The initial search retrieved 2368 unique articles to be screened, of which 107 were selected for full-text evaluation after screening based on titles and abstracts. The reasons for the unsuitability are given in Fig. 1. After a greater detailed appraisal of the full texts, 72 studies were discharged due to following reasons: a) had only data on other inflammatory markers (n = 50) instead of the outcomes of interest; b) did not provide adequate data to be suitable for the final meta-analyses (n = 5); c) administrated an LCD which was accompanied by other dietary interventions (n = 12); d) prescribed extremely short-term treatments (n = 4) and f) an LCD was defined as > 45% from calorie (n = 1). Ultimately, from 35 citations [30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64] were considered potentially eligible for qualitative assessments, 2 studies [41, 46] were dropped owing to reporting the implausible baseline concentration of outcomes of interest, resulting in 33 studies [30,31,32,33,34,35,36,37,38,39,40, 42,43,44,45, 47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64] included in the quantitative analyses. In aggregate, there were 24 unique RCTs [30, 31, 34,35,36, 38,39,40, 42, 44, 45, 47,48,49,50, 52, 56,57,58,59,60,61,62,63] with 26 datasets for TNF-α and 28 unique citations [30, 32,33,34,35,36,37,38,39,40, 42,43,44,45, 47, 49,50,51,52,53,54,55,56, 58, 59, 61, 63, 64] with 30 datasets for IL- 6 to be included in this meta-analysis.

PRISMA flow diagram for study selection process

The main characteristics of included trials are summarized in Table 2. Based on this table, the majority of studies were carried out in the United States [31, 33, 37,38,39,40, 47, 49, 54, 55, 57,58,59, 62,