A longitudinal study of antenatal and perinatal risk factors in early childhood cognition: Evidence from Growing Up in New Zealand

The effects of prenatal smoke exposure on language development ‐ a systematic review

“…As with previous reports, there was some inconsistency regarding the nature of the relationship between maternal smoking or exposure and language development; eight studies concluded that smoking directly impaired early language (Heinonen et al, 2011;Hernandez-Martinez et al, 2017;Hsieh et al, 2008;Julvez et al, 2007;Lee et al, 2019;Mohamed et al, 2018;Neumann et al, 2019;Polanska et al, 2017) while five concluded that the observed effects could be explained by confounding factors (Alati et al, 2008;Eriksen et al, 2012;Gilman et al, 2008;Huijbregts et al, 2006;MacArthur et al, 2001). These confounding factors varied between study designs making it hard to make a conclusion about the direction of effects.…”

“…Of the 14 studies included in this review, 13 (93%) reported negative associations between maternal pre‐pregnancy smoking, smoking during pregnancy or exposure to smoke and childhood language outcomes (Alati et al, 2008; Eriksen et al, 2012; Gilman et al, 2008; Heinonen et al, 2011; Hernandez‐Martinez et al, 2017; Hsieh et al, 2008; Huijbregts et al, 2006; Julvez et al, 2007; Lee et al, 2019; MacArthur et al, 2001; Mohamed et al, 2018; Neumann et al, 2019; Polanska et al, 2017) (Figure 2, Table 2). Differences in study design and reporting methods make it difficult to directly compare effects between studies but six of the fourteen studies reviewed (43%) report highly significant effects (Alati et al, 2008; Gilman et al, 2008; Hernandez‐Martinez et al, 2017; Hsieh et al, 2008; Huijbregts et al, 2006; MacArthur et al, 2001) ( p ≤ 0.001) and seven found marginal effects (0.001 < p < 0.05) (Eriksen et al, 2012; Heinonen et al, 2011; Julvez et al, 2007; Lee et al, 2019; Mohamed et al, 2018; Neumann et al, 2019; Polanska et al, 2017) (Figure 2, Table 2). Although most studies found consistent associations, their conclusions differed; five concluded that the effects of smoking on child language could be explained by indirect effects, primarily socioeconomic in nature (Alati et al, 2008; Eriksen et al, 2012; Gilman et al, 2008; Huijbregts et al, 2006; MacArthur et al, 2001) (Figure 2, Table 2).…”

“…See each individual paper for details of measures, models and effects. Julvez et al, 2007;Lee et al, 2019;Mohamed et al, 2018;Neumann et al, 2019;Polanska et al, 2017) which remained significant even after correcting for possible socioeconomic confounders (Figure 2, Table 2).…”

“…Studies that employed VIQ as a language outcome measure (Alati et al, 2008;Eriksen et al, 2012;Gilman et al, 2008;Julvez et al, 2007;MacArthur et al, 2001) generally considered an older age group (mean age 79 months, range 4 years to 11 years). In addition, these were more likely to report p-values ≤0.001 (Alati et al, 2008;Gilman et al, 2008;MacArthur et al, 2001) and indirect effects (Alati et al, 2008;Eriksen et al, 2012;Gilman et al, 2008;MacArthur et al, 2001) than studies of developmental language indices (Hernandez-Martinez et al, 2017;Hsieh et al, 2008;Lee et al, 2019;Mohamed et al, 2018;Moore et al, 2020;Polanska et al, 2017) or vocabulary (Heinonen et al, 2011;Hernandez-Martinez et al, 2017;Huijbregts et al, 2006;Moore et al, 2020;Neumann et al, 2019); these generally involved testing at younger ages (mean age 34 months, range 2 to 5 years)…”

“…and were more likely to report marginal p-values (0.001 ≥ p ≤ 0.05) (Heinonen et al, 2011;Hsieh et al, 2008;Lee et al, 2019;Mohamed et al, 2018;Neumann et al, 2019;Polanska et al, 2017) and direct effects (Heinonen et al, 2011;Hernandez-Martinez et al, 2017;Hsieh et al, 2008;Lee et al, 2019;Mohamed et al, 2018;Neumann et al, 2019;Polanska et al, 2017).…”