I am in the process of writing my systematic review in which I asses the longitudinal association between baseline endothelial function and diabetes type II at follow up in healthy participants. we selected only cohort studies. most of the studies didn't report the analysis between these two measures so we had to ask the authors for multivariate regression and they reported beta, standard error and significance. the analysis they did for endothelial function measure (FMD or RHI) and diabetes status measure (fasting plasma glucose or HbA1c) reported by 3 studies is insignificant as p value was insignificant. only one study reported baseline endothelial function and incidence of diabetes at follow up with significant relative ratio. and it matches with our hypothesis. now I have to conclude and discuss this two different outcome and build my case around it.
i want to ask that what things and factors I should consider in building my discussion. Any help in any direction would be very helpful. I feel like having two different outcomes is the key here but what could I use to differentiate between the two outcomes?
Thank you
If the p-value is insignificant it's impossible to come to any conclusions... Maybe there is no correlation, but on the other hand, it doesn't exclude the thesis entirely. Let's think if the size of the study group was sufficient, what was a statistical power, and differences in the test protocols that could differentiate the results of these studies.
Monika Klimek-Tulwin thanks for your reply.
The insignificant p is for the 3 study with different follow up and population(smaller). however, one study with almost 840 participants showed the positive association but the type of outcome was different in this study. so it makes me think that we cannot completely invalidate the hypothesis. I am now thinking if it could be because of this difference in type of outcome(i.e diabetic measure Fasting plasma glucose, Hba1c & incidence of diabetes). for example, incidence of diabetes is due to many factors like environmental, family history etc.
- Badges
- Science topic
- Similar topics
- Filing