Pattern of birth in restrictive and bulimic eating disorders

Season of birth bias and anorexia nervosa: Results from an international collaboration

“…9,10 These findings are strengthened by a pooled meta-analyses of some of the UK samples 11 and a narrative review of the field. 12 In contrast, other studies do not support these findings, e.g., in Europe, [13][14][15][16][17][18] the USA, 19 Singapore, 20 or Australia. 21 There are two main explanations posed for any season of birth bias for AN: (i) altered neuropsychological functioning due to maternal infections during pregnancy, nutritional changes or sunlight exposure during gestation or the post partum period, [8][9][10]17,18 and (ii) altered parental fertility or reproductive patterns due to cultural influences, cyclothymia, or due to disordered eating patterns in the mothers which in combination with environmental temperature may influence fertility patterns.…”

“…12 In contrast, other studies do not support these findings, e.g., in Europe, [13][14][15][16][17][18] the USA, 19 Singapore, 20 or Australia. 21 There are two main explanations posed for any season of birth bias for AN: (i) altered neuropsychological functioning due to maternal infections during pregnancy, nutritional changes or sunlight exposure during gestation or the post partum period, [8][9][10]17,18 and (ii) altered parental fertility or reproductive patterns due to cultural influences, cyclothymia, or due to disordered eating patterns in the mothers which in combination with environmental temperature may influence fertility patterns. 9,10,13 There are three explanations for the contradictory findings of no season of birth bias: (i) low statistical power, 8 (ii) geographical differences in the bias, 20 and (iii) that there may indeed not be any bias.…”

“…9,10,13 There are three explanations for the contradictory findings of no season of birth bias: (i) low statistical power, 8 (ii) geographical differences in the bias, 20 and (iii) that there may indeed not be any bias. 18,21 Methodological limitations associated with the previously published studies includes: (i) absence of any a priori test-specific power calculations, thereby increasing the risk of both Type I and II errors, (ii) multiple subanalyses, which sometimes do not include a control group consisting of the general population. Such an apparent bias may in fact reflect the distribution of births in the general population, 22 (iii) a general lack of mention of effect sizes, contrary to the recommendations of The American Psychological Associations Production Manual 23 and a number of journals including International Journal of Eating Disorders (IJED) (see IJED Statistical Reporting Guidelines).…”

Season of birth bias in eating disorders—Fact or fiction?

“…4 The smallest sample used in analysis was 17 participants when comparing binge-purge AN to restrictive AN. 10 The other subgroups in the reviewed papers varied between around 100-300 participants. Future research needs to address the study's power, and ensure that the sizes of the (sub)groups are sufficient.…”

“…Six of the reviewed papers used this method. 7,[9][10][11][12]14 The main concern with such methods, as Chi-square that test goodness-of-fit between observed events and sine-waves for small samples, is the increased risk of Type I error.…”