-
Notifications
You must be signed in to change notification settings - Fork 0
/
LSR2_animal_analysis.Rmd
2749 lines (1850 loc) · 248 KB
/
LSR2_animal_analysis.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
---
title: "LSR2 analysis: Mechanisms through which exercise reduces symptom severity and/or functional impairment in posttraumatic stress disorder (PTSD)"
author: "Francesca Tinsdeall, Fiona Ramage, Virginia Chiocchia and Malcolm Macleod"
date: "2024-02-20"
output:
html_document:
fig_width: 9
toc: TRUE
bibliography: grateful-refs.bib
editor_options:
chunk_output_type: console
---
Date report generated: `r Sys.Date()`
```{r setup, message=F, echo=F, include=F}
### libraries
library(devtools)
library(dosresmeta)
library(dplyr)
library(grid)
library(gtools)
library(kableExtra)
library(graphics)
library(patchwork)
library(forcats)
library(knitr)
library(ggplot2)
library(Matrix)
library(meta)
library(metafor)
library(orchaRd)
library(readxl)
library(readr)
library(rje)
library(rms)
library(stringr)
library(tibble)
library(tidyr)
library(tools)
library(rlang)
#devtools::install_github("mcguinlu/robvis")
#install_github("mcguinlu/robvis")
library(robvis)
library(PRISMA2020)
library(grateful)
library(xtable)
library(scales)
#define LSR
LSR <- 'LSR2'
# define date of processing
DoP <- Sys.Date()
# All function needed to run this notebook (analyses, etc.) should be in a util.R file
source("utils/util.R")
# obtain the data and prepare them for analysis - all data cleaning routines should be in this .R script
source("wrangling/wrangling_functions.R", local = TRUE)
source("wrangling/data_wrangle_script.R")
source("utils/development.R")
#Round off results to two digits
options(scipen=100, digits=3)
# Import data
file2load <- paste0(LSR,'_clean_data_',DoP,'.csv')
df <- read_csv(file2load)
```
# 1. Flow of study selection and descriptives
```{r article metadata, eval = TRUE, echo = FALSE}
# Retrieve article metadata omitted from 'clean_data.csv' and join
article_metadata <- df %>%
select(StudyId, Title, Year) %>%
distinct()
```
The flow of study selection is shown in Figure 1. Studies included were published between `r min(article_metadata$Year)` and `r max(article_metadata$Year)`. Overall, this analysis includes `r length(table(df$StudyId))` studies containing `r nrow(df)` comparisons.
**Figure 1**
```{r PRISMA flowchart, eval = TRUE, echo = FALSE, warning=FALSE, message=FALSE}
prisma_data <- read_csv("data/LSR2_prisma_200224.csv")
dfp <- PRISMA_data(prisma_data)
PRISMA_flowdiagram(dfp, interactive=FALSE, previous=FALSE, other=FALSE,
detail_databases=FALSE, detail_registers=FALSE, fontsize=12, font="Helvetica",
title_colour="Goldenrod1", greybox_colour="Gainsboro", main_colour="Black",
arrow_colour="Black", arrow_head="normal", arrow_tail="none", side_boxes=TRUE )
```
```{r Included studies, eval = TRUE, echo = FALSE, warning=FALSE, message=FALSE}
RoB <- unique(df[,c(4,6,12)])
#change studyId to Author, year
RoB$Study <- RoB$StudyId_I
RoB$StudyId <- toupper(paste0(str_extract(RoB$Authors_I,"\\b\\w+\\b"),', ',RoB$Year_I))
# fix >1 publication per first author in a year
unique_study_ids <- unique(RoB$StudyId)
suffix_list <- character(length = nrow(RoB))
for (study_id in unique_study_ids) {
indices <- RoB$StudyId == study_id
if (sum(indices) > 1) {
suffix_list[indices] <- letters[seq_along(suffix_list[indices])]
}
}
RoB$suffix <- suffix_list
# Add the suffix to the original column
RoB$StudyId <- paste(RoB$StudyId, RoB$suffix, sep = "")
# Remove the 'suffix' column if you no longer need it
RoB <- select(RoB, -suffix)
RoB <- RoB[order(RoB$StudyId),]
colnames(RoB) <- c('Study', 'Authors', 'Year', 'StudyId')
colnamesdf <- c('StudyId_I', 'Authors_I','Year_I','Strain', 'Label', 'outcome_type',
'NumberOfAnimals', 'NumberOfAnimals_C', 'NumberOfAnimals_I', 'SortLabel')
df3 <- df[,colnamesdf]
df3$N <- as.numeric(df3$NumberOfAnimals_C) + as.numeric(df3$NumberOfAnimals_I)
df3$CompLable <- df3$Label
trow <- merge(df3, RoB, by.x = 'StudyId_I', by.y = 'StudyId')
colnames(trow)[16] <- "Id"
trowmodel <- subset(trow, trow$SortLabel == 'CvS')
trowint <- subset(trow, trow$SortLabel =='TvC')
tab1m <- trowmodel[,c(16,4,6,11)]
tab1i <- trowint[,c(16,4,5,6,11)]
tab2i <- tab1i %>%
group_by(Id, Strain, Label, outcome_type) %>%
summarise(subjects_sum = sum(N))
tab2m <- tab1m %>%
group_by(Id, Strain, outcome_type) %>%
summarise(subjects_sum = sum(N))
colnames(tab2i) <- c("Study", "Strain", "Comparison", "Outcome","N")
colnames(tab2m) <- c("Study", "Strain", "Outcome","N")
tab2m <- tab2m %>%
arrange(Study, Strain, Outcome)
tab2i <- tab2i %>%
arrange(Study, Strain, Comparison, Outcome)
write.csv(tab2i, 'tab2i.csv')
write.csv(tab2m, 'tab2m.csv')
```
**Table 1** below gives a summary of the included studies for the effect of exercise interventions. N represents an aggregate of animals contributing to outcomes reported from control and treatment groups, and if the same control group has contributed to more than one experiment, it will be counted twice.
```{r results="asis", echo = FALSE, warning=FALSE, message=FALSE}
tab2i <- read_csv("data/tab2i.csv")
tab2i$N <- as.numeric(tab2i$N)
original_data <- tab2i
# Rows to have bold lines beneath
rows_with_bold_lines <- c(2,6,10,15,18,23,25,28,31,34,36,40)
# Generate HTML table with adjusted styles
cat('<div style="text-align: center;">')
cat('<table style="width: 100%; border-collapse: collapse;">')
# Table header
cat('<tr style="border-bottom: 2px solid black;">')
for (col_name in names(original_data)) {
cat(paste('<th style="padding: 8px; text-align: center;">', col_name, '</th>', sep = ''))
}
cat('</tr>')
# Table rows
for (i in seq_len(nrow(original_data))) {
cat('<tr>')
for (j in seq_len(ncol(original_data))) {
cell_content <- original_data[i, j]
cell_style <- ifelse(i %in% rows_with_bold_lines, 'border-bottom: 2px solid black; padding: 8px; text-align: center;', 'border-bottom: 1px solid black; padding: 8px; text-align: center;')
if (j == 1) {
cell_style <- gsub('text-align: center;', 'text-align: left;', cell_style)
}
if (i == nrow(original_data)) {
cell_style <- gsub('border-bottom: 1px solid black;', 'border-bottom: 2px solid black;', cell_style)
}
cat(paste('<td style="', cell_style, '">', cell_content, '</td>', sep = ''))
}
cat('</tr>')
}
cat('</table>')
cat('</div>')
```
**Table 2** below gives a summary of the included studies for the effect of model induction. N represents an aggregate of animals contributing to outcomes reported from control and treatment groups, and if the same control group has contributed to more than one experiment, those animals will be counted more than once.
```{r results="asis", echo = FALSE, warning=FALSE, message=FALSE}
tab2m <- read_csv("data/tab2m.csv")
tab2m$N <- as.numeric(tab2m$N)
original_data <- tab2m
# Rows to have bold lines beneath
rows_with_bold_lines <- c(1,2,5,9,14,17,22,24,27,30,33,35,39)
# Generate HTML table with adjusted styles
cat('<div style="text-align: center;">')
cat('<table style="width: 100%; border-collapse: collapse;">')
# Table header
cat('<tr style="border-bottom: 2px solid black;">')
for (col_name in names(original_data)) {
cat(paste('<th style="padding: 8px; text-align: center;">', col_name, '</th>', sep = ''))
}
cat('</tr>')
# Table rows
for (i in seq_len(nrow(original_data))) {
cat('<tr>')
for (j in seq_len(ncol(original_data))) {
cell_content <- original_data[i, j]
cell_style <- ifelse(i %in% rows_with_bold_lines, 'border-bottom: 2px solid black; padding: 8px; text-align: center;', 'border-bottom: 1px solid black; padding: 8px; text-align: center;')
if (j == 1) {
cell_style <- gsub('text-align: center;', 'text-align: left;', cell_style)
}
if (i == nrow(original_data)) {
cell_style <- gsub('border-bottom: 1px solid black;', 'border-bottom: 2px solid black;', cell_style)
}
cat(paste('<td style="', cell_style, '">', cell_content, '</td>', sep = ''))
}
cat('</tr>')
}
cat('</table>')
cat('</div>')
```
References of included studies are located in the appendix. Included studies used `r length(table(df$ModelID))` unique disease model induction procedures.
