The Influence of Metformin on Weight

Type 2 Diabetes Mellitus and Obesity

Type 2 diabetes mellitus has been growing in epidemic proportions worldwide, closely linked to the increasing prevalence of obesity. About 80-90% of patients with type 2 diabetes are overweight or obese (1). These diseases are so interconnected that since the 1970s, literature has referred to them with a single term: 'diabesity' (from Diabetes and obesity) (1). In patients with diabetes, obesity increases insulin resistance and hyperglycemia, making the treatment of diabetes inseparable from lifestyle modification recommendations aimed at weight reduction (1). Hypoglycemic drugs should stabilize or ideally reduce weight instead of increasing it (1).

Importance of Weight Reduction

Weight loss increases insulin sensitivity, lowers blood pressure, improves endothelial function, and reduces inflammatory and procoagulant markers, cholesterol, and triglyceride concentrations in the blood (1). However, not all antidiabetic drugs lead to weight reduction. Sulfonylureas (SU), thiazolidinediones (TZD), and meglitinides contribute to weight gain, while dipeptidyl peptidase-4 (DPP-4) inhibitors do not impact weight (1). Glucagon-like peptide 1 (GLP-1) receptor agonists and sodium-glucose cotransporter 2 (SGLT-2) inhibitors promote weight loss (1, 2). The scientific literature describes the impact of metformin on weight ambiguously. Global diabetes treatment recommendations also vary: the 2012 ADA and EASD guidelines state that metformin has a neutral effect on weight (2), while the 2013 AACE recommendations suggest a weight-reducing effect of metformin (3). This article reviews the mechanism of action of metformin concerning its effect on weight, clinical trial results, and practical principles of metformin use.

Mechanism of Action of Metformin

Metformin, one of the oldest and most widely used antidiabetic drugs, still does not have a fully understood molecular action. Several hypotheses exist about metformin's mechanism of action. Researchers believe that metformin inhibits the mitochondrial respiratory chain (complex I), activates AMP-activated protein kinase (AMPK), and inhibits glucagon-induced cyclic adenosine monophosphate (cAMP) elevation and protein kinase A (PKA) activation (4). By increasing insulin sensitivity, metformin reduces the frequency of postprandial hypoglycemia, common in diabetic patients due to delayed insulin secretion and a later insulin peak (5). Less postprandial hypoglycemia leads to fewer cravings for carbohydrates and fewer episodes of overeating, particularly of carbohydrates (6). Additionally, metformin inhibits gluconeogenesis, reducing glucose production in the liver. In diabetic patients, the liver's gluconeogenesis is on average three times more intense than in healthy individuals, and metformin reduces this by one-third (7). Metformin also inhibits glucose absorption in the gastrointestinal tract, decreasing glucose supply for energy storage in adipose tissue (7). Unlike hypocaloric diets, metformin-induced weight loss occurs primarily through the reduction of adipose tissue rather than muscle tissue (7).

By increasing peripheral tissue sensitivity to insulin, metformin lowers insulin secretion in the pancreas, decreasing insulin exposure of receptors in abdominal adipose tissue and possibly reducing energy absorption and abdominal adipose tissue mass (8). Researchers found that metformin reduces weight in obese patients without insulin resistance (normal HOMA or Matsuda indices) and/or with normal glycemia (8). Thus, other mechanisms might contribute to weight loss in patients treated with metformin besides the reduction of insulin resistance. Metformin reduces appetite, suggesting it might have anorectic properties (9). Leptin reduction in adipose tissue and serum observed in vivo studies may indicate reduced leptin resistance (10, 11). Additionally, metformin treatment significantly increases GLP-1 levels, which could contribute to weight loss (12, 13).

Monotherapy and Comparative Studies

Studies show that metformin monotherapy's weight-reducing effect compared to placebo during the first year of diabetes treatment is not very significant or is neutral (14). However, compared to other antidiabetic drugs, which often cause weight gain, the difference becomes clinically significant. A review of 12 randomized trials in 2011 found an average weight gain of 2.7 kg in patients treated with sulfonylureas compared to those treated with metformin (15).

Longer treatment periods show more significant differences: in eight studies lasting up to 24 weeks, the weight difference was 1.9 kg (95% CI 1.4-2.5 kg), while in four studies over 24 weeks, the average difference was 3.6 kg (95% CI 3.1-4.1 kg) (15). Switching from SU drugs to metformin resulted in a weight loss of 3.8 kg, whereas patients continuing SU treatment showed no weight changes (16). Comparing metformin to TZDs showed a significant weight difference in favor of metformin. In eight comparative randomized trials lasting up to 1 year, the average weight difference was 2.6 kg (95% CI 1.2-4.1 kg) (15). In a 4-year double-blind randomized trial (ADOPT), the average weight difference between TZD (rosiglitazone) or metformin was 6.9 kg (95% CI 6.3-7.4 kg) (17).

Combination Therapy Studies

Three small-scale randomized trials comparing metformin and DPP-4 inhibitors on weight also showed greater weight loss with metformin, with an average difference of 1.4 kg (18, 19, 20). Two small-scale studies evaluating weight differences between metformin and meglitinides found a weight difference in favor of metformin, ranging from 2.0 kg to 3.4 kg (21, 22). Studies comparing metformin monotherapy with metformin and TZD, SU, and DPP-4 combinations showed lower weight gain in the metformin monotherapy group: -2.2 kg, -2.3 kg, and -0.2 kg respectively (15).

Higher doses of metformin in combinations corresponded to greater weight loss. Comparing metformin monotherapy to combinations with meglitinides showed a slight weight gain (0.9 kg) in the combination group (22, 23). Among metformin combinations, the greatest weight gain occurred with metformin and TZD: 0.9 kg higher than metformin and SU, and a 0.3-1.5 kg increase compared to a 0.4-1.2 kg weight loss with metformin and sitagliptin (15). Treatment with metformin and GLP-1 significantly reduced patient weight, showing a weight difference of 2.7 kg compared to metformin and TZD, and 2.4 kg compared to metformin and DPP-4 (24, 25).

Two randomized trials comparing metformin/GLP-1 and metformin/sulfonylurea combinations showed a significantly greater weight loss in the metformin/GLP-1 combination group: -3.8 kg and -12.3 kg (26, 27). Combining metformin with insulin resulted in less weight gain compared to insulin monotherapy. A 2004 Cochrane meta-analysis found a statistically significant lower weight gain (-3.7 kg) in patients using basal insulin with metformin compared to insulin monotherapy, a trend confirmed in subsequent studies (28, 29, 30, 31).

Practical Implications

Research indicates that treating type 2 diabetes with metformin, compared to placebo, usually results in a neutral or slightly positive effect on weight. However, compared to other antidiabetic drugs (SU, TZDs, meglitinides, DPP-4 inhibitors), patients on metformin show up to 6 kg less weight gain. Lower weight gain also occurs with metformin combined with other antidiabetic drugs and insulin. Even a modest weight difference carries clinical significance, as increases in body weight discourage treatment continuation and medication adherence (34, 35). Conversely, small weight losses (about 5% of body weight) significantly reduce cardiovascular risk, which is notably high in diabetes patients.