Andrea Glenn

Dr. Andrea Glenn

orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
2 December 2024 in The American Journal of Clinical Nutrition
Verified
Journal Article
Closed access
AGAndrea Glenn
Andrea Glenn
FWFenglei Wang
Fenglei Wang
ATAnne-Julie Tessier
Anne-Julie Tessier
JMJoAnn E Manson
JoAnn E Manson
EREric B Rimm
Eric B Rimm
FHFrank B Hu
Frank B Hu

22 April 2025 in European Journal of Clinical Nutrition
Verified
Journal Article
Closed access
PDF file
AGAndrea Glenn
Andrea Glenn
ATAnne-Julie Tessier
Anne-Julie Tessier
MKMeaghan E. Kavanagh
Meaghan E. Kavanagh
GMGloria A. Morgan
Gloria A. Morgan
CCClary B. Clish
Clary B. Clish
John SievenpiperJohn Sievenpiper
John Sievenpiper

2 September 2025 in The Journal of Nutrition
Verified
Journal Article
Closed access
AGAndrea Glenn
Andrea Glenn
FHFrank B Hu
Frank B Hu

28 March 2025 in JMIR Cardio
Verified
Journal Article
Open access
MKMeaghan E Kavanagh
Meaghan E Kavanagh
Laura ChiavaroliLaura Chiavaroli
Laura Chiavaroli
SQSelina M Quibrantar
Selina M Quibrantar
GVGabrielle Viscardi
Gabrielle Viscardi
KRKimberly Ramboanga
Kimberly Ramboanga
John SievenpiperJohn Sievenpiper
John Sievenpiper
Background: The Portfolio Diet is a dietary pattern for cardiovascular disease (CVD) risk reduction with 5 key categories including nuts and seeds; plant protein from specific food sources; viscous fiber sources; plant sterols; and plant-derived monounsaturated fatty acid sources. To enhance implementation of the Portfolio Diet, we developed the PortfolioDiet.app, an automated, web-based, multicomponent, patient-facing health app that was developed with psychological theory. Objective: We aimed to evaluate the effect of the PortfolioDiet.app on dietary adherence and its acceptability among adults with a high risk of CVD over 12 weeks. Methods: Potential participants with evidence of atherosclerosis and a minimum of one additional CVD risk factor in an ongoing trial were invited to participate in a remote web-based ancillary study by email. Eligible participants were randomized in a 1:1 ratio using a concealed computer-generated allocation sequence to the PortfolioDiet.app group or a control group for 12 weeks. Adherence to the Portfolio Diet was assessed by weighed 7-day diet records at baseline and 12 weeks using the clinical Portfolio Diet Score, ranging from 0 to 25. Acceptability of the app was evaluated using a multifaceted approach, including usability through the System Usability Scale ranging from 0 to 100, with a score >70 being considered acceptable, and a qualitative analysis of open-ended questions using NVivo 12. Results: In total, 41 participants were invited from the main trial to join the ancillary study by email, of which 15 agreed, and 14 were randomized (8 in the intervention group and 6 in the control group) and completed the ancillary study. At baseline, adherence to the Portfolio Diet was high in both groups with a mean clinical Portfolio Diet Score of 13.2 (SD 3.7; 13.2/25, 53%) and 13.7 (SD 5.8; 13.7/25, 55%) in the app and control groups, respectively. After the 12 weeks, there was a tendency for a mean increase in adherence to the Portfolio Diet by 1.25 (SD 2.8; 1.25/25, 5%) and 0.19 (SD 4.4; 0.19/25, 0.8%) points in the app and control group, respectively, with no difference between groups (P=.62). Participants used the app on average for 18 (SD 14) days per month and rated the app as usable (System Usability Scale of mean 80.9, SD 17.3). Qualitative analyses identified 4 main themes (user engagement, usability, external factors, and added components), which complemented the quantitative data obtained. Conclusions: Although adherence was higher for the PortfolioDiet.app group, no difference in adherence was found between the groups in this small ancillary study. However, this study demonstrates that the PortfolioDiet.app is considered usable by high-risk adults and may reinforce dietitian advice to follow the Portfolio Diet when it is a part of a trial for CVD management. Trial Registration: ClinicalTrials.gov NCT02481466; https://clinicaltrials.gov/study/NCT02481466

29 July 2026 in bioRxiv
Verified
Preprint
Open access
PDF file
SSSam Sterling
Sam Sterling
LBLauren T. Berube
Lauren T. Berube
AGAndrea Glenn
Andrea Glenn
ASAasma Shaukat
Aasma Shaukat
SBSouptik Barua
Souptik Barua
ATAristotelis Tsirigos
Aristotelis Tsirigos
Background: Dietary assessment is the cornerstone of clinical management and research studies evaluating diet and health. Traditional methods such as food diaries and 24-hour recalls can be burdensome, prone to recall bias, and difficult to adhere to. Image-based dietary assessment using vision-language models (VLMs) offers a potential solution. Objective: Our goal was to benchmark state-of-the-art VLMs for automated food recognition, weight estimation, and calorie estimation using Google’s Nutrition5k dataset. Methods: We evaluated 3,229 food images using ten approaches: proprietary VLMs (Gemini 2.0 Flash, 2.5 Flash, 3.0 Flash, and 3.1 Flash Lite; GPT- 4o, GPT-4o-mini, and GPT-5 Mini; and Claude Haiku 4.5), an open-source VLM (Qwen2-VL-7B), and a commercial food recognition API (FatSecret). We assessed calorie and weight estimation using Lin’s Concordance Correlation Coefficient (CCC) and component detection using Jaccard similarity. Results: Gemini 3.0 Flash achieved the best calorie estimation (CCC 0.767, MAE 80.7 kcal), while Gemini 3.1 Flash Lite offered very comparable accuracy (CCC 0.754) with the highest ingredient recognition (Jaccard 0.655) at the lowest cost among top-performing models ($0.59/1K images). Among earlier-generation models, Gemini 2.0 Flash remained competitive (CCC 0.742, Jaccard 0.621) at a fraction of the cost ($0.10/1K images). A human validation study in which four annotators reviewed 440 images revealed systematic omissions in the original Nutrition5k labels. After correction, the extrapolated ingredient-overlap score for Gemini 2.0 Flash increased from 0.62 to an estimated 0.82, suggesting that raw Jaccard scores substantially underestimate true model performance. Conclusions: Current VLMs can perform automated dietary assessment with reasonable accuracy from single overhead photographs. Our results inform model selection for dietary assessment applications and highlight remaining challenges in calorie estimation and component detection for complex, multi-item meals.