## 1.1 Description of experiment types and methodological approach
Within the literature we identified distinct categories of experiments and the data presented would allow several meta-analytical contrasts to be drawn:
**Treatment vs control**. These were experiments investigating the effect of performing exercise, reported in `r nrow(df %>% filter(SortLabel == "TvC"))` experiments from `r nrow(df %>% filter(SortLabel == "TvC") %>% distinct(StudyId))` publications.
In these studies the:
- **Control group** is a group of animals that is (1) subjected to a PTSD model induction paradigm and (2) administered a control treatment (vehicle) or no treatment.
- **Intervention group** is a group of animals that is (1) subjected to a PTSD model induction paradigm and (2) performing exercise.
- **Sham group** is a group of animals that is (1) not subjected to a PTSD model induction paradigm and (2) administered a control treatment (vehicle) or no treatment. These data are required to allow a 'normalised mean difference' effect size to be calculated, given by
$$
\frac{{\bar{\mu}_C - \bar{\mu}_T}}{{\bar{\mu}_C - \bar{\mu}_S}} \times 100
$$
where $\bar{\mu}_C$, $\bar{\mu}_T$, $\bar{\mu}_S$ are the mean reported scores in the control, treatment, and sham groups respectively.
**Effects of disease modelling**. These are experiments investigating the effect of models of PTSD, reported in `r nrow(df %>% filter(SortLabel == "CvS"))` experiments from `r nrow(df %>% filter(SortLabel == "CvS") %>% distinct(StudyId))` publications.
In these studies the:
- **Control group** is a group of animals that is (1) not subjected to a PTSD model induction paradigm and (2) is administered a control treatment (vehicle) or no treatment.
- **Intervention group** is a group of animals that is (1) subjected to a PTSD model induction paradigm and (2) is administered a control treatment (vehicle) or no treatment.
Outcomes with ≥2 independent effect sizes were considered for meta-analysis. In this iteration of the review, this includes `r df %>% group_by(outcome_type) %>% filter(n_distinct(StudyId) > 1) %>% summarise(n = n_distinct(StudyId)) %>% arrange(desc(n)) %>% pull(outcome_type) %>% unique() %>% tolower() %>% { if (length(.) > 1) paste(paste(head(., -1), collapse = ", "), "and", tail(., 1)) else .}`.
All analyses were conducted allowing for the following hierarchical levels in a random effects model, which accounts for features common to experimental contrasts such as a shared control group:
- **Level 1: Rodent strain** - effect sizes measured across experiments using the same rodent strain.
- **Level 2: Study** - effect sizes measured from different experiments presented in the same publication.
- **Level 3: Experiment** - effect sizes measured in the same experiment within a study, where often a control group contributes to several effect sizes.
Each level for the hierarchy was only included in the model if more than 4 categories were present for at least one of these levels. Where more than 4 categories are not present for all levels, the variance attributable to levels with fewer than 5 categories is reported as zero.
The hierarchical grouping may therefore be considered thus: **Strains** of laboratory animals are included in several **Studies**, each of which can report one or more **Experiments**, and each Experiment is comprised of at least two **Cohorts** which are considered identical except for differing in the experimental manipulation (the **Intervention**) or not being exposed to the disease modelling procedures (a **Sham** cohort, these only being used to provide a baseline for outcome measures to allow Normalised Mean Difference meta-analysis). An **Experiment** can include several **experimental contrasts**, for instance where different doses of drugs are compared to the same control group. Several outcomes can be measured and reported from the same cohort of animals.
We constructed multilevel models without Hartung-Knapp adjustments as these are not available for rma.mv class objects in the metafor package. Instead, the model is set to `test = "t"` to use t- and F-distributions for making inferences, and `dfs="contain"` to improve the method of approximating degrees of freedom of these distributions.
The scales and units used to measure outcomes in preclinical studies often differ between studies although they may measure the same underlying biological construct. The primary effect size used for meta-analysis of preclinical studies is therefore the standardised mean difference (SMD, Hedge's g). For experiments testing the effects of interventions we also present a sensitivity analysis using normalised mean difference (NMD), where there are sufficient data for sham procedures to allow this. This analysis is not possible for studies of the effect of single prolonged stress (SPS).
# 2 Exercise v Control
```{r # Split df by experiment and outcome type into dataframes to simply inline code below - Exercise v Control, eval = TRUE, echo = FALSE}
df_S <- filter(df, SortLabel == "TvC")
df_S_NT <- filter_experiment_outcome_type(df, "TvC", "Neurotransmitter levels")
df_S_L <- filter_experiment_outcome_type(df, "TvC", "Locomotor")
df_S_B <- filter_experiment_outcome_type(df, "TvC", "BDNF")
df_S_OB <- filter_experiment_outcome_type(df, "TvC", "Other behavioural")
df_S_OthN <- filter_experiment_outcome_type(df, "TvC", "Other neurobiological")
df_S_FM <- filter_experiment_outcome_type(df, "TvC", "Fear memory")
df_S_Fr <- filter_experiment_outcome_type(df, "TvC", "Freezing")
df_S_StR<- filter_experiment_outcome_type(df, "TvC", "Stress response")
```
`r length(table(df_S$StudyId))` studies (`r nrow(df%>%filter(SortLabel == "TvC"))` comparisons) investigated the effects of exercise versus control. The number of studies and individual effect sizes for each outcome were:
- Locomotor outcomes: `r length(table(df_S_L$StudyId))` studies and `r nrow(df_S_L)` comparisons in `r {strains <- unique(df_S_L$Strain); if (length(strains) > 1) paste(length(strains), "strains") else paste(length(strains), "strain")}`
- Fear Memory `r length(table(df_S_FM$StudyId))` studies and `r nrow(df_S_FM)` comparisons in `r {strains <- unique(df_S_FM$Strain); if (length(strains) > 1) paste(length(strains), "strains") else paste(length(strains), "strain")}`
- Freezing: `r length(table(df_S_Fr$StudyId))` studies and `r nrow(df_S_Fr)` comparisons in `r {strains <- unique(df_S_Fr$Strain); if (length(strains) > 1) paste(length(strains), "strains") else paste(length(strains), "strain")}`
- Other behaviours: `r length(table(df_S_OB$StudyId))` studies and `r nrow(df_S_OB)` comparisons in `r {strains <- unique(df_S_OB$Strain); if (length(strains) > 1) paste(length(strains), "strains") else paste(length(strains), "strain")}`
- BDNF: `r length(table(df_S_B$StudyId))` studies and `r nrow(df_S_B)` comparisons in `r {strains <- unique(df_S_B$Strain); if (length(strains) > 1) paste(length(strains), "strains") else paste(length(strains), "strain")}`
- Neurotransmitter levels: `r length(table(df_S_NT$StudyId))` studies and `r nrow(df_S_NT)` comparisons in `r {strains <- unique(df_S_NT$Strain); if (length(strains) > 1) paste(length(strains), "strains") else paste(length(strains), "strain")}`
- Biochemical stress responses: `r length(table(df_S_StR$StudyId))` studies and `r nrow(df_S_StR)` comparisons in `r {strains <- unique(df_S_StR$Strain); if (length(strains) > 1) paste(length(strains), "strains") else paste(length(strains), "strain")}`
- Other neurobiological outcomes: `r length(table(df_S_OthN$StudyId))` studies and `r nrow(df_S_OthN)` comparisons in `r {strains <- unique(df_S_OthN$Strain); if (length(strains) > 1) paste(length(strains), "strains") else paste(length(strains), "strain")}`
## 2.1 Outcome 1: Locomotor activity
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_L <- run_SMD(df, "TvC", "Locomotor")
#TvC levels 4,4,2 so stop
```
Multilevel analysis is only performed if there are 5 levels or more for at least one of Strain, Study and Experiment, and that is not the case here. `r SMD_S_L[["SMD_ML"]][["k"]]` experimental comparisons were reported in `r length(unique(SMD_S_L[["SMD_ML"]]$data$ExperimentID_I))` experiments from `r length(unique(SMD_S_L[["SMD_ML"]]$data$StudyId))` publications and involving `r length(unique(SMD_S_L[["SMD_ML"]]$data$Strain))` different animal strain(s). We provide a conventional random effects model to illustrate the data. No subgroup analysis is performed. We also now include these locomotor outcomes under the 'other behavioural' heading below.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE}
forest_metafor(SMD_S_L, "TvC", "Locomotor")
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE}
# since locomotor actvity no longer considered separately, edit df to 'other behavioural'
column_name <- 'outcome_type'
condition <- df$SortLabel == "TvC" & df$outcome_type == "Locomotor"
df[condition, column_name] <- 'Other behavioural'
```
## 2.2 Outcome 2: Fear memory
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_FM <- run_SMD(df, "TvC", "Fear memory")
#TvC levels 4-4-1
```
Multilevel analysis is only performed if there are 5 levels or more for at least one of Strain, Study and Experiment, and that is not the case here. `r SMD_S_FM[["SMD_ML"]][["k"]]` experimental comparisons were reported in `r length(unique(SMD_S_FM[["SMD_ML"]]$data$ExperimentID_I))` experiments from `r length(unique(SMD_S_FM[["SMD_ML"]]$data$StudyId))` publications and involving `r length(unique(SMD_S_FM[["SMD_ML"]]$data$Strain))` different animal strain(s). We provide a conventional random effects model to illustrate the data. No subgroup analysis is performed. We also now include these fear memory outcomes under the 'other behavioural' heading below.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE}
forest_metafor(SMD_S_FM, "TvC", "Fear memory")
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE}
# since fear memory no longer considered separately, edit df to 'other behavioural'
column_name <- 'outcome_type'
condition <- df$SortLabel == "TvC" & df$outcome_type == "Fear memory"
df[condition, column_name] <- 'Other behavioural'
```
## 2.3 Outcome 3: Freezing
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_Fr <- run_SMD(df, "TvC", "Freezing")
# TvC levels 3-3-2
```
Multilevel analysis is only performed if there are 5 levels or more for at least one of Strain, Study and Experiment, and that is not the case here. `r SMD_S_Fr[["SMD_ML"]][["k"]]` experimental comparisons were reported in `r length(unique(SMD_S_Fr[["SMD_ML"]]$data$ExperimentID_I))` experiments from `r length(unique(SMD_S_Fr[["SMD_ML"]]$data$StudyId))` publications and involving `r length(unique(SMD_S_Fr[["SMD_ML"]]$data$Strain))` different animal strain(s). We provide a conventional random effects model to illustrate the data. No subgroup analysis is performed. We also now include these Freezing outcomes under the 'other behavioural' heading below.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE}
forest_metafor(SMD_S_Fr, "TvC", "Freezing")
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE}
# since freezing no longer considered separately, edit df to 'other behavioural'
column_name <- 'outcome_type'
condition <- df$SortLabel == "TvC" & df$outcome_type == "Freezing"
df[condition, column_name] <- 'Other behavioural'
```
## 2.4 Outcome 4: Other behavioural outcomes
### 2.4.1 Risks of bias
Figure 2.4.1 shows the risk of bias summary for studies investigating the effect of exercise on other behavioural outcomes in animals. The risk of bias assessment was performed using the SyRCLE RoB tool.
**Figure 2.4.1**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, fig.height=8}
SyRCLE_RoB_traffic(df, "TvC", "Other behavioural")
```
### 2.4.2 Reporting completeness
Figure 2.4.2 shows the reporting completeness summary for studies investigating the effect of exercise on other behavioural outcomes in animals. The reporting completeness assessment was performed using the ARRIVE guidelines.
**Figure 2.4.2**
```{r message = FALSE, warning=FALSE, eval = TRUE, echo = FALSE, fig.height = 10}
ARRIVE_traffic(df, "TvC", "Other behavioural")
```
### 2.4.3 Meta-analysis
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB <- run_ML_SMD(df, "TvC", "Other behavioural", 0.5)
```
The effect of exercise on other behavioural outcomes in animals using SMD as the effect size is shown in Figure 2.4.3. The pooled estimate for SMD across all individual comparisons is displayed as a diamond shape at the bottom of the plot. Dotted lines indicate the prediction interval of the pooled estimate.
**Figure 2.4.3**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, fig.height=15}
forest_metafor(SMD_S_OB, "TvC", "other behavioural")
```
Exercise interventions had a pooled effect on other behavioural outcomes induced by single prolonged stress of SMD = `r round(SMD_S_OB$SMD_ML[['beta']], 3)`, (95% CI: `r round(SMD_S_OB$SMD_ML[['ci.lb']], 3)` to `r round(SMD_S_OB$SMD_ML[['ci.ub']], 3)`; 95% PrI: `r round(SMD_S_OB[["pred_interval"]][["pi.lb"]], 3)` to `r round(SMD_S_OB[["pred_interval"]][["pi.ub"]], 3)`).
`r SMD_S_OB$SMD_ML[["k"]]` experimental comparisons were reported in `r length(unique(SMD_S_OB[["SMD_ML"]]$data$ExperimentID_I))` experiments from `r length(unique(SMD_S_OB[["SMD_ML"]]$data$StudyId))` publications and involving `r length(unique(SMD_S_OB[["SMD_ML"]]$data$Strain))` different animal strain(s).
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Study | `r ifelse('StudyId' %in% SMD_S_OB[["SMD_ML"]]$s.names, SMD_S_OB[["SMD_ML"]]$s.nlevels[[which(SMD_S_OB[["SMD_ML"]]$s.names == 'StudyId')]], 0)` | `r ifelse('StudyId' %in% SMD_S_OB[["SMD_ML"]]$s.names, SMD_S_OB[["SMD_ML"]]$sigma2[[which(SMD_S_OB[["SMD_ML"]]$s.names == 'StudyId')]], 'NA')` |
| Study x Experiment | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_OB[["SMD_ML"]]$s.names, SMD_S_OB[["SMD_ML"]]$s.nlevels[[which(SMD_S_OB[["SMD_ML"]]$s.names == 'StudyId/ExperimentID_I')]], 0)` | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_OB[["SMD_ML"]]$s.names, SMD_S_OB[["SMD_ML"]]$sigma2[[which(SMD_S_OB[["SMD_ML"]]$s.names == 'StudyId/ExperimentID_I')]], 'NA')` |
### 2.4.4 Subgroup analyses and meta-regressions
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
options(scipen=100, digits=3)
```
For each outcome, the covariates of interest for subgroup analyses and meta-regressions were:
- **Sex**
- **Voluntary or forced exercise**
- **Duration of exercise**
- **Intensity of exercise**
We also conducted subgroup analyses using **(1) SyRCLE Risk of Bias** and **(2) ARRIVE reporting completeness** assessment scores as covariates to evaluate their influence on effect size estimates. These were not specified in the study protocol, but evaluation of risk of bias is required for the Summary of Evidence table, and no studies were considered entirely at low risk of bias or of high reporting completeness to allow such a sensitivity analysis.
The significance (p-value) reported is that for a test of whether the moderators are significantly different one from another, rather than whether the effect is significantly different from 0.
#### **Sex**
Figure 2.4.4.1 displays the estimates for the pooled SMD's when comparisons are stratified by sex of the animal. Whiskers indicate the 95% confidence interval of each estimate. The overall pooled SMD, not stratified by sex, is displayed as a diamond shape at the bottom of the plot, with the 95% prediction interval shown as a red bar.
**Figure 2.4.4.1 - Effect of exercise on other behavioural outcomes by Sex**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_sex <- subgroup_analysis(df, "TvC", "Other behavioural", "Sex", 0.5)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide', fig.height=3}
plot_subgroup_analysis(df, "TvC", "Other behavioural", "Sex", SMD_S_OB_sex, SMD_S_OB)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_sex_noI <- subgroup_SMD(df, "TvC", "Other behavioural", "Sex", 0.5)
```
The p-value for the association between the sex of animal groups used and outcome reported was `r round(SMD_S_OB_sex_noI$QMp,3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Study | `r ifelse('StudyId' %in% SMD_S_OB_sex$analysis$s.names, SMD_S_OB_sex$analysis$s.nlevels[[which(SMD_S_OB_sex$analysis$s.names == 'StudyId')]], 0)` | `r ifelse('StudyId' %in% SMD_S_OB_sex$analysis$s.names, SMD_S_OB_sex$analysis$sigma2[[which(SMD_S_OB_sex$analysis$s.names == 'StudyId')]], 'NA')` |
| Study x Experiment | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_OB_sex$analysis$s.names, SMD_S_OB_sex$analysis$s.nlevels[[which(SMD_S_OB_sex$analysis$s.names == 'StudyId/ExperimentID_I')]], 0)` | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_OB_sex$analysis$s.names, SMD_S_OB_sex$analysis$sigma2[[which(SMD_S_OB_sex$analysis$s.names == 'StudyId/ExperimentID_I')]], 'NA')` |
#### **Voluntary or forced exercise**
Figure 2.4.4.2 displays the estimates for the pooled SMD when comparisons are stratified by whether the exercise was voluntary or forced (VoF). Whiskers indicate the 95% confidence interval of each estimate. The overall pooled SMD, not stratified by exercise type, is displayed as a diamond shape at the bottom of the plot, with the 95% prediction interval shown as a red bar.
**Figure 2.4.4.2 - Effect of exercise on other behavioural outcomes by voluntary or forced exercise**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_VoF <- subgroup_analysis(df, "TvC", "Other behavioural", "VoF", 0.5)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide', fig.height=3}
plot_subgroup_analysis(df, "TvC", "Other behavioural", "VoF", SMD_S_OB_VoF, SMD_S_OB)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_VoF_noI <- subgroup_SMD(df, "TvC", "Other behavioural", "VoF", 0.5)
```
The p-value for the association between the VoF of animal groups used and outcome reported was `r round(SMD_S_OB_VoF_noI$QMp,3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Study | `r ifelse('StudyId' %in% SMD_S_OB_VoF$analysis$s.names, SMD_S_OB_VoF$analysis$s.nlevels[[which(SMD_S_OB_VoF$analysis$s.names == 'StudyId')]], 0)` | `r ifelse('StudyId' %in% SMD_S_OB_VoF$analysis$s.names, SMD_S_OB_VoF$analysis$sigma2[[which(SMD_S_OB_VoF$analysis$s.names == 'StudyId')]], 'NA')` |
| Study x Experiment | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_OB_VoF$analysis$s.names, SMD_S_OB_VoF$analysis$s.nlevels[[which(SMD_S_OB_VoF$analysis$s.names == 'StudyId/ExperimentID_I')]], 0)` | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_OB_VoF$analysis$s.names, SMD_S_OB_VoF$analysis$sigma2[[which(SMD_S_OB_VoF$analysis$s.names == 'StudyId/ExperimentID_I')]], 'NA')` |
#### **Duration of treatment**
We provide a meta-regression of the number of weeks of treatment as a continuous variable.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_DOE <- metaregression_analysis(df, "TvC", "Other behavioural", "DurationOfTreatmentWeeks1", 0.5)
```
**Figure 2.4.4.3 - Effect of exercise on other behavioural outcomes by Duration of treatment**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE,}
SMD_S_OB_DOE$regression_plot
```
The p-value for the association between duration of treatment and outcome reported was `r round(SMD_S_OB_DOE$metaregression_summary$pval[2],3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Strain | `r ifelse('Strain' %in% SMD_S_OB_DOE$metaregression_summary$s.names, SMD_S_OB_DOE$metaregression_summary$s.nlevels[[which(SMD_S_OB_DOE$metaregression_summary$s.names == 'Strain')]], 0)` | `r ifelse('Strain' %in% SMD_S_OB_DOE$metaregression_summary$s.names, SMD_S_OB_DOE$metaregression_summary$sigma2[[which(SMD_S_OB_DOE$metaregression_summary$s.names == 'Strain')]], NA)` |
| Study x Strain | `r ifelse('Strain/StudyId' %in% SMD_S_OB_DOE$metaregression_summary$s.names, SMD_S_OB_DOE$metaregression_summary$s.nlevels[[which(SMD_S_OB_DOE$metaregression_summary$s.names == 'Strain/StudyId')]], 0)` | `r ifelse('Strain/StudyId' %in% SMD_S_OB_DOE$metaregression_summary$s.names, SMD_S_OB_DOE$metaregression_summary$sigma2[[which(SMD_S_OB_DOE$metaregression_summary$s.names == 'Strain/StudyId')]], NA)` |
| Study x Strain x Experiment | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_OB_DOE$metaregression_summary$s.names, SMD_S_OB_DOE$metaregression_summary$s.nlevels[[which(SMD_S_OB_DOE$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 0)` | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_OB_DOE$metaregression_summary$s.names, SMD_S_OB_DOE$metaregression_summary$sigma2[[which(SMD_S_OB_DOE$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 'NA')` |
#### **Exercise Intensity**
We provide a meta-regression where exercise intensity (expressed as m/min) is considered as a continuous variable.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_EI <- metaregression_analysis(df, "TvC", "Other behavioural", "EI1", 0.5)
```
**Figure 2.4.4.4 - Effect of exercise on other behavioural outcomes by exercise intensity**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE,}
SMD_S_OB_EI$regression_plot
```
The p-value for the association between exercise intensity and the outcome reported was `r round(SMD_S_OB_EI$metaregression_summary$pval[2],3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Strain | `r ifelse('Strain' %in% SMD_S_OB_EI$metaregression_summary$s.names, SMD_S_OB_EI$metaregression_summary$s.nlevels[[which(SMD_S_OB_EI$metaregression_summary$s.names == 'Strain')]], 0)` | `r ifelse('Strain' %in% SMD_S_OB_EI$metaregression_summary$s.names, SMD_S_OB_EI$metaregression_summary$sigma2[[which(SMD_S_OB_EI$metaregression_summary$s.names == 'Strain')]], NA)` |
| Study x Strain | `r ifelse('Strain/StudyId' %in% SMD_S_OB_EI$metaregression_summary$s.names, SMD_S_OB_EI$metaregression_summary$s.nlevels[[which(SMD_S_OB_EI$metaregression_summary$s.names == 'Strain/StudyId')]], 0)` | `r ifelse('Strain/StudyId' %in% SMD_S_OB_EI$metaregression_summary$s.names, SMD_S_OB_EI$metaregression_summary$sigma2[[which(SMD_S_OB_EI$metaregression_summary$s.names == 'Strain/StudyId')]], NA)` |
| Study x Strain x Experiment | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_OB_EI$metaregression_summary$s.names, SMD_S_OB_EI$metaregression_summary$s.nlevels[[which(SMD_S_OB_EI$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 0)` | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_OB_EI$metaregression_summary$s.names, SMD_S_OB_EI$metaregression_summary$sigma2[[which(SMD_S_OB_EI$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 'NA')` |
#### **Total exercise**
We provide a meta-regression where the total exercise as the product of the number of sessions (n), session duration (min), and session intensity (km/min), with total exercise expressed in km (n \* min \* km/min), is considered as a continuous variable.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_TE <- metaregression_analysis1(df, "TvC", "Other behavioural", "TotalExercise1", 0.5)
```
**Figure 2.4.4.5**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE,}
SMD_S_OB_TE$regression_plot
```
The p-value for the association between total exercise and the outcome reported was `r ifelse(SMD_S_OB_TE$metaregression_summary$pval[2] < 0.001, "<0.001", sprintf("%.3f", SMD_S_OB_TE$metaregression_summary$pval[2]))`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Strain | `r ifelse('Strain' %in% SMD_S_OB_TE$metaregression_summary$s.names, SMD_S_OB_TE$metaregression_summary$s.nlevels[[which(SMD_S_OB_TE$metaregression_summary$s.names == 'Strain')]], 0)` | `r ifelse('Strain' %in% SMD_S_OB_TE$metaregression_summary$s.names, SMD_S_OB_TE$metaregression_summary$sigma2[[which(SMD_S_OB_TE$metaregression_summary$s.names == 'Strain')]], NA)` |
| Study x Strain | `r ifelse('Strain/StudyId' %in% SMD_S_OB_TE$metaregression_summary$s.names, SMD_S_OB_TE$metaregression_summary$s.nlevels[[which(SMD_S_OB_TE$metaregression_summary$s.names == 'Strain/StudyId')]], 0)` | `r ifelse('Strain/StudyId' %in% SMD_S_OB_TE$metaregression_summary$s.names, SMD_S_OB_TE$metaregression_summary$sigma2[[which(SMD_S_OB_TE$metaregression_summary$s.names == 'Strain/StudyId')]], NA)` |
| Study x Strain x Experiment | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_OB_TE$metaregression_summary$s.names, SMD_S_OB_TE$metaregression_summary$s.nlevels[[which(SMD_S_OB_TE$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 0)` | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_OB_TE$metaregression_summary$s.names, SMD_S_OB_TE$metaregression_summary$sigma2[[which(SMD_S_OB_TE$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 'NA')` |
#### **SyRCLE RoB assessment considered as a categorical variable**
Figure 2.4.4.6 displays the estimates for the pooled SMD when comparisons are stratified by how many of the SyRCLE risk of bias assessment criteria (of which there are 10) the experiment met. Whiskers indicate the 95% confidence interval of each estimate. The overall pooled SMD, not stratified by SyRCLE Risk of Bias, is displayed as a diamond shape at the bottom of the plot, with the 95% prediction interval shown as a red bar.
**Figure 2.4.4.6 - Effect of exercise on other behavioural outcomes by SyRCLE** **risk of bias criteria**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide', out.width = "100%"}
SMD_S_OB_SyRCLERoB <- subgroup_analysis(df, "TvC", "Other behavioural", "RoBScore", 0.5)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide', fig.height=3}
plot_subgroup_analysis(df, "TvC", "Other behavioural", "RoBScore", SMD_S_OB_SyRCLERoB, SMD_S_OB)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_SyRCLERoB_noI <- subgroup_SMD(df, "TvC", "Other behavioural", "RoBScore", 0.5)
```
The p-value for the association between SyRCLE Risks of Bias reporting and the outcome reported was `r round(SMD_S_OB_SyRCLERoB_noI$QMp,3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Study | `r ifelse('StudyId' %in% SMD_S_OB_SyRCLERoB$analysis$s.names, SMD_S_OB_SyRCLERoB$analysis$s.nlevels[[which(SMD_S_OB_SyRCLERoB$analysis$s.names == 'StudyId')]], 0)` | `r ifelse('StudyId' %in% SMD_S_OB_SyRCLERoB$analysis$s.names, SMD_S_OB_SyRCLERoB$analysis$sigma2[[which(SMD_S_OB_SyRCLERoB$analysis$s.names == 'StudyId')]], 'NA')` |
| Study x Experiment | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_OB_SyRCLERoB$analysis$s.names, SMD_S_OB_SyRCLERoB$analysis$s.nlevels[[which(SMD_S_OB_SyRCLERoB$analysis$s.names == 'StudyId/ExperimentID_I')]], 0)` | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_OB_SyRCLERoB$analysis$s.names, SMD_S_OB_SyRCLERoB$analysis$sigma2[[which(SMD_S_OB_SyRCLERoB$analysis$s.names == 'StudyId/ExperimentID_I')]], 'NA')` |
#### **SyRCLE RoB assessment considering those studies where any item is at low risk of bias**
Figure 2.4.4.7 displays the estimates for the pooled SMD when comparisons are stratified by whether of not any of the SyRCLE Risk of bias domains were rated as low risk of bias. Whiskers indicate the 95% confidence interval of each estimate. The overall pooled SMD, not stratified by SyRCLE Risk of Bias, is displayed as a diamond shape at the bottom of the plot, with the 95% prediction interval shown as a red bar.
**Figure 2.4.4.7 - Effect of exercise on other behavioural outcomes by low SyRCLE** **RoB**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide', out.width = "100%"}
SMD_S_OB_SyRCLERoBTF <- subgroup_analysis(df, "TvC", "Other behavioural", "RoBTF", 0.5)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide', fig.height=3}
plot_subgroup_analysis(df, "TvC", "Other behavioural", "RoBTF", SMD_S_OB_SyRCLERoBTF, SMD_S_OB)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_SyRCLERoBTF_noI <- subgroup_SMD(df, "TvC", "Other behavioural", "RoBTF", 0.5)
```
The p-value for the association between low SyRCLE Risks of Bias reporting and the outcome reported was `r round(SMD_S_OB_SyRCLERoBTF_noI$QMp,3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Study | `r ifelse('StudyId' %in% SMD_S_OB_SyRCLERoBTF$analysis$s.names, SMD_S_OB_SyRCLERoBTF$analysis$s.nlevels[[which(SMD_S_OB_SyRCLERoBTF$analysis$s.names == 'StudyId')]], 0)` | `r ifelse('StudyId' %in% SMD_S_OB_SyRCLERoBTF$analysis$s.names, SMD_S_OB_SyRCLERoBTF$analysis$sigma2[[which(SMD_S_OB_SyRCLERoBTF$analysis$s.names == 'StudyId')]], 'NA')` |
| Study x Experiment | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_OB_SyRCLERoBTF$analysis$s.names, SMD_S_OB_SyRCLERoBTF$analysis$s.nlevels[[which(SMD_S_OB_SyRCLERoBTF$analysis$s.names == 'StudyId/ExperimentID_I')]], 0)` | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_OB_SyRCLERoBTF$analysis$s.names, SMD_S_OB_SyRCLERoBTF$analysis$sigma2[[which(SMD_S_OB_SyRCLERoBTF$analysis$s.names == 'StudyId/ExperimentID_I')]], 'NA')` |
#### **ARRIVE reporting completeness guidelines**
We provide a meta-regression where the number of ARRIVE items met is considered as a continuous variable.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_ARR2 <- metaregression_analysis(df, "TvC", "Other behavioural", "ARRIVEScore", 0.5)
```
**Figure 2.4.4.8 - Effect of exercise on other behavioural outcomes by ARRIVE reporting completeness**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE,}
SMD_S_OB_ARR2$regression_plot
```
The p-value for the association between ARRIVE reporting completeness and outcome reported was `r round(SMD_S_OB_ARR2$metaregression_summary$pval[2],3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Strain | `r ifelse('Strain' %in% SMD_S_OB_ARR2$metaregression_summary$s.names, SMD_S_OB_ARR2$metaregression_summary$s.nlevels[[which(SMD_S_OB_ARR2$metaregression_summary$s.names == 'Strain')]], 0)` | `r ifelse('Strain' %in% SMD_S_OB_ARR2$metaregression_summary$s.names, SMD_S_OB_ARR2$metaregression_summary$sigma2[[which(SMD_S_OB_ARR2$metaregression_summary$s.names == 'Strain')]], NA)` |
| Study x Strain | `r ifelse('Strain/StudyId' %in% SMD_S_OB_ARR2$metaregression_summary$s.names, SMD_S_OB_ARR2$metaregression_summary$s.nlevels[[which(SMD_S_OB_ARR2$metaregression_summary$s.names == 'Strain/StudyId')]], 0)` | `r ifelse('Strain/StudyId' %in% SMD_S_OB_ARR2$metaregression_summary$s.names, SMD_S_OB_ARR2$metaregression_summary$sigma2[[which(SMD_S_OB_ARR2$metaregression_summary$s.names == 'Strain/StudyId')]], NA)` |
| Study x Strain x Experiment | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_OB_ARR2$metaregression_summary$s.names, SMD_S_OB_ARR2$metaregression_summary$s.nlevels[[which(SMD_S_OB_ARR2$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 0)` | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_OB_ARR2$metaregression_summary$s.names, SMD_S_OB_ARR2$metaregression_summary$sigma2[[which(SMD_S_OB_ARR2$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 'NA')` |
#### **Heterogeneity explained by covariates (Effect of exercise on other behaviours)**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide', out.width = "100%"}
SMD_S_OB_sexI <- subgroup_SMDI(df, "TvC", "Other behavioural", "Sex", 0.5)
SMD_S_OB_VoFI <- subgroup_SMDI(df, "TvC", "Other behavioural", "VoF", 0.5)
SMD_S_OB_SyRCLERoBI <- subgroup_SMDI(df, "TvC", "Other behavioural", "RoBScore", 0.5)
```
The table below shows which of the covariates, if any, explain some of the heterogeneity observed in the effect of exercise on other behaviours. We present marginal R^2^, which measures the proportion of variance explained by including moderators in the model (the % change in the between-studies variance when the covariate is included in the model, in other words the % of the heterogeneity explained by the variable). The coefficients are derived form an RMA model fitted with an intercept (and so represent, for each category, the point estimate and 95% CIs of the effect in that category).
| Moderator | Category | $\beta$ | 95% CI | Marginal R^2^ (%) |
|:-----------:|:-----------:|:-----------:|:--------------------:|:-----------:|
| Overall effect | \- | `r SMD_S_OB[["SMD_ML"]]$beta[1]` | `r SMD_S_OB[["SMD_ML"]]$ci.lb` to `r SMD_S_OB[["SMD_ML"]]$ci.ub` | \- |
| Sex | \- | \- | \- | `r round((r2_ml(SMD_S_OB_sexI)[1]*100),1)`% |
| \- | *Female* | `r SMD_S_OB_sexI$beta[1]` | `r SMD_S_OB_sexI$ci.lb[1]` to `r SMD_S_OB_sexI$ci.ub[1]` | \- |
| \- | *Male* | `r SMD_S_OB_sexI$beta[2]` | `r SMD_S_OB_sexI$ci.lb[2]` to `r SMD_S_OB_sexI$ci.ub[2]` | \- |
| Voluntary or forced | \- | \- | \- | `r round((r2_ml(SMD_S_OB_VoFI)[1]*100),1)`% |
| \- | *Forced* | `r SMD_S_OB_VoFI$beta[1]` | `r SMD_S_OB_VoFI$ci.lb[1]` to `r SMD_S_OB_VoFI$ci.ub[1]` | \- |
| \- | *Voluntary* | `r SMD_S_OB_VoFI$beta[2]` | `r SMD_S_OB_VoFI$ci.lb[2]` to `r SMD_S_OB_VoFI$ci.ub[2]` | \- |
| Duration of treatment | \- | \- | \- | `r round((r2_ml(SMD_S_OB_DOE$metaregression)[1]*100),1)`% |
| \- | *per weeks of treatment increase* | `r SMD_S_OB_DOE$metaregression_summary$beta[2]` | `r SMD_S_OB_DOE$metaregression_summary$ci.lb[2]` to `r SMD_S_OB_DOE$metaregression_summary$ci.ub[2]` | \- |
| Exercise intensity | \- | \- | \- | `r round((r2_ml(SMD_S_OB_EI$metaregression)[1]*100),1)`% |
| \- | *per unit (m/min) increase* | `r SMD_S_OB_EI$metaregression_summary$beta[2]` | `r SMD_S_OB_EI$metaregression_summary$ci.lb[2]` to `r SMD_S_OB_EI$metaregression_summary$ci.ub[2]` | \- |
| Total exercise | \- | \- | \- | `r round((r2_ml(SMD_S_OB_TE$metaregression)[1]*100),1)`% |
| \- | *per km increase* | `r SMD_S_OB_TE$metaregression_summary$beta[2]` | `r SMD_S_OB_TE$metaregression_summary$ci.lb[2]` to `r SMD_S_OB_TE$metaregression_summary$ci.ub[2]` | \- |
| Risk of Bias | \- | \- | \- | `r round((r2_ml(SMD_S_OB_SyRCLERoBI)[1]*100),1)`% |
| \- | *0 criteria met* | `r SMD_S_OB_SyRCLERoBI$beta[1]` | `r SMD_S_OB_SyRCLERoBI$ci.lb[1]` to `r SMD_S_OB_SyRCLERoBI$ci.ub[1]` | \- |
| \- | *1 criteria met* | `r SMD_S_OB_SyRCLERoBI$beta[2]` | `r SMD_S_OB_SyRCLERoBI$ci.lb[2]` to `r SMD_S_OB_SyRCLERoBI$ci.ub[2]` | \- |
| Reporting completeness | \- | \- | \- | `r round((r2_ml(SMD_S_OB_ARR2$metaregression)[1]*100),1)`% |
| \- | *per unit increase* | `r SMD_S_OB_ARR2$metaregression_summary$beta[2]` | `r SMD_S_OB_ARR2$metaregression_summary$ci.lb[2]` to `r SMD_S_OB_ARR2$metaregression_summary$ci.ub[2]` | \- |
### 2.4.5 Sensitivity Analyses
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
options(scipen = 100, digits = 2)
```
We examine the robustness of the findings for other behaviours by performing the following sensitivity analyses.
#### Imputed rho values of 0.2 and 0.8
In the previous analyses for the effect of exercise on other behavioural outcomes, we imputed a $\rho$ value - the imputed within-study correlation between observed effect sizes - of 0.5. Here, we examine the effect of imputing $\rho$ values of 0.2 and 0.8.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_rho0.2 <- run_ML_SMD(df, "TvC", "Other behavioural", 0.2)
```
When the $\rho$ value is assumed to be 0.2, exercise had a pooled effect on other behavioural outcomes of **SMD = `r round(SMD_S_OB_rho0.2[["SMD_ML"]][["beta"]],3)`** (95% CI: `r round(SMD_S_OB_rho0.2[["SMD_ML"]][["ci.lb"]],3)` to `r round(SMD_S_OB_rho0.2[["SMD_ML"]][["ci.ub"]],3)`; 95% PrI: `r round(SMD_S_OB_rho0.2[["pred_interval"]][["pi.lb"]],3)` to `r round(SMD_S_OB_rho0.2[["pred_interval"]][["pi.ub"]],3)`).
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_OB_rho0.8 <- run_ML_SMD(df, "TvC", "Other behavioural", 0.8)
```
When the $\rho$ value is assumed to be 0.8, exercise had a pooled effect on other behavioural outcomes of **SMD = `r SMD_S_OB_rho0.8[["SMD_ML"]][["beta"]]`** (95% CI: `r SMD_S_OB_rho0.8[["SMD_ML"]][["ci.lb"]]` to `r SMD_S_OB_rho0.8[["SMD_ML"]][["ci.ub"]]`; 95% PrI: `r round(SMD_S_OB_rho0.8[["pred_interval"]][["pi.lb"]],3)` to `r round(SMD_S_OB_rho0.8[["pred_interval"]][["pi.ub"]],3)`).
For reference the pooled effect size when rho is assumed to be 0.5 is `r SMD_S_OB[["SMD_ML"]][["beta"]]` (95% CI: `r SMD_S_OB[["SMD_ML"]][["ci.lb"]]` to `r SMD_S_OB[["SMD_ML"]][["ci.ub"]]`).
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
df_S_OB <- filter_experiment_outcome_type(df, "TvC", "Other behavioural")
```
#### NMD
For behavioural outcomes, an NMD was calculable for `r df_S_OB %>% filter(NMD_possible == "TRUE") %>% nrow()` out of `r nrow(df_S_OB)` comparisons, i.e. `r ((df_S_OB %>% filter(NMD_possible == "TRUE") %>% nrow())/(nrow(df_S_OB)))*100` % of comparisons.
The effect of exercise on other behavioural outcomes in animals using NMD as the effect size is shown in Figure 2.4.5. The pooled estimate for NMD across all individual comparisons is displayed as a diamond shape at the bottom of the plot. Dotted lines indicate the prediction interval of the pooled estimate.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
NMD_S_OB <- run_ML_NMD(df, "TvC", "Other behavioural", 0.5)
```
**Figure 2.4.5**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, fig.height = 14}
forest_metafor_NMD(NMD_S_OB, "Other behavioural")
```
Exercise had a pooled effect on other behavioural outcomes of NMD = `r NMD_S_OB[["beta"]][1]` (95% CI: `r NMD_S_OB[["ci.lb"]]` to `r NMD_S_OB[["ci.ub"]]`) with a prediction interval of `r predict(NMD_S_OB)$pi.lb` to `r predict(NMD_S_OB)$pi.ub`). For reference the pooled effect size for SMD was `r SMD_S_OB[["SMD_ML"]][["beta"]][1]` (95% CI: `r SMD_S_OB[["SMD_ML"]][["ci.lb"]]` to `r SMD_S_OB[["SMD_ML"]][["ci.ub"]]`).
`r NMD_S_OB[["k"]]` experimental comparisons were reported in `r NMD_S_OB[["s.nlevels"]][3]` experiments reported from `r length(unique(NMD_S_OB$data$StudyId))` publications and involving `r NMD_S_OB[["s.nlevels"]][1]` different animal strains.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:-------------:|:---------------------------:|:---------------------------:|
| Strain | `r ifelse('Strain' %in% NMD_S_OB$s.names, NMD_S_OB$s.nlevels[[which(NMD_S_OB$s.names == 'Strain')]], 0)` | `r ifelse('Strain' %in% NMD_S_OB$s.names, NMD_S_OB$sigma2[[which(NMD_S_OB$s.names == 'Strain')]], NA)` |
| Study x Strain | `r ifelse('Strain/StudyId' %in% NMD_S_OB$s.names, NMD_S_OB$s.nlevels[[which(NMD_S_OB$s.names == 'Strain/StudyId')]], 0)` | `r ifelse('Strain/StudyId' %in% NMD_S_OB$s.names, NMD_S_OB$sigma2[[which(NMD_S_OB$s.names == 'Strain/StudyId')]], NA)` |
| Study x Strain x Experiment | `r ifelse('Strain/StudyId/ExperimentID_I' %in% NMD_S_OB$s.names, NMD_S_OB$s.nlevels[[which(NMD_S_OB$s.names == 'Strain/StudyId/ExperimentID_I')]], 0)` | `r ifelse('Strain/StudyId/ExperimentID_I' %in% NMD_S_OB$s.names, NMD_S_OB$sigma2[[which(NMD_S_OB$s.names == 'Strain/StudyId/ExperimentID_I')]], 'NA')` |
### 2.4.6 Reporting bias/small-study effects
Because of the relationship between SMD effect sizes and variance inherent in their calculation, where study size is small the standard approach to seeking evidence of small-study effects (regression based tests including Egger's regression test for multilevel meta-analysis) can lead to over-estimation of small-study effect (see for instance [10.7554/eLife.24260](10.7554/eLife.24260)). To address this we used Egger's regression test for multilevel meta-analysis, with regression of SMD effect size against 1/√N, where N is the total number of animals involved in an experiment.
```{r warning=FALSE, eval = TRUE, echo = FALSE}
run_sse_plot_SMD_L(df, 'TvC', 'Other behavioural')
#run_sse_NMD(df)
```
Egger regression based on `r run_sse_SMD_L(df, 'TvC', 'Other behavioural')[["k"]]` studies of modelling of depression where sucrose preference was measured showed a coefficient for a small-study effect of `r run_sse_SMD_L(df, 'TvC', 'Other behavioural')[["beta"]][2]` (95% CI: `r run_sse_SMD_L(df, 'TvC', 'Other behavioural')[["ci.lb"]][2]` to `r run_sse_SMD_L(df, 'TvC', 'Other behavioural')[["ci.ub"]][2]`; p = `r ifelse(run_sse_SMD_L(df, 'TvC', 'Other behavioural')[["pval"]][2] < 0.001, "<0.001", sprintf("%.3f", run_sse_SMD_L(df, 'TvC', 'Other behavioural')[["pval"]][2]))`).
## 2.5 Outcome 5: BDNF
### 2.5.1 Risks of bias
Figure 2.5.1 shows the risk of bias summary for studies investigating the effect of exercise on BDNF in animals. The risk of bias assessment was performed using the SyRCLE RoB tool.
**Figure 2.5.1**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE}
SyRCLE_RoB_traffic(df, "TvC", "BDNF")
```
### 2.5.2 Reporting completeness
Figure 2.5.2 shows the reporting completeness summary for studies investigating the effect of exercise on BDNF in animals. The reporting completeness assessment was performed using the ARRIVE guidelines.
**Figure 2.5.2**
```{r message = FALSE, warning=FALSE, eval = TRUE, echo = FALSE, fig.height=8}
ARRIVE_traffic(df, "TvC", "BDNF")
```
### 2.5.3 Meta-analysis
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_B <- run_ML_SMD(df, "TvC", "BDNF", 0.5)
```
The effect of exercise on BDNF in animals using SMD as the effect size is shown in Figure 2.5.3. The pooled estimate for SMD across all individual comparisons is displayed as a diamond shape at the bottom of the plot. Dotted lines indicate the prediction interval of the pooled estimate.
**Figure 2.5.3**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, fig.height=8}
forest_metafor(SMD_S_B, "TvC", "BDNF")
```
Exercise had a pooled effect on BDNF of SMD = `r round(SMD_S_B[["SMD_ML"]][['beta']], 3)` , (95% CI: `r round(SMD_S_B[["SMD_ML"]][['ci.lb']], 3)` to `r round(SMD_S_B[["SMD_ML"]][['ci.ub']], 3)`; 95% PrI: `r round(SMD_S_B[["pred_interval"]][["pi.lb"]], 3)` to `r round(SMD_S_B[["pred_interval"]][["pi.ub"]], 3)`).
`r SMD_S_B[["SMD_ML"]][["k"]]` experimental comparisons were reported in `r length(unique(SMD_S_B[["SMD_ML"]]$data$ExperimentID_I))` experiments from `r length(unique(SMD_S_B[["SMD_ML"]]$data$StudyId))` publications and involving `r length(unique(SMD_S_B[["SMD_ML"]]$data$Strain))` different animal strain(s).
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Study | `r ifelse('StudyId' %in% SMD_S_B[["SMD_ML"]]$s.names, SMD_S_B[["SMD_ML"]]$s.nlevels[[which(SMD_S_B[["SMD_ML"]]$s.names == 'StudyId')]], 0)` | `r ifelse('StudyId' %in% SMD_S_B[["SMD_ML"]]$s.names, SMD_S_B[["SMD_ML"]]$sigma2[[which(SMD_S_B[["SMD_ML"]]$s.names == 'StudyId')]], 'NA')` |
| Study x Experiment | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_B[["SMD_ML"]]$s.names, SMD_S_B[["SMD_ML"]]$s.nlevels[[which(SMD_S_B[["SMD_ML"]]$s.names == 'StudyId/ExperimentID_I')]], 0)` | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_B[["SMD_ML"]]$s.names, SMD_S_B[["SMD_ML"]]$sigma2[[which(SMD_S_B[["SMD_ML"]]$s.names == 'StudyId/ExperimentID_I')]], 'NA')` |
### 2.5.4 Subgroup analyses and meta-regressions
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
options(scipen=100, digits=3)
```
For each outcome, the covariates of interest for subgroup analyses and meta-regressions were:
- **Sex**
- **Voluntary or forced exercise**
- **Duration of exercise**
- **Intensity of exercise**
We also conducted subgroup analyses using **(1) SyRCLE Risk of Bias** and **(2) ARRIVE reporting completeness** assessment scores as covariates to evaluate their influence on effect size estimates. These were not specified in the study protocol, but evaluation of risk of bias is required for the Summary of Evidence table, and no studies were considered entirely at low risk of bias or of high reporting completeness to allow such a sensitivity analysis.
The significance (p-value) reported is that for a test of whether the moderators are significantly different one from another, rather than whether the effect is significantly different from 0.
#### **Sex**
Figure 2.5.4.1 displays the estimates for the pooled SMD when comparisons are stratified by sex of the animal. Whiskers indicate the 95% confidence interval of each estimate. The overall pooled SMD, not stratified by sex, is displayed as a diamond shape at the bottom of the plot, with the 95% prediction interval shown as a red bar.
**Figure 2.5.4.1 - Effect of exercise on BDNF by Sex**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_B_sex <- subgroup_analysis(df, "TvC", "BDNF", "Sex", 0.5)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide', fig.height=3}
plot_subgroup_analysis(df, "TvC", "BDNF", "Sex", SMD_S_B_sex, SMD_S_B)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_B_sex_noI <- subgroup_SMD(df, "TvC", "BDNF", "Sex", 0.5)
```
The p-value for the association between the sex of animal groups used and outcome reported was `r round(SMD_S_B_sex_noI$QMp,3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Study | `r ifelse('StudyId' %in% SMD_S_B_sex$analysis$s.names, SMD_S_B_sex$analysis$s.nlevels[[which(SMD_S_B_sex$analysis$s.names == 'StudyId')]], 0)` | `r ifelse('StudyId' %in% SMD_S_B_sex$analysis$s.names, SMD_S_B_sex$analysis$sigma2[[which(SMD_S_B_sex$analysis$s.names == 'StudyId')]], 'NA')` |
| Study x Experiment | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_B_sex$analysis$s.names, SMD_S_B_sex$analysis$s.nlevels[[which(SMD_S_B_sex$analysis$s.names == 'StudyId/ExperimentID_I')]], 0)` | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_B_sex$analysis$s.names, SMD_S_B_sex$analysis$sigma2[[which(SMD_S_B_sex$analysis$s.names == 'StudyId/ExperimentID_I')]], 'NA')` |
#### **Voluntary or forced exercise**
Figure 2.5.4.2 displays the estimates for the pooled SMD when comparisons are stratified by whether the exercise was voluntary or forced (VoF). Whiskers indicate the 95% confidence interval of each estimate. The overall pooled SMD, not stratified by exercise type, is displayed as a diamond shape at the bottom of the plot, with the 95% prediction interval shown as a red bar.
**Figure 2.5.4.2 - Effect of exercise on other BDNF by voluntary or forced exercise**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_B_VoF <- subgroup_analysis(df, "TvC", "BDNF", "VoF", 0.5)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide', fig.height=3}
plot_subgroup_analysis(df, "TvC", "BDNF", "VoF", SMD_S_B_VoF, SMD_S_B)
```
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_B_VoF_noI <- subgroup_SMD(df, "TvC", "BDNF", "VoF", 0.5)
```
The p-value for the association between the VoF of animal groups used and outcome reported was `r round(SMD_S_B_VoF_noI$QMp,3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Study | `r ifelse('StudyId' %in% SMD_S_B_VoF$analysis$s.names, SMD_S_B_VoF$analysis$s.nlevels[[which(SMD_S_B_VoF$analysis$s.names == 'StudyId')]], 0)` | `r ifelse('StudyId' %in% SMD_S_B_VoF$analysis$s.names, SMD_S_B_VoF$analysis$sigma2[[which(SMD_S_B_VoF$analysis$s.names == 'StudyId')]], 'NA')` |
| Study x Experiment | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_B_VoF$analysis$s.names, SMD_S_B_VoF$analysis$s.nlevels[[which(SMD_S_B_VoF$analysis$s.names == 'StudyId/ExperimentID_I')]], 0)` | `r ifelse('StudyId/ExperimentID_I' %in% SMD_S_B_VoF$analysis$s.names, SMD_S_B_VoF$analysis$sigma2[[which(SMD_S_B_VoF$analysis$s.names == 'StudyId/ExperimentID_I')]], 'NA')` |
#### **Duration of treatment**
We provide a meta-regression of the number of weeks of treatment as a continuous variable.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_B_DOE <- metaregression_analysis(df, "TvC", "BDNF", "DurationOfTreatmentWeeks1", 0.5)
```
**Figure 2.5.4.3 - Effect of exercise on BDNF by duration of treatment**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE,}
SMD_S_B_DOE$regression_plot
```
The p-value for the association between duration of treatment and outcome reported was `r round(SMD_S_B_DOE$metaregression_summary$pval[2],3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Strain | `r ifelse('Strain' %in% SMD_S_B_DOE$metaregression_summary$s.names, SMD_S_B_DOE$metaregression_summary$s.nlevels[[which(SMD_S_B_DOE$metaregression_summary$s.names == 'Strain')]], 0)` | `r ifelse('Strain' %in% SMD_S_B_DOE$metaregression_summary$s.names, SMD_S_B_DOE$metaregression_summary$sigma2[[which(SMD_S_B_DOE$metaregression_summary$s.names == 'Strain')]], NA)` |
| Study x Strain | `r ifelse('Strain/StudyId' %in% SMD_S_B_DOE$metaregression_summary$s.names, SMD_S_B_DOE$metaregression_summary$s.nlevels[[which(SMD_S_B_DOE$metaregression_summary$s.names == 'Strain/StudyId')]], 0)` | `r ifelse('Strain/StudyId' %in% SMD_S_B_DOE$metaregression_summary$s.names, SMD_S_B_DOE$metaregression_summary$sigma2[[which(SMD_S_B_DOE$metaregression_summary$s.names == 'Strain/StudyId')]], NA)` |
| Study x Strain x Experiment | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_B_DOE$metaregression_summary$s.names, SMD_S_B_DOE$metaregression_summary$s.nlevels[[which(SMD_S_B_DOE$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 0)` | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_B_DOE$metaregression_summary$s.names, SMD_S_B_DOE$metaregression_summary$sigma2[[which(SMD_S_B_DOE$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 'NA')` |
#### **Exercise Intensity**
We provide a meta-regression where exercise intensity (expressed at m/min) is considered as a continuous variable.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_B_EI <- metaregression_analysis(df, "TvC", "BDNF", "EI1", 0.5)
```
**Figure 2.5.4.4 - Effect of exercise on BDNF by exercise intensity**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE,}
SMD_S_B_EI$regression_plot
```
The p-value for the association between exercise intensity and outcome reported was `r round(SMD_S_B_EI$metaregression_summary$pval[2],3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Strain | `r ifelse('Strain' %in% SMD_S_B_EI$metaregression_summary$s.names, SMD_S_B_EI$metaregression_summary$s.nlevels[[which(SMD_S_B_EI$metaregression_summary$s.names == 'Strain')]], 0)` | `r ifelse('Strain' %in% SMD_S_B_EI$metaregression_summary$s.names, SMD_S_B_EI$metaregression_summary$sigma2[[which(SMD_S_B_EI$metaregression_summary$s.names == 'Strain')]], NA)` |
| Study x Strain | `r ifelse('Strain/StudyId' %in% SMD_S_B_EI$metaregression_summary$s.names, SMD_S_B_EI$metaregression_summary$s.nlevels[[which(SMD_S_B_EI$metaregression_summary$s.names == 'Strain/StudyId')]], 0)` | `r ifelse('Strain/StudyId' %in% SMD_S_B_EI$metaregression_summary$s.names, SMD_S_B_EI$metaregression_summary$sigma2[[which(SMD_S_B_EI$metaregression_summary$s.names == 'Strain/StudyId')]], NA)` |
| Study x Strain x Experiment | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_B_EI$metaregression_summary$s.names, SMD_S_B_EI$metaregression_summary$s.nlevels[[which(SMD_S_B_EI$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 0)` | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_B_EI$metaregression_summary$s.names, SMD_S_B_EI$metaregression_summary$sigma2[[which(SMD_S_B_EI$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 'NA')` |
#### **Total exercise**
We provide a meta-regression where the total exercise as the product of the number of sessions, session duration, and session intensity, expressed in km, is considered as a continuous variable.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_B_TE <- metaregression_analysis1(df, "TvC", "BDNF", "TotalExercise1", 0.5)
```
**Figure 2.5.4.5 - Effect of exercise on BDNF by total exercise**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE,}
SMD_S_B_TE$regression_plot
```
The p-value for the association between total exercise and outcome reported was `r ifelse(SMD_S_B_TE$metaregression_summary$pval[2] < 0.001, "<0.001", sprintf("%.3f", SMD_S_B_TE$metaregression_summary$pval[2]))`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Strain | `r ifelse('Strain' %in% SMD_S_B_TE$metaregression_summary$s.names, SMD_S_B_TE$metaregression_summary$s.nlevels[[which(SMD_S_B_TE$metaregression_summary$s.names == 'Strain')]], 0)` | `r ifelse('Strain' %in% SMD_S_B_TE$metaregression_summary$s.names, SMD_S_B_TE$metaregression_summary$sigma2[[which(SMD_S_B_TE$metaregression_summary$s.names == 'Strain')]], NA)` |
| Study x Strain | `r ifelse('Strain/StudyId' %in% SMD_S_B_TE$metaregression_summary$s.names, SMD_S_B_TE$metaregression_summary$s.nlevels[[which(SMD_S_B_TE$metaregression_summary$s.names == 'Strain/StudyId')]], 0)` | `r ifelse('Strain/StudyId' %in% SMD_S_B_TE$metaregression_summary$s.names, SMD_S_B_TE$metaregression_summary$sigma2[[which(SMD_S_B_TE$metaregression_summary$s.names == 'Strain/StudyId')]], NA)` |
| Study x Strain x Experiment | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_B_TE$metaregression_summary$s.names, SMD_S_B_TE$metaregression_summary$s.nlevels[[which(SMD_S_B_TE$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 0)` | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_B_TE$metaregression_summary$s.names, SMD_S_B_TE$metaregression_summary$sigma2[[which(SMD_S_B_TE$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 'NA')` |
#### **SyRCLE RoB assessment considered as a categorical variable**
No studies met any RoB criteria.
#### **SyRCLE RoB assessment considering those studies where any item is at low risk of bias**
No studies met any RoB criteria.
#### **ARRIVE reporting completeness guidelines**
We provide a meta-regression where the number of ARRIVE items met is considered as a continuous variable.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_B_ARR2 <- metaregression_analysis(df, "TvC", "BDNF", "ARRIVEScore", 0.5)
```
**Figure 2.5.4.6 - Effect of exercise on BDNF by ARRIVE reporting completeness**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE,}
SMD_S_B_ARR2$regression_plot
```
The p-value for the association between ARRIVE reporting completeness and outcome reported was `r round(SMD_S_B_ARR2$metaregression_summary$pval[2],3)`.
| Level | Number of categories for that level included in this analysis | Attributable variance |
|:------------:|:---------------------------:|:---------------------------:|
| Strain | `r ifelse('Strain' %in% SMD_S_B_ARR2$metaregression_summary$s.names, SMD_S_B_ARR2$metaregression_summary$s.nlevels[[which(SMD_S_B_ARR2$metaregression_summary$s.names == 'Strain')]], 0)` | `r ifelse('Strain' %in% SMD_S_B_ARR2$metaregression_summary$s.names, SMD_S_B_ARR2$metaregression_summary$sigma2[[which(SMD_S_B_ARR2$metaregression_summary$s.names == 'Strain')]], NA)` |
| Study x Strain | `r ifelse('Strain/StudyId' %in% SMD_S_B_ARR2$metaregression_summary$s.names, SMD_S_B_ARR2$metaregression_summary$s.nlevels[[which(SMD_S_B_ARR2$metaregression_summary$s.names == 'Strain/StudyId')]], 0)` | `r ifelse('Strain/StudyId' %in% SMD_S_B_ARR2$metaregression_summary$s.names, SMD_S_B_ARR2$metaregression_summary$sigma2[[which(SMD_S_B_ARR2$metaregression_summary$s.names == 'Strain/StudyId')]], NA)` |
| Study x Strain x Experiment | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_B_ARR2$metaregression_summary$s.names, SMD_S_B_ARR2$metaregression_summary$s.nlevels[[which(SMD_S_B_ARR2$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 0)` | `r ifelse('Strain/StudyId/ExperimentID_I' %in% SMD_S_B_ARR2$metaregression_summary$s.names, SMD_S_B_ARR2$metaregression_summary$sigma2[[which(SMD_S_B_ARR2$metaregression_summary$s.names == 'Strain/StudyId/ExperimentID_I')]], 'NA')` |
#### **Heterogeneity explained by covariates (Effect of exercise on BDNF)**
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide', out.width = "100%"}
SMD_S_B_sexI <- subgroup_SMDI(df, "TvC", "BDNF", "Sex", 0.5)
SMD_S_B_VoFI <- subgroup_SMDI(df, "TvC", "BDNF", "VoF", 0.5)
SMD_S_B_ARRIVEI <- subgroup_SMDI(df, "TvC", "BDNF", "ARRIVEScoreCat", 0.5)
```
The table below shows which of the covariates, if any, explain some of the heterogeneity observed in the effect of exercise on BDNF. We present marginal R^2^, which measures the proportion of variance explained by including moderators in the model (the % change in the between-studies variance when the covariate is included in the model, in other words the % of the heterogeneity explained by the variable). The coefficients are derived form an RMA model fitted with an intercept (and so represent, for each category, the point estimate and 95% CIs of the effect in that category).
| Moderator | Category | $\beta$ | 95% CI | Marginal R^2^ (%) |
|:-----------:|:-----------:|:-----------:|:--------------------:|:-----------:|
| Overall effect | \- | `r SMD_S_B[["SMD_ML"]]$beta[1]` | `r SMD_S_B[["SMD_ML"]]$ci.lb` to `r SMD_S_B[["SMD_ML"]]$ci.ub` | \- |
| Sex | \- | \- | \- | `r round((r2_ml(SMD_S_B_sexI)[1]*100),1)`% |
| \- | *Female* | `r SMD_S_B_sexI$beta[1]` | `r SMD_S_B_sexI$ci.lb[1]` to `r SMD_S_B_sexI$ci.ub[1]` | \- |
| \- | *Male* | `r SMD_S_B_sexI$beta[2]` | `r SMD_S_B_sexI$ci.lb[2]` to `r SMD_S_B_sexI$ci.ub[2]` | \- |
| Voluntary or forced | \- | \- | \- | `r round((r2_ml(SMD_S_B_VoFI)[1]*100),1)`% |
| \- | *Forced* | `r SMD_S_B_VoFI$beta[1]` | `r SMD_S_B_VoFI$ci.lb[1]` to `r SMD_S_B_VoFI$ci.ub[1]` | \- |
| \- | *Voluntary* | `r SMD_S_B_VoFI$beta[2]` | `r SMD_S_B_VoFI$ci.lb[2]` to `r SMD_S_B_VoFI$ci.ub[2]` | \- |
| Duration of treatment | \- | \- | \- | `r round((r2_ml(SMD_S_B_DOE$metaregression)[1]*100),1)`% |
| \- | *per weeks of treatment increase* | `r SMD_S_B_DOE$metaregression_summary$beta[2]` | `r SMD_S_B_DOE$metaregression_summary$ci.lb[2]` to `r SMD_S_B_DOE$metaregression_summary$ci.ub[2]` | \- |
| Exercise intensity | \- | \- | \- | `r round((r2_ml(SMD_S_B_EI$metaregression)[1]*100),1)`% |
| \- | *per unit (m/min) increase* | `r SMD_S_B_EI$metaregression_summary$beta[2]` | `r SMD_S_B_EI$metaregression_summary$ci.lb[2]` to `r SMD_S_B_EI$metaregression_summary$ci.ub[2]` | \- |
| Total exercise | \- | \- | \- | `r round((r2_ml(SMD_S_B_TE$metaregression)[1]*100),1)`% |
| \- | *per km increase* | `r SMD_S_B_TE$metaregression_summary$beta[2]` | `r SMD_S_B_TE$metaregression_summary$ci.lb[2]` to `r SMD_S_B_TE$metaregression_summary$ci.ub[2]` | \- |
| Reporting completeness | \- | \- | \- | `r round((r2_ml(SMD_S_B_ARR2$metaregression)[1]*100),1)`% |
| \- | *per unit increase* | `r SMD_S_B_ARR2$metaregression_summary$beta[2]` | `r SMD_S_B_ARR2$metaregression_summary$ci.lb[2]` to `r SMD_S_B_ARR2$metaregression_summary$ci.ub[2]` | \- |
### 2.5.5 Sensitivity Analyses
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
options(scipen = 100, digits = 2)
```
We examine the robustness of the findings for BDNF by performing the following sensitivity analyses.
#### Imputed rho values of 0.2 and 0.8
In the previous analyses for the effect of exercise on BDNF, we imputed a $\rho$ value - the imputed within-study correlation between observed effect sizes - of 0.5. Here, we examine the effect of imputing $\rho$ values of 0.2 and 0.8.
```{r message=FALSE, warning=FALSE, eval = TRUE, echo = FALSE, results='hide'}
SMD_S_B_rho0.2 <- run_ML_SMD(df, "TvC", "BDNF", 0.2)
```