Visual habituation and dishabituation in preterm infants: A review and meta-analysis

Abstract

We review comparative studies of infant habituation and dishabituation performance focusing on preterm infants. Habituation refers to cognitive encoding, and dishabituation refers to discrimination and memory. If habituation and dishabituation constitute basic information-processing skills, and preterm infants suffer cognitive disadvantages, then preterms should show diminished habituation and dishabituation performance. Our review provides evidence that preterm infants’ habituation and dishabituation are impoverished relative to term infants. On the whole, effect sizes indicated that the differences between preterms and terms are of a medium magnitude. We also find that preterms’ performance is moderated by risk factors, stimulus materials, procedural variables, and age. These factors need to be taken into account in the construction of tests in which habituation–dishabituation tasks are employed. Overall, the habituation–dishabituation paradigm presents a promising approach in the diagnosis of cognitive status and development in preterm infants.

Introduction

The term infant is delivered about 38–42 weeks after the mother's last menstrual period. The World Health Organization (WHO) defines preterm birth as delivery before 37 completed weeks of gestation (and low birth weight if born under 2500 g). Rates of preterm birth vary around 10% and with country. Among the 4 million new births each year in the United States, approximately 12.3% of children are born too early—that is, approximately 1 in 8 babies (Martin et al., 2007). Among European countries, preterm birth rates vary widely, ranging from 5.3% in Lithuania to 11.4% in Austria. In Germany, the preterm birth rate amounts to 8.9% (EURO-PERISTAT Project, 2008, Macfarlane and Blondel, 2005). Overall, according to Beck et al. (2009), the highest rates of preterm birth are in North America and Africa (10.6% and 11.9%), and the lowest are in Europe (6.2%). Moreover, wherever trend data are available, rates of preterm birth are increasing. For example, the premature birth rate in the United States increased by more than 30% between 1981 and 2003 (Martin et al., 2007). Women pregnant through certain infertility treatments, poor women, and those under age 16 or over 35 have increased risk; even single babies conceived by in vitro fertilization are more likely to be preterm (Jackson, Gibson, Wu, & Croughan, 2004). Other factors that are associated with preterm birth include poor diet, maternal stress, lack of prenatal care and smoking, increased use of caesarean deliveries, and growth in multiple birth rates as well as ongoing technological advances in neonatal care that promote the viability of very small infants (Behrman and Butler, 2006, Davidoff et al., 2006, Goldenberg et al., 2008, Hamilton et al., 2004). Moutquin (2003) described the etiological heterogeneity of preterm birth leading to taxonomy of three main categories: medically indicated (iatrogenic) preterm birth, preterm premature rupture of membranes (PPROM), and spontaneous (idiopathic) preterm birth. The cause of spontaneous preterm birth tends to be unknown, and therefore it is difficult to predict and prevent (Behrman and Butler, 2006, Steer, 2005).

Preterm infants are commonly classified according to gestational age, the period of time between conception and birth (the number of weeks that the baby has been in utero) as well as birth weight. Postmenstrual age (PA) is gestational age plus chronological age, the time elapsed between birth and date of assessment. Age of preterms is usually described in terms of corrected age (Wilson & Cradock, 2004), denoting the age of the child from the expected date of delivery. Corrected age is determined by subtracting the number of weeks born before 40 weeks of gestation from the child's chronological age. Gestational age is normally trichotomized as “mild preterm birth” (32–36 weeks gestational age); “very preterm birth” (28–31 weeks gestational age); and “extremely preterm birth” (<28 weeks gestational age). Birth weight is trichotomized as “low birth weight (LBW)” 1500–2499 g; “very low birth weight” (VLBW) 1000–1499 g; “extremely low birth weight” (ELBW) ≤1000 g).

Preterm low birth-weight babies have average hospital stays of 45–50 days, and between one-third and one-half experience one or more rehospitalizations during the first 3 years of life (Behrman & Butler, 2006). Serious health problems and developmental delays are more pronounced among very preterm and very low birth weight babies, who account for between 14% and 15% of all preterm, low-birth-weight births in the United States. In Europe, very preterm births account for about 1% of all births. The Institute of Medicine (Behrman & Butler, 2006) estimated the annual societal economic burden associated with preterm birth in the United States to be $26.2 billion in 2005 (or $51,600 per infant born preterm). In Germany, the cost difference between a term and a preterm delivery amounts to about €10,550 (Kirschner, Halle, & Pogonke, 2009).

