LSR2 r markdown_mediators.Rmd

---
title: Mechanisms through which exercise reduces symptom severity and/or functional impairment in posttraumatic stress disorder (PTSD) - Research question II
author: Virginia Chiocchia
date: "`r format(Sys.Date(), '%d %b %Y')`"
output: html_notebook
mainfont: Arial
---

```{r setup,message=F,warning=F,echo=F,include=F}

### libraries
library(meta)
library(tidyr)
library(tibble)
library(stringr)
library(readxl)
library(grid)
library(dplyr)
library(rjags)
library(Matrix)
library(knitr)
library(kableExtra)
library(rms) 
library(meta)
library(devtools)
library(dosresmeta)
library(tools)
library(metafor)
library(rje)
library(gtools)


load("data/data_2023-12-4_LSR2_Awks2.RData")
load("data/LSR2weeksStata2insufdata.RData")

load("data/LSR2_studychartab_wks_140124.RData")
load("data/LSR2_studychartab_insuf_140124.RData")

load("data/LSR2_interventions_150224.RData")


```

# Results

## Flow of study selection and descriptives

The flow of study selection is shown in Figure 1 (*to be included*). We examined potential mediators affecting the relationship between the *intervention* and the *outcome(s),* as shown in Figure 2*.* Overall, `r length(table(LSR2_mediators$study))` studies reported data on association between *intervention* and the *outcome(s):*

-   `r ###` studies reported data on *mediator 1*

-   `r ###` studies reported data on *mediator 2*

-   ...

The earliest studies were performed in `r ###` while the most recent studies were performed in `r ###`

The median sample size across the studies was `r ###` participants per arm. The median of the mean participant age was `r ###` years (ranged from `r ###` to `r ###` years).

*Include DAG(s) for relevant mediators*

![*Fig.2 Example of a mediation of the effect of exercise on improving PTSD symptoms via changing the levels of BDNF. The indirect effect of exercise on the outcome A~Indirect~ is obtained as the product of B and C effect sizes. The direct effect A~Direct~ would typically not be reported in studies; instead, the total effect A~Total~= A~Direct~ +A~Indirect~ is reported.*](Picture%201.png){width="15cm"}

### Descriptive associations

In total, `r ###` studies only reported the associations descriptively.

*Add the reported statements about associations.*

### Mediation analysis

In total, `r ###` studies performed proper mediation analysis and reported the following indirect and direct effects for the mediators.

```{r,echo=F }

# maybe better if dataset is in long format (if studies have more than one mediators)
kable(LSR2_mediators %>% select(study, mediator, Atotal, 
                                Aindirect, Adirect),
      digits = 0,
      caption = "Table 1: Mediators, direct and indirect effects for the included studies",
      col.names = c("Study","Mediator/Mechanism",
                    "Total Effect Intervention Outcome",
                    "Indirect Effect Intervention Outcome", 
                    "Direct Effect Intervention Outcome")) %>%
  footnote() %>%
  kable_paper("striped", full_width = F) #%>%
  #scroll_box(width = "500px", height = "200px") 

  
```

| Study | Mediator/Mechanism | Total Effect | Indirect Effect | Direct Effect |
|-------|--------------------|--------------|-----------------|---------------|
|       |                    |              |                 |               |
|       |                    |              |                 |               |

## Synthesis of mediator(s) estimates

*If relevant add results from appropriate methodology as described in protocol.*

### Sensitivity analyses

*Add any if relevant*

### Subgroup analyses

*Add any if relevant*

## Risk of bias assessment

...

## Summary of evidence on mediators

...