Surprising Findings: Spilling the Tea
I thought I was seeing double! As an avid tea drinker, an article investigating tea consumption with cancer risk caught my eye. But, spoiler alert! This “surprising finding” isn’t specifically about tea or cancer risk. During a perusal of the literature, I had stumbled upon not one, but two meta-analyses assessing tea consumption and cancer risk in observational studies published within months of each other in the same journal by different authors. While repetition and validation are important aspects of the scientific process, it is not every day that articles appearing to address the same research question and objectives are published in the same journal so close together, unless there is a debate on the topic. So, I was curious to investigate “why?”
My investigation had two main questions.
First Question: What did the authors find and did the findings between the two articles differ?
To provide a bit more background, both research teams meta-analyzed observational findings examining the association of black tea, green tea or total tea intake on the risk of any type of cancer (1,2). Neither study reported excluding articles from their analyses based on the type of participants being studied, the duration of the follow up or the setting of where the study took place.
Surprisingly, overall, there appeared to be little evidence to support an association between tea drinking with cancer risk; however, the findings did differ between the two articles.
Kim and colleagues were the first to publish their research showing a significant inverse association between tea consumption and the risk of oral cancer (1). Inverse associations with other types of cancer were also reported; however, the authors noted the findings were not supported by consistent or convincing evidence.
Zhao and colleagues showed, except for an inverse association between lymphoid neoplasms with green tea intake, no consistent associations were observed between tea consumption and cancer risk when there was more than one cohort, either for assessment of the highest compared to the lowest tea intake nor in dose-response analyses (2).
Second Question: What could have led to these differences?
A key difference was in the methodology, specifically the type of meta-analysis conducted. Kim and colleagues conducted an umbrella review (i.e. a meta-analysis of meta-analyses, which syntheses information from all systematic reviews and meta-analyses on a given topic) (1), whereas Zhao and colleagues meta-analyzed individual prospective cohorts (2).
An umbrella SRMA has been called one of the ‘next-generation’ SRMAs being touted to “raise the bar and help shape a new generation of more reliable evidence synthesis” (3). While an umbrella SRMA certainly has the potential to enable an efficient method for higher level synthesis of evidence allowing for detection associations, effects, uncertainties and knowledge gaps, its strength is still dependent on the type of data available and included. Kim and colleagues present one example of an umbrella SRMA acknowledging limitations based on the types of studies included (1). Specifically, the authors reported inclusion of systematic reviews with both cohort and case-control studies, and due to the potential biases that may affect case-control studies, such as recall and selection bias, recommended further analysis of only prospective studies.
Zhao and colleagues met this call as they purposely included only prospective cohort studies in their analyses to reduce potential bias (2). As well, they provided separate analyses by tea type (black, green, total) (2), which had not been done by Kim and colleagues (1).
Additional methodological comparisons consisted of the umbrella SRMA including only English language publications (1), whereas the articles included by Zhao and colleagues had no language restriction (2). Only the Zhao article reported publicly registering their study protocol (2), which allows for greater transparency. Given the observational nature of both studies, a causal relationship between tea and cancer could not be inferred in either case. Despite limitations, both articles should be acknowledged for their strengths, specifically their comprehensive designs affording an overview of at least 25 cancer sites using standardized SRMA methods.
These studies serve as a reminder of the importance of assessing the strengths and limitations of research methods. One of the earliest research lessons I received was to read and critically assess the methods of a paper first (even before the abstract!). This advice highlights the importance of research methodology and continues to stay front of mind as I ponder the next article over a cup of tea.
If you were hoping to learn more about tea and/or cancer, don’t worry the PEN
® System has you covered! Check out this
content.
Written by:
Stephanie Nishi RD, MSc, PhD(c)
PEN Evidence Analyst
References
- Kim TL, Jeong GH, Yang JW, Lee KH, Kronbichler A, van der Vliet HJ, et al. Tea Consumption and Risk of Cancer: An Umbrella Review and Meta-Analysis of Observational Studies. Adv Nutr. 2020 Jul 15:nmaa077. doi: 10.1093/advances/nmaa077. Epub ahead of print. PMID: 32667980. Abstract available from: https://pubmed.ncbi.nlm.nih.gov/32667980
- Zhao LG, Li ZY, Feng GS, Ji XW, Tan YT, Li HL, Gunter MJ, Xiang YB. Tea Drinking and Risk of Cancer Incidence: A Meta-Analysis of Prospective Cohort Studies and Evidence Evaluation. Adv Nutr. 2020 Oct 1:nmaa117. doi: 10.1093/advances/nmaa117. Epub ahead of print. PMID: 33002099. Abstract available from: https://pubmed.ncbi.nlm.nih.gov/33002099
- Ioannidis J. Next-generation systematic reviews: prospective meta-analysis, individual-level data, networks and umbrella reviews. Br J Sports Med. 2017 Oct;51(20):1456-1458. doi: 10.1136/bjsports-2017-097621. Epub 2017 Feb 21. PMID: 28223307. Link available from: https://pubmed.ncbi.nlm.nih.gov/28223307
November 2020 ·
Volume 10
(11)