMC) | PMCID: PMC7362903 (PMC Full Text)
• Study Design & Cohort: Multicenter prospective study across 12 US centers ($N=153$ healthy non-diabetic individuals aged $7–80\text{ years}$ with baseline HbA1c $<5.7\%$ wearing blinded Dexcom G6 Pro CGMs for up to 10 days).
• Quantitative Findings & Diagnostic Thresholds:
  • Mean 24-hour glucose was $98–99\text{ mg/dL}$ ($93\text{ mg/dL}$ in youth/young adults, $104\text{ mg/dL}$ in adults $\ge 60$).
  • Median Time in Range ($70–140\text{ mg/dL}$) was $96\%$ ($93–98\%$ IQR); median time $>140\text{ mg/dL}$ was only $2.1\%$ ($30\text{ min/day}$); mean Standard Deviation (SD) was $17\text{ mg/dL}$ and Coefficient of Variation (CV) was $17\pm 3\%$.

2. Monnier et al. (2006) — Activation of Oxidative Stress by Acute Glucose Fluctuations Compared With Sustained Chronic Hyperglycemia

• Citation: Monnier, L., Mas, E., Ginet, C., Michel, F., Villon, L., Cristol, J. P., & Colette, C. (2006). Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemia in patients with type 2 diabetes. JAMA, 295(14), 1681–1687.
• Identifiers: DOI: 10.1001/jama.295.14.1681 (DOI · SciDB · Scholar) | PMID: 16609090 (PubMed · Europe PMC)
• Study Design & Cohort: Case-control CGM and lipid peroxidation study ($N=42$: 21 patients monitored by 48-hour CGM and 21 healthy controls; oxidative stress quantified via 24h urinary excretion of 8-iso-prostaglandin ${\text{F}}_{2\alpha }$ [8-iso-${\text{PGF}}_{2\alpha }$] by LC-MS/MS).
• Quantitative Findings & Diagnostic Thresholds:
  • Urinary 8-iso-${\text{PGF}}_{2\alpha }$ excretion showed an exceptionally strong linear correlation with the Mean Amplitude of Glycemic Excursions (MAGE: $r=0.86,p<0.001$) and postprandial incremental AUC ($r=0.55,p=0.009$), whereas correlation with HbA1c or mean 24-hour glucose was non-significant in multivariable models, proving that acute glycemic swings trigger far greater vascular oxidative stress than chronic steady glucose.

3. Hall et al. (2018) — Glucotypes Reveal New Patterns of Glucose Dysregulation

• Citation: Hall, H., Perelman, D., Breschi, A., Limcaoco, P., Kellogg, R., McLaughlin, T., & Snyder, M. (2018). Glucotypes reveal new patterns of glucose dysregulation. PLoS Biology, 16(7), e2005143.
• Identifiers: DOI: 10.1371/journal.pbio.2005143 (DOI · SciDB · Scholar) | PMID: 30040822 (PubMed · Europe PMC) | PMCID: PMC6057684 (PMC Full Text)
• Study Design & Cohort: Stanford longitudinal multi-omic CGM study ($N=57$ participants without prior diabetes diagnosis undergoing 2–4 weeks of CGM alongside standardized meal challenges and gold-standard steady-state plasma glucose [SSPG] insulin resistance testing).
• Quantitative Findings & Diagnostic Thresholds:
  • Spectral clustering classified individuals into Low (Glucotype A), Moderate (Glucotype B), and Severe (Glucotype C) glycemic variability phenotypes.
  • 25% of individuals classified as completely normoglycemic by standard fasting glucose ($<100\text{ mg/dL}$) and HbA1c ($<5.7\%$) exhibited the Severe Glucotype, reaching prediabetic ($>140\text{ mg/dL}$) $15\%$ of the time and diabetic ($>200\text{ mg/dL}$) $2\%$ of the time, which correlated directly with true insulin resistance on SSPG testing.

4. Zeevi et al. (2015) — Personalized Nutrition by Prediction of Glycemic Responses

• Citation: Zeevi, D., Korem, T., Zmora, N., Israeli, D., Rothschild, D., Weinberger, A., … & Segal, E. (2015). Personalized nutrition by prediction of glycemic responses. Cell, 163(5), 1079–1094.
• Identifiers: DOI: 10.1016/j.cell.2015.11.001 (DOI · SciDB · Scholar) | PMID: 26590418 (PubMed · Europe PMC)
• Study Design & Cohort: Prospective 800-person cohort monitored continuously for 7 days ($46,898$ total meals profiled via CGM, shotgun gut metagenomics, anthropometrics, and blood tests), followed by a blinded randomized dietary intervention ($N=26$).
• Quantitative Findings & Diagnostic Thresholds:
  • Demonstrated high interpersonal variability in postprandial glycemic response (PPGR) to identical standardized meals ($p<0.0001$); machine-learning integration of microbiome and meal features predicted individual PPGRs with $R=0.70$ and reduced postprandial glucose spikes and MAGE by $>35\%$ ($p<0.001$) in the randomized intervention arm.

5. Shukla et al. (2015) — Food Order Has a Significant Impact on Postprandial Glucose and Insulin Levels

• Citation: Shukla, A. P., Iliescu, R. G., Thomas, C. E., & Aronne, L. J. (2015). Food order has a significant impact on postprandial glucose and insulin levels. Diabetes Care, 38(7), e98–e99.
• Identifiers: DOI: 10.2337/dc15-0429 (DOI · SciDB · Scholar) | PMID: 26106234 (PubMed · Europe PMC) | PMCID: PMC4876745 (PMC Full Text)
• Study Design & Cohort: Within-subject crossover trial comparing identical isocaloric meals ($628\text{ kcal}$: $55\text{ g}$ protein, $68\text{ g}$ carbohydrate, $16\text{ g}$ fat) consumed carbohydrates-first vs. protein-and-vegetables-first (15 minutes prior to carbohydrates).
• Quantitative Findings & Diagnostic Thresholds:
  • Consuming protein and vegetables before carbohydrates reduced postprandial glucose levels by 28.6% at 30 min ($p=0.001$), 36.7% at 60 min ($p=0.001$), and 16.8% at 120 min ($p=0.003$), and reduced the incremental area under the glucose curve (${\text{iAUC}}_{0–120}$) by 73% ($p=0.001$) and insulin ${\text{iAUC}}_{0–120}$ by 48% ($p=0.003$).