Findings

Children of older first-time mothers
Children's physical health and development
Children's behaviour and cognitive development
Discussion

Children of older first-time mothers

Socio-demographic characteristics

Based on this sample from the NLSCY, about 11% of first-born children were born to older mothers (Table 1). A further 24% of children were born to middle mothers, 34% to reference mothers, 23% to young mothers, and 8% to teenaged mothers.

Children of older mothers had a similar socio-demographic profile as with children born to reference mothers (Table 1). Children born to teenaged and young mothers, however, were significantly more likely to be in single-mother and low-income households, and to have mothers with no more than a high school education compared to children of reference mothers.

Prenatal and birth-related characteristics

Children of older mothers tended to experience more prenatal and birth-related complications than children of reference mothers (Table 1). Children of older mothers were more likely to have a mother who suffered from gestational hypertension and to be born preterm (Figure 2). A greater proportion of children of older mothers were also delivery by caesarean.

These findings are supported by numerous studies of older mothers that have found increased risk of gestational hypertension (Carolan, 2003; Prysak et al., 1995; Tough, Tofflemire, Benzies, Fraser-Lee, & Newburn-Cook, 2007), and preterm birth (Carolan, 2003; Tough et al., 2006; Heck, Schoendorf, Ventura, & Kiely, 1997; Ziadeh & Yahaya, 2001; Tough et al., 2007) and caesarean delivery (Carolan, 2003; Heck et al., 1997; Ziadeh & Yahaya, 2001; Prysak et al., 1995; Tough et al., 2007).

NLSCY data indicated that about 7% of children of older mothers were exposed to maternal smoking throughout the pregnancy (Table 1). While this proportion was significantly lower than that among children of teenaged mothers (29%), it did not differ from that among children of reference mothers (10%).

Children of older mothers were significantly more likely to have been breastfed and breastfed longer compared to children of reference mothers. Forty-three percent (43%) of children of older mothers were breastfed for more than 6 months compared to 26% of children of reference mothers (Table 1). These rates were significantly lower among children of teenaged (15%) and young mothers (20%).

Children's physical health and development

Children who received special care at birth

When women were first interviewed in the NLSCY, they were asked about their pregnancy and birth experiences, including the health of their child at birth. One measure of the child's health was whether the child received any sort of specialized care at birth. Overall, 21% of children required some form of specialized care at birth (Table 2). Of these children, 39% spent time in the intensive care unit and 20% were put on ventilation or oxygen. On average, children who received specialized care were in this type of care for 7.3 days, although the median was much lower at 2.0 days.

Maternal age. More than one in four (26%) children of older mothers required special medical care when they were born, but this did not differ significantly from children of reference mothers (19%). Among children who required specialized care, there was no difference by maternal age group in the type of care that was received or the length of time spent in specialized care (Table 2).

In multivariate analyses controlling for socio-demographic and birth-related characteristics (Table 3), children of older mothers were as likely as children of reference mothers to have required specialized care at birth.

Other factors. The only variables to be significantly associated with receiving specialized care at birth in the multivariate model were gestational age and weight. Children who were born preterm, whether they were of low birth weight, were significantly more likely to have required specialized medical care at birth than children who were not born prematurely (Table 3).

Children's health in early childhood (ages 0 through 5)

At every cycle, mothers were asked about the health of their child. The majority of children in this study were reported by their mothers as being in excellent or very good health throughout early childhood: 93% at ages 0 to 1, 92% at ages 2 to 3 and 91% at ages 4 to 5 (Table 4). Because children are generally in good health, they do not exhibit many chronic conditions. However, asthma is a very prevalent chronic condition among children (Akinbami & Schoendorf, 2002; Asher et al., 2006; Mannino, Homa, Akinbami, Moorman, Gwynn, & Redd, 2002; Millar & Hill, 1998). Among the first-born children in this study, the prevalence of doctor-diagnosed asthma increased as children grew older, from 4% among 0 to 1-year-olds to nearly 16% four years later among 4- to 5-year-olds.

Excellent or very good general health

Maternal age. Overall, the general health of children born to older mothers did not differ from that of children of reference mothers (Table 4). Children of teenaged mothers were significantly less likely than children of reference mothers to be in excellent or very good health when they were ages 0 to 1. Conversely, children of middle mothers were significantly more likely to be in excellent or very good health when they were 4 to 5 years old, compared to children of reference mothers.

After controlling for the effects of other characteristics, maternal age was not significantly associated with children's general health at any age. The exception was that children born to middle mothers remained significantly more likely than children of reference mothers to be in excellent or very good health at ages 4 to 5 (Table 5).

Other factors. Few other characteristics showed significant associations with children's general health. Children born preterm and at low birth weight were significantly less likely than full-term children to be in excellent or very good health at ages 2 to 3, while children whose mother had no more than a high school education were significantly less likely to be in excellent or very good health at ages 4 to 5 (Table 5).

Asthma

Maternal age. There were no significant differences between maternal age groups in the proportion of children diagnosed with asthma (Table 4). After adjusting for the effect of socio-demographic factors as well as gestational age, birth weight and duration of breastfeeding (Table 6), maternal age continued to have no association with the diagnosis of asthma among children.

Other factors. From the multivariate models, boys were significantly more likely to be diagnosed with asthma than girls at ages 4 to 5, and children whose mothers had asthma were significantly more likely to receive an asthma diagnosis themselves (Table 6). These findings are consistent with results from other studies of childhood asthma (Garner & Kohen, 2008; Millar & Hill, 1998).

Developmental milestones

Mothers were asked to report the age in months at which their child first achieved several developmental milestones. On average, mothers reported that their children were 5.8 months old when they started eating solid food, 6.0 months old when they first sat up by themselves, 9.6 months old when they first fed themselves, 9.7 months old when they said their first word, and 11.4 months old when they took their first step.

Examining differences across maternal age groups, age at achieving these milestones tended to increase with mother's age at birth (data table not shown). Children of older mothers were significantly older than children of reference mothers when they first started eating solid food (6.3 vs. 5.5 months old), when they first fed themselves (10.3 vs. 9.4 months old) and when they said their first word (10.5 vs. 9.7 months old; Figure 3). These differences represent nearly a month-long lag in achieving these milestones.

Many textbooks or reference documents present the average age by which the majority of children will have achieved various developmental milestones (Bayley, 1969; Shaffer, Wood, & Willoughby, 2005), while others present the typical age range for the acquisition of these skills (WHO Multicentre Growth Reference Study Group, 2006). Based on responses for the present sample, cut-points were derived to identify children who were late achievers of certain developmental milestones (see the subsection Outcome variable definitions for cut-point definitions).

Examining these developmental milestones further, two milestones - eating solid food and feeding oneself - are highly dependent on the behaviour of the mother (parent) and may not be fully reflective of the physical development of the child. Therefore, further examination of the late development of milestones was limited to the remaining three measures: (i) late to sit up by self, (ii) late to say first word, and (iii) late to take first step.

Late to sit up by self

Maternal age. A significantly greater proportion of children of middle (17%) and older mothers (20%) were late achievers of sitting up compared to children of reference mothers (11%; Table 7). In multivariate models, children of middle and older mothers had significantly greater odds of being late (8 months or older) to first sit up by themselves (Table 8).

Other factors. Gestational age and birth weight were associated with a child being late to sit up by themselves. Children born prematurely but at a normal birth weight had significantly higher odds of being late to sit up by themselves compared to full-term children (Table 8).

Late to say first word

Maternal age. On average, 14% of children of older mothers were late to say their first word. This was not significantly different from children of reference mothers (Table 7). In the multivariate model, maternal age was not significantly associated with the likelihood of children's being late (13 months or older) to say their first word (Table 8).

Other factors. In the model, only children born prematurely but not of low birth weight were found to have significantly increased odds of being late to say their first word (Table 8).

Late to take first steps

Maternal age. On average, 19% of children of older mothers were late to take their first steps. This was not significantly different from children of reference mothers (Table 7). In the multivariate model, maternal age was not significantly associated with the likelihood of children being late (14 months or older) to take their first steps.

Other factors. Children whose mothers had no more than a high school education were significantly more likely to be late in taking their first steps relative to children whose mothers had higher education. Children born prematurely, whether they were of low birth weight or not, were also at increased odds of being late in achieving this milestone (Table 8).

Motor and social development (MSD)

Mothers responded to a series of age-specific questions regarding the motor and social skills possessed by their child, including gross and fine motor skills, perception and cognition, communication and language, and social development. These questions