The paper that broke the simple 'metformin works through AMPK' story. Metformin still suppressed hepatic glucose production in mice and hepatocytes lacking AMPK or LKB1, acting instead through a fall in hepatic energy charge. Any claim that metformin's benefit is AMPK-mediated has to answer this result.
Tier C because it is an animal intervention or observation study measuring an organismal outcome (model: Mouse (liver-specific AMPK-null and LKB1-null); primary hepatocytes); tier describes study design, not quality -- animal evidence can be rigorous and still sit below direct human data.
| Evidence tier | C Animal in vivo |
| Study type | 4 - Animal Study |
| Model system | Mouse (liver-specific AMPK-null and LKB1-null); primary hepatocytes |
| Journal | The Journal of Clinical Investigation |
| Year | 2010 |
| Peer reviewed | Yes |
| Record last updated | 2026-07-29 |
| Source | DOI 10.1172/JCI40671 · PMID 20577053 |
| Intervention | Metformin in AMPK-null / LKB1-null mouse liver |
| Target | Hepatic energy state (AMPK-independent) |
| Model | Mouse |
| Effect | Metformin suppresses hepatic gluconeogenesis without AMPK or LKB1, via reduced hepatic energy charge |
Barton, O. (2026). Metformin inhibits hepatic gluconeogenesis in mice independently of the LKB1/AMPK pathway via a decrease in hepatic energy state — evidence-graded record FOR2010. In Oliver's mTOR Atlas. https://mtor-atlas.org/study/FOR2010/ · Dataset DOI 10.5281/zenodo.22059963
@misc{atlas_FOR2010,
author = {Barton, Oliver},
title = {{Metformin inhibits hepatic gluconeogenesis in mice independently of the LKB1/AMPK pathway via a decrease in hepatic energy state} --- evidence-graded record FOR2010},
howpublished = {Oliver's mTOR Atlas},
year = {2026},
url = {https://mtor-atlas.org/study/FOR2010/},
note = {Dataset DOI: 10.5281/zenodo.22059963}
}