Ji et al. (2021).

Population:	Large White x DM (Duroc x Min Pig [YDM]), (China).
Sample size:	Before weaning: 12 litters (133 piglets, 11.2 ± 1.5 piglets per pen). After weaning: 12 litters (120 piglets), selected 10 piglets per litter (four litters in each treatment group), selected with similar body weight, sex balanced.
Intervention details:	Pre-weaning treatment groups: Piglets were randomly allocated into a treatment group. Group 1: control group, no contact or socialisation (CON), n = 4 litters (43 piglets). Housing design: littermate-only with sow in a farrowing crate until weaning. Group 2: intermittent contact / socialisation group (IM), n = 4 litters (46 piglets). Socialisation housing design: littermate-only with sow in a farrowing crate on days 1–14, then two neighbouring litters socialised in shared activity area on days 14–35 for 3 hours per day. Socialisation group size: two litters per pen (socialisation group size is the number of litters that interact with one another during lactation). Group 3: continuous contact / socialisation group (CM), n = 4 litters (44 piglets). Socialisation housing design: littermate-only with sow in a farrowing crate on days 1–14, then two neighbouring litters socialised in shared activity area on days 14–35 with 24 hours access. Socialisation group size: two litters per pen.
Study design:	Randomised control trial.
Outcome studied:	Weaning age: 35 days. Socialisation length: 21 days. Regroup condition: CON: Randomly mixed according to weight and sex, unfamiliar piglets. IM: Mixed with non-adjacent litters in same treatment group, unfamiliar piglets. CM: Mixed with non-adjacent litters in same treatment group, unfamiliar piglets. Post-weaning group size: 10 piglets per group / pen.  Physiological indicators: Blood, day 42 (1 week after mixing). Interleukin-1β (IL1β). Cortisol (COR). Interleukin-6 (IL6). Interleukin-10 (IL10). Brain derived neurotrophic factor (BDNF). Growth performance: Day 63. Group uniformity: coefficient of variation (CV). Feed to meat ratio. Average daily weight gain (ADWG). Behavioural indicators: Days 35–36, and days 35–38, 30-second scan sampling every 10 minutes. Aggressive behaviours: fighting, head knocks. Social interactions: oral-nasal contact, mounting, no reaction, avoiding, return approach.
Main Findings (relevant to PICO question)	Physiological indicators: IL1β: CM > CON > IM (significant, P < 0.001). COR: IM significantly lower than CON and CM (P = 0.005). IM < CON < CM (trend only). BDNF: IM > CON (significant, P = 0.024). Growth performance: From weaning until the end of nursery (day 63). CV: IM significantly lower CV than CM and CON (P = 0.043). Feed to meat ratio: CM significantly higher than IM and CON (P = 0.001). Average daily weight gain (ADWG): IM > CM (significant, P = 0.005), but no significant difference between CON and IM / CM. Behavioural indicators: Aggressive behaviours: Fighting: CON > IM > CM (significant on days 35–36, P = 0.001; days 35– 38, P < 0.001). Head knocks: control group significantly more than socialised group (days 35–36, P = 0.003; days 35–38, P = 0.001). Social interactions: Oral-nasal contact: CON significantly more than socialised group (days 35–36, P < 0.001), CON > IM > CM (significant on days 35–38, P < 0.001). Mounting: no significant difference (days 35–36), IM significantly lower than the other groups (days 35–38, P = 0.036). No reaction: CM significantly lower than the other groups (days 35–36, P = 0.018; days 35–38, P < 0.001). Avoiding: CM significantly higher than the other groups (days 35–38, P = 0.008). Return approach: CM significantly higher than the other groups (days 35–36, P = 0.001; days 35–38, P < 0.001).
Limitations:	The mechanism of randomisation for treatment group allocation was not reported. The study was not blinded. The socialisation group (IM and CM) had more space allowance than the control group (CON) during lactation, which might be significant.  The proportion of familiar piglet to unfamiliar piglet (familiarity ratio) in a pen after regrouping was not specified. There was no baseline for the physiological indicators established prior to regrouping. Behavioural observations were recorded by one observer only.