Very preterm, very low birth-weight infants are at increased risk for brain complications, such as intraventricular hemorrhage (IVH) and periventricular leukomalacia, both of which are associated with significant developmental delay (Allin, 2006). Intraventricular hemorrhage is defined by bleeding in areas surrounding the lateral cerebral ventricles (Luciana, 2003). IVH is classified into one of four grades with Grade 1 being the mildest degree of severity. IVH can injure the hippocampus, a site of recognition memory (e.g., Aylward, 2005, Kirwan et al., 2008). Beauchamp et al. (2008) found that very preterm infants with relatively small hippocampal volumes displayed working memory deficits at age 2 years. Reduced myelination has also been found in preterm, as compared to term, infants’ white matter (Mewes et al., 2006, Volpe, 2003, Woodward et al., 2004). Even preterms without early brain injury are characterized by a potential disruption in the development of brain structures such as the corpus callosum (Aylward, 2005). Birth before 28–30 weeks gestation can result in lung tissue which is very fragile, increasing the risk that this tissue will be injured. Injured lung tissue tends to trap air, collapse, or fill with mucus. Respiratory distress syndrome (RDS) is associated with infants born younger than 32 weeks of gestational age, with 80% of infants born younger than 27 weeks’ gestation developing RDS (Behrman and Butler, 2006, Verma, 1995). Sometimes bronchopulmonary dysplasia (BPD) or chronic lung disease (CLD) follow RDS in preterm infants (Vanhatalo, Ekblad, Kero, & Erkkola, 1994). Respiratory distress syndrome is indexed by the length of time on oxygen.

Bhutta and Anand (2001) linked cumulative brain injuries to observed cognitive deficits. Neuropsychological studies provide evidence that long-term cognitive and behavioral outcomes of preterm birth range from severe impairments, such as language disorders, to less severe problems, such as mild cognitive delays or visuomotor difficulties (e.g., Case-Smith et al., 1998, Feldman, 2009, Moster et al., 2008). Wood et al. (2000) found about 50% of their cohort of 30-month-old preterm infants from the United Kingdom and Ireland had a disability in mental or psychomotor development, neuromotor function, or sensory and communication function domains with approximately 25% reaching criteria for severe disability. In turn, these disabilities were predictive of child outcome at 6 years when the rate of moderate or severe disability observed was 46% (Marlow, Wolke, Bracewell, & Samara, 2005). Anderson, Doyle, and the Victorian Infant Collaborative Study Group (2003) reported that 55% of survivors born very preterm and extremely low birth weight in Australia exhibited clinically significant neurobehavioral impairment in middle childhood. Lefebvre, Glorieux, and St-Laurent-Gagnon (1996) reported a 30% overall incidence of abnormality in their cohort of Canadian preterm infants (born between 23 and 29 weeks gestation). Therefore, across countries findings have shown similar rates of atypical development, which have tended to occur in approximately 50% of extremely preterm infants.

Neurodevelopmental outcome effects appear to be much stronger in preterms who experience multiple risk factors including IVH, RDS, and a birth weight of less than 1500 g (e.g., Foulder-Hughes and Cooke, 2003, Wolke et al., 1994). Dividing all samples that are included in our review into preterms with IVH or RDS problems, preterms with other disturbances (excluding IVH or RDS problems), and preterms without additional risk factors, is confounded with both differences in birth weight and, to a lesser extent, gestational age. All preterm samples with no additional problems are characterized by low birth weight and by gestational ages between about 30 and 36 weeks (see Table 2). Most preterm samples with IVH or RDS have very low birth weight and a gestational age between about 30 and 33 weeks (see Table 4). The group of preterms with complications other than IVH or RDS (see Table 4) is, for the most part, comparable to the group of preterms with no additional complications in both birth weight and gestational age. Separation of these two groups is, thus, probably artificial. It is possible that preterms who are classified in the present review as non-risk de facto suffer similar complications as preterms with additional complications (except IVH or RDS problems), but that depictions of these complications were omitted in the research reports.

Premature birth is a major cause of developmental delay. In recent years, concern has shifted from the survival of preterm (and low birth weight) infants toward their long-term prognosis and quality of life. Despite improved survival rates, disability rates associated with preterm status have remained stable, leading to more survivors with disabilities and impairments as an absolute number (Anderson et al., 2003, Goldberg and DiVitto, 1983, Hintz et al., 2005, Lefebvre et al., 1996, Vohr et al., 2005, Zwicker and Harris, 2008).

Poor long-term outcomes have been documented in preterm infants in various domains of development, including motor, sensory, cognitive, and behavioral (for reviews see Anderson and Doyle, 2008, Bhutta et al., 2002, Salt and Redshaw, 2006, Zwicker and Harris, 2008). Vohr et al. (2000) found that ∼50% of a cohort of extremely preterm and extremely low birth weight infants in the United States had abnormal neurodevelopmental and sensory assessments. Similar incidence in cohorts of preterm and low birth weight infants have been found in various countries (Anderson et al., 2003, Khan et al., 2006, Lefebvre et al., 1996, Marlow et al., 2005, Wood et al., 2000).

In their meta-analysis, Bhutta et al. (2002; see also Bhutta, 2004) pointed out that preterm birth is associated with lower cognitive test scores at school age, a conclusion supported by Anderson and Doyle's (2008) review. Impairments in preterms’ (recognition) memory performance have been documented, not only in the first year of life (e.g., Rose, Feldman, & Jankowski, 2001), but in later childhood as well (e.g., Beauchamp et al., 2008, Isaacs et al., 2000, Luciana et al., 1999, Rose and Feldman, 1995). Furthermore, gestational age and birth weight appear to be directly proportional to cognitive performance: the younger the gestational age and lower the birth weight, the lower the cognitive score. Scores on the Mental Development Index (MDI) of the Bayley Scales of Infant Development have been shown to be significantly lower in preterm, as compared to term, children from 18 months to 30 months (Foster-Cohen et al., 2007, Hintz et al., 2005, Khan et al., 2006, Rose et al., 2005, Vohr et al., 2000, Wood et al., 2000). Similar results have been reported for the Griffiths Mental Development Scales (Lefebvre et al., 1996). Cognitive deficits, as documented by IQ scores, have been shown to persist into the early school years (6–7 years; Anderson et al., 2003, Marlow et al., 2005, Wiener et al., 1965), middle childhood (12 years; Constable et al., 2008), and adolescence and young adulthood (15 and 19.5 years; Allin et al., 2007, Allin et al., 2008).

Cognitive development is impaired not only in high-risk, but also in low-risk preterms, that is in preterms without neurological deficits such as cerebral palsy or mental retardation or hearing loss (e.g. Atkinson and Braddick, 2007, Caravale et al., 2005, de Haan et al., 2000, Luoma et al., 1998). In Caravale et al. (2005), low-risk preterms, at 3–4 years of age, obtained relatively lower scores in an intelligence test, a visual perception test, a location memory test, and a sustained attention test.

Research in preterm structure or function must consider age matching. Infants can be matched for either maturational age or experiential age (Matthews, Ellis, & Nelson, 1996). Maturational age comparisons use infants who were conceived at the same time and so are of the same postmenstrual age; experiential age comparisons use infants who are all tested the same amount of time after birth and thus have the same chronological or postnatal age. Age correction is controversial (Brandt and Sticker, 1991, DiPietro and Allen, 1991, Ross and Lawson, 1997) because of concerns about mistakenly overcorrecting or mis-estimating when conception actually occurred. Matching preterm and term infants on the basis of postmenstrual age controls for biological maturity; therefore, all infants are developmentally equivalent. This type of matching is based on the argument that development of preterm infants proceeds at the same rate as their term peers with a lag equal to the degree of prematurity. Preterm infants’ age is routinely adjusted in this way when estimating expected age of achievement of developmental milestones. Preterm infants are expected to arrive at milestones at an equivalent postmenstrual (but not postnatal) age to term infants. An advantage of this basis of matching is that biological maturity is controlled. Postmenstrual age matching has been used to examine differences between preterm and term infant brain development (Boardman et al., 2007, Woodward et al., 2004), neuromotor status (Gorga, Stern, Ross, & Nagler, 1988), information processing (Rose, Feldman, & Jankowski, 2002), language acquisition (Foster-Cohen et al., 2007, Sansavini et al., 2006), school behaviors (for example, aggression and shyness; Nadeau, Tessier, Boivin, Lefebvre, & Robaey, 2003), and temperament (Oberklaid et al., 1991, Sajaniemi et al., 1998). However, this type of matching has often masked developmental problems in preterm infants (Brachfeld, Goldberg, & Sloman, 1980).

To ensure equivalent biological maturity, preterm and term infants necessarily differ on postnatal experience. Matching preterm and term infants on the basis of postnatal age equates groups for postnatal experience. This type of matching has tended to be used in older children. For example, Marlow et al. (2005) investigated cognitive and motor impairments in 6-year-olds, and Carmody et al. (2006) investigated the impact of medical and environmental risk in infancy on 15–16-year-olds; both used this standard of matching. Although postnatal age matching allows comparisons of infants with equivalent postnatal experience, it fails to account for additional experience (for example, through antenatal classes) or preparations parents of term infants have taken due to longer pregnancies. Another limitation of postnatal age matching is that preterm infants are developmentally younger than their term peers at any given assessment; in consequence, results may represent effects of immaturity rather than prematurity (Brachfeld et al., 1980). Whether postnatal experiences have extra effects appears to depend on the domain being studied (Goldberg & DiVitto, 1983). Comparisons in which preterm infants have more postnatal experience than term infants provide preterm infants with some apparent advantages.

Brachfeld et al. (1980) suggested using both postnatal and postconceptional age mates as comparisons or controls. Piper, Byrne, Darrah, and Watt (1989), who followed this recommendation, compared motor development of a group of moderately preterm infants to very preterm infants in one group at 8 and 12 months postnatal age and another group at 8 and 12 months postterm age, rather than have a term control group. The use of both postnatal and postterm age allowed Piper et al. (1989) to demonstrate the differential impact of biological maturity on gross and fine motor development. Gross motor function was determined to develop based on biological age; therefore, neurologically intact infants developed at normal rates based on postterm age regardless of gestational age. However, fine motor development was not programmed solely by biological maturity.

Most research in prematurity favors some form of age correction to help determine whether the aspect of development in question is under maturational control or is susceptible to extrauterine experience. For example, Siegel (1983) conducted a longitudinal study in which preterm and term infants were repeatedly assessed over the first 5 years of life. Examining correlations between measures of infants’ corrected and uncorrected ages and later cognitive status, she found that age correction was appropriate in the early months but not later, suggesting that environmental influences grew in importance. Siegel's results are consistent with later recommendations for either full or half correction for prematurity during the first 2 years of life, but no correction thereafter (Blasco, 1989, Brandt and Sticker, 1991).