Effect of ketogenic diets on insulin-like growth factor (IGF)-1 in humans: A systematic review and meta-analysis

<bold>Background: </bold>Insulin-like growth factor (IGF)-1 plays a role in aging and cancer biology, with fasting known to reduce serum IGF-1 levels in human adults. However, the impact of ad libitum ketogenic diets (KDs) on IGF-1 levels remains unclear. <bold>Methods: </bold>Adhering to PRISMA guidelines, we conducted a meta-analysis of human trials by systematically searching Ovid, PubMed, Scopus, and CENTRAL Libraries until June 2023. Eligible studies prescribed KDs to adults of any health status, confirmed ketosis, and measured serum IGF-1. Protocols involving prescribed fasting or energy restriction were excluded. Mean differences (MD) and 95 % confidence intervals (CIs) were calculated longitudinally between pre- and post-intervention measurements for the KD groups. <bold>Results: </bold>Among twelve publications meeting the inclusion criteria, 522 individuals participated, with 236 completing KDs. The intervention duration ranged from 1 to 20 weeks. Pooled results from ten trials showed a significant reduction in serum IGF-1 levels post-intervention (MD: -24.9 ng/mL [95 % CI -31.7 to -18.1]; p<0.0001) with low heterogeneity across studies (I-2=27 %, p=0.19). KDs were also associated with significantly decreased fasting insulin (MD: -2.57 mU/L [95 % CI -4.41 to -0.74], p=0.006) and glucose (MD: -7.30 mg/dL [95 % CI -11.62 to -2.98], p=0.0009), although heterogeneity was significant. Subgroup analyses on study design, gender, dietary duration, and oncological status revealed no significant differences. <bold>Conclusion: </bold>Ad libitum KDs (>55 % fat) effectively induce ketosis and can lower serum IGF-1 by 20 %, fasting glucose by 6 % and insulin by 29 %. This clinically notable reduction in IGF-1 can be attained without the need for a prescribed fasting or severe calorie restriction regimen. Further investigation is warranted to explore the impact of KDs on ageing biomarkers and cancer management.