org355: Tree trial | LSHTM Moodle

From The Guardian

"Previous studies have shown statistical associations between exposure to microbial diversity and the development of a well-functioning immune system. But this is the first study to deliberately change the children’s environment and therefore indicate a causal link."

1. What kinds of studies were these previous studies? What aspects of these studies made it so a causal link was difficult to establish?

"The study involved 75 children in two cities in Finland, a relatively small number for a trial. “But when we saw the results, we were very surprised because they were so strong,” said Aki Sinkkonen, at Natural Resources Institute Finland, who led the work."

2. What does Sinkkonen mean by "strong" ?

3. Comment on the clinical endpoint of this study. Is it meaningful?

"The children were between three and five years old and spread between 10 similar daycare centres. In four centres, turf from natural forest floors, complete with dwarf shrubs, blueberries, crowberry, and mosses, were installed in previously bare play areas."

4. What is the name of this study design?

"The researchers gave all the children the same meals each day and excluded the small number who had been given probiotic supplements by their parents."

5. Why did the researchers give meals? Why did they exclude the small number?

Re: Tree trial

by | Olivia El Jassar - Thursday, 29 October 2020, 2:47 PM

1. Observational studies. This is the first INTERVENTION study in which the exposure is introduced by the researcher

2. A ‘powerful’ measure of effect was found. Statistically significant (p<0.05) may be arbitrarily used. Confidence intervals not including the null, of one?

3. After 28 days, children's skin and gut were tested for microbe diversity and blood samples were taken for immunological markers. 28 days is a very short period of time of exposure. Repeat testing was not performed thus we don't know how long lived these effects are - for example, another month down the line, any effect on the immune system may have reverted back to normal or even reversed

4. Intervention trial

5. Controlling for potential confounding factors (effect of meals on gut microbiome). Probiotics are a major confounding factor here.

Re: Tree trial

by | Selene Corsini - Tuesday, 3 November 2020, 10:19 AM

1. the previous studies 'have shown statistical associations' between the exposure of interest and the outcome, so they should have been analytical studies (ie case-control, cohort..). In these studies the association found may or may not be causal, as opposed to the intervention study described in the article, which gives more weight to a possible causal relationship.

2. he means that it is very likely that the results obtained from the study show a significant association between a diverse environment and a more functional immune system, therefore a reduction of immune mediated diseases. this measure is generally obtained by calculating confidence intervals and a p value, which is not given in the article.

3. children were exposed for 90 minutes per day to a more diverse playground environment and were encouraged to play with the soil and plants. samples from skin, guts and blood were collected showing increased microbial diversity on skin and guts and a higher level of immune cells in the blood.
In order to state whether these changes were meaningful in regards to providing a better immunity (and possibly lower predisposition to immune mediated diseases), we would need
-to estimate the prevalence of immune mediated disorders in the control group and observe a statistically significant difference between the two groups (ie higher prevalence in the control)
-prove that the presence of the above cells in the blood stream at one point in time is directly related to reduction of occurence of immune mediated disorders.

4. it is an intervention trial, we don't know whether the samples were randomised.

5. the meals were given in order to minimise differences between the 2 groups and therefore create comparability.
the small number of children receiving probiotics was excluded to not incur in counfounding

Re: Tree trial

by | OLGA VIACHESLAVOVNA KOZHAEVA - Saturday, 28 November 2020, 6:48 PM

Hi all
to add to the above

4.
- it seems it was a cluster trial, as allocation to the intervention was at daycare centre level.
- indeed it is somewhat unclear if the trial was randomised or not .. although, the article makes the claim of strong evidence and speaks of causality , and one could imagine that randomisation took place . BUT: when one looks at the original article, randomisation is not mentioned. would anyone have further thoughts on non-randomised trials and whether it is fair to claim causality on this basis?

5. another remark is that they used resriction to control for confouding (ie excluding children taking probiotics) also because the sample size was small and they wanted to ensure control of this strong confounder. if the sample size was larger, probiotics may have been balanced at baseline and/or controllled for at analysis stage. (does this make sense?)

Best wishes
Olga