The redundant landscape of meta-analyses: Evaluating exercise interventions for older adults with cognitive...

Exercise to benefit cognition in older adults with cognitive impairment is well-studied. A recent proliferation of synthesis studies might be a positive contribution to the science; however, redundancy in research can be wasteful and detrimental to drawing confident conclusions about the evidence. This synthesis-based method study was designed to analyze: 1) the frequency and growth patterns of meta-analyses (MAs) and randomized controlled trials (RCTs) on this topic; and 2) the redundancy rate of the RCT studies included in MAs.

This study was borne of a living systematic review, following standard synthesis methodology (PROSPERO registration, librarian-assisted search algorithms developed for multiple databases, searches updated regularly with most recent search in 2025). Frequency counts determined the number of RCTs and how many times they were included across the MAs.

Forty MAs were identified and included in the synthesis representing a total of 728 RCT studies. After reviewing RCT duplicates, 276 (37.9%) unique RCT studies comprised this body of evidence. Among these, 153 RCTs were cited in only one MA and 123 were cited in 2-19 different MAs. Thus, 452 (62.1%) of all RCTs used across the 40 MAs were redundant (i.e., represented in more than one MA).

This study found substantial redundancy in the RCTs used in MAs evaluating the cognitive impact of exercise for older adults with cognitive impairment. Replication is common practice in research but reliance on the same RCTs in multiple MAs creates an illusion of robustness when, in fact, the strength and diversity of the evidence may be more limited. Research redundancy is wasteful and may actually stagnate advancement of this science.