Original: Examples explain what Is Meta Regression and how to use Meta Regression to publish articles!

Meta analysis, clinical data analysis one-on-one guidance, in line with academic norms, direct submission at the end of the course, +tjzgBL yo!

As we all know, reading high-quality literature can quickly improve your scientific research ability and help yourself quickly understand some jerky concepts

Before we get started, let's talk about some basic knowledge

Meta regression is a statistical analysis method

The article shared today is a meta-analysis article published in the top issue of Crit Rev Food Sci Nutr, which evaluated the effects of artificial sweeteners and stevia sweeteners on lipid profile markers through a systematic review and meta-analysis of randomized controlled trials (RCTs) and finally reached conclusions

The paper, which has an impact factor of nearly 12, was published in the first quarter of 2022 and was one of the few articles

Statistical light - Meta regression

Topics

The article is: "The effects of artificial- and stevia-based sweeteners on lipid profile in adults: a GRADE-assessed systematic review, meta-analysis, and meta-regression of randomized clinical trials , (PROSPERO registration number: CRD42021250025) This study explored the effects

As we said in the previous Meta Analysis Topic Selection Course, meta-analysis topic selection should adhere to the principles

Because many clinical studies have produced inconsistent results within the framework of whether NNS affects lipid levels, and the effect of NNS on lipid distribution is not clear, it is necessary

Statistical light - Meta regression

Controversial: whether NNS affects blood lipid levels, many clinical studies have produced inconsistent results;

Innovative: the effect of NNS on lipid distribution is unclear;

literature search

The study systematically searched three databases, Pubmed, Scopus and EMABASE, with a deadline of April

Search strategies were developed according to PICOS, P study subjects, I intervention modalities, C comparative measures, outcome measures of the O study, S study type

Randomized controlled trials of [non-nutritive sweeteners] versus [adults] [lipids], so the article retrieved:

I Intervention mode - non-nutritive sweeteners

O Outcome measure - lipids

S Study Type - Randomized Controlled Trial

At the time of screening, the population was limited, and the initial screening and re-screening excluded non-adult study subjects

Statistical light - Meta regression

Criteria and literature screening

Statistical light - Meta regression

At the time of the initial screening, of the 3212 articles retrieved, 1243 duplicates, 1178 unrelated, 518 animal studies, 150 reviews and 29 conference papers were excluded, and the remaining 94 documents were included in the re-screening

During the re-screening, 80 literature was excluded according to the criteria for exclusion, and 14 literature were finally included as studies

Statistical light - Meta regression

The flowchart here can actually be written in more detail, in the first step, write which libraries were searched, and how many documents were retrieved

Extracts

Two independent researchers extracted information from the original literature of the included randomized controlled trials

Statistical light - Meta regression

Statistical light - Meta regression

Statistical analysis

Continuous variables use the mean and standard deviation to obtain the overall effect size

It is worth noting here that the covariates included in the Meta Regression Analysis are some of the characteristics of the study or experimental level, and cannot be a single numeric value

Statistical light - Meta regression

Subgroup analysis is a highlight of this article, and the setting is very detailed
.
This article considers the source, duration of use, dosage, and country of sweeteners, and conducts sensitivity and subgroup analyses of health status BMI in the study population
.

The use of Meta regression is mainly used to explore the sources of
heterogeneity.
Because when a large number of studies are included in a topic, sometimes there will be inconsistencies between the results of one study and other studies, which may be related to the research protocol, the time and area of the study, the quality of the study, the method of study, the age and sex of the participants, etc.
, then meta-regression needs to be used to clarify the source
of heterogeneity between studies.

The summary analysis of this article showed no statistical differences in TG, TC, LDL, and HDL, but the subgroup analysis showed that in subjects with normal LDL levels (<100 mg/dl), NNS may be associated with a small increase in LDL (wMD: 4.
23, 95% Ci: 0.
50, 7.
96 mg/dl), with statistical differences
.

It was eventually concluded that the intake of artificial sweeteners and stevia sweeteners was not associated
with changes in blood lipid levels in adults.

Quality assessment

The type of studies included was RCTs, so the included literature was assessed using the Cochrane quality assessment and the grade review
.

Statistical light - Meta regression

summary

Diets rich in sucrose can induce hyperglycemia and hyperlipidemia, while non-nutritive sweeteners (NNS) are considered to be low-calorie and sugar-surrogate compounds, so they are used by many companies in dietary beverages, claiming to be "sugar-free" and can be used for weight loss, even for obese and diabetic patients
.

However, whether NNS intake has an impact on health has always been controversial, especially the impact of NNS on
lipid distribution.
Some studies have shown that NNS can be effective in reducing TG, TC and LDL; Some said there was no significant impact
.

This issue investigates all types of artificial sweeteners and stevia sweeteners through systematic search strategies and compares
them across different subgroups using Meta regression.
It was concluded that ingestion of artificial sweeteners and stevia sweeteners may not affect serum TG, TC, LDL and HDL levels
.

The paper uses the common meta-analysis statistical process, from information extraction to statistical analysis is not particularly complicated, but the author is still successfully published in high-quality journals based on highly controversial and innovative topic selection and clear writing ideas
.

Although the results of the study show that sugar-free beverages have no significant effect on the body's lipid distribution, it does not mean that sugar-free beverages are healthy
.

Statistical light - Meta regression

If you also want to publish a meta-analysis article this year, or want to understand meta-regression in more detail, as well as other meta-analysis types, Meta-analysis one-on-one full-process guidance is recommended to you, from topic selection to submission
.

Not only can you master Meta, but also make your own results at the same time, and achieve the goal
of improving scientific thinking, statistical ability, and writing ability in an all-round way.

Whether you are a medical student in school, facing graduation problems!

Still a front-line clinician, facing the problem of busy scientific research!

Or maybe I'm just interested in meta-analysis, and I've learned a lot of tutorials but haven't published an article! All right!

Statistical light - Meta regression

Responsible Editor: