Table 3. Amount of positive (HR) and negative (parasympathetic) significant reported effects.
| HR (35) | Parasympathetic (67) | ||||
|---|---|---|---|---|---|
| Median split | Range | # positive effects | Range | # negative effects | |
| Sample size | lower | 20–95 | 4 (18) | 17–135 | 4 (34) |
| higher | 100–1552 | 8 (17) | 159–9924 | 9 (33) | |
| Sex proportion | lower | 0% females | 7 (18) | 0–12% females | 1 (35) |
| higher | 8–100% femalesa | 4 (16) | 19–100% females b | 11 (30) | |
| Age | lower | 26.9–42.5 | 5 (17) | 29.1–41.6 | 5 (30) |
| higher | 42.8–51.2 | 7 (17) | 41.9–51.2c | 5 (28) |
The number of reported outcomes is in brackets. HR = heart rate. All reported outcomes were median split on sample size, sex or age.
a A median split was performed on the basis of the percentage of females as in a general median split it would be arbitrary which of the female outcomes would be included in the higher % sample.
b The median split was performed on the basis of the percentage of females. The sample was not exactly split in half because the median included a study with 6 reported outcomes. Therefore it would be arbitrary which of the reported outcomes would be included in the higher or lower percentage sample. We avoided this problem by including 30 outcomes in the higher % sample and 35 in the lower % sample.
c The median split was performed on the basis of age. The sample was not exactly split in half because the median included a study with 6 reported outcomes. Therefore it would be arbitrary which of the reported outcomes would be included in the higher or lower age sample. By including 28 outcomes in the higher age sample (and 30 in the lower) we avoided this problem.