Anales de la RANM

129 A N A L E S R A N M R E V I S T A F U N D A D A E N 1 8 7 9 THE IMPACT OF MIGRATION IN FIRST-EPISODE PSYCHOSIS López Ortega A et al. An RANM. 2024;141(02): 126 - 133 discord; (ix) expelled or suspended from school and/or high-school. Each item was dichotomised using the most conservative published cut-off points (21). A “total adversity” variable (range 0-9) was calculated with the dichotomised item scores. It was recoded on a categorical variable of 0 (none), 1 (one or two adverse experiences), 2 (multiple adverse experiences). For bullying, the participants were asked if they had experienced any type of bullying before the age of 17 (having been verbally abused or made fun of; having been ignored, excluded, or left out on purpose; having been hit, kicked, shoved, or locked in a room; or any other type of bullying). For the analyses, a categorical variable with three groups 0 (never), 1 (occasionally), 2 (frequently) was recoded. f. Statistical analyses Analyses were performed using SPSS v25.0. Data normality distribution was explored using Kolmogorov-Smirnov test. Continuous variables are described by mean, median, and standard deviation. Chi-squared tests x 2 frequency analyses were employed to describe the differences in categorical variables and T-tests were used to test differences in continuous variables for descriptive analyses. First, exploratory analyses were made to examine the differences in clinical presentation (i.e. severity of symptoms) between first generation migrants and those considered as “natives” (non-migrants and 2nd generation migrants) using T-test. Further analyses explored the differences between the 3 migration groups (non-migrants, 1 st genera- tion migrants and 2nd generation migrants), using analyses of variance ANOVA tests. Levene’s test was conducted to assess the homogeneity of variances, guiding the selection of the corres- pondent post-hoc tests. Post-hoc analyses were then used to identify differences between specific groups, with the Games-Howel test being used for non-homogeneous variances and the Scheffe test for homogeneous variances. A non-parame- tric Kruskal-Wallis test was used to analyse the differences between categorical variables in the three groups. Secondary analyses were conducted in the first- generation migration FEP group only, in order to further explore the plausible influence of other factors that had previously been associated with migration status (such as the presence of childhood adversities, sex, age and the age at migration) on clinical presentation in this group. A Multiple General Lineal Model (GLM) was conducted to explore the variables that might influence the presence of significant differences in symptom presentation at baseline among first- generation migrants. Predictive variables (i.e. sex, childhood adversities, age, age at migration) were selected according to the literature on the clinical presentation and migration status of first-episode psychosis (22–25). For all analyses, p values < 0.05 were used to indicate statistical significance. RESULTS a. Sample characteristics Out of the sample of 363 subjects, there were 114 who were 1 st generation migrants (34,2%) and 249 native-born patients (non-migrants and 2nd generation migrants) (65,95%). No significant differences were found in age, sex, and SES. Signifi- cant differences (p < 0,001) were found between “natives” and 1 st generation migrants in terms of ethnicity. The socio-demographic characteristics are presented in Table 1. b. Clinical presentation On the clinical scales, a significant difference (p < 0.01) was found for negative symptom presen- tation, assessed using the total of the PANSS Negative subscale, between migrants and natives; with 1 st generation migrants scoring 2 points higher on average than natives. No other significant differences were found. All the t-tests performed on the correspondent clinical scales are listed in Table 2. Further analyses weremade to explore the differences between the 3 groups (1 st Generation migrants, 2 nd Generation migrants and non-migrants). ANOVA analyses replicated a significant difference (p < 0,027) for the PANSS Negative Total between groups. Post-hoc analysis demonstrated the signifi- cant difference in the severity of negative symptom presentation was derived by higher average scores in 1 st generation migrants vs. non-migrants, with the first-generation migrant group having more severe (i.e. two points higher) negative symptoms than the later non-migrant group. c. Childhood adversities and bullying Kruskal-Wallis test revealed a significant difference in childhood adversity scores across the 3 groups (non-migrants, 1 st G migrants and 2 nd G migrants) ( x 2 = 8.767, p < .012 ). Post-hoc compari- sons demonstrated that the median of childhood adversity for the 1 st Generation migrants ( Md = 2) was significantly ( p < .011) higher than the median for the non-migrant group ( Md = 1). No significant difference was found for bullying scores between the 3 groups. d. Other plausible factors related to clinical presentation or migration The Multivariate Generalized Linear Model (GLM) including sex, childhood adversities, age and age at migration was not statistically significant ( R 2= 0.029, F(5,95) = 1.607, p = 0.166) for the PANSS negative in the first-generation migration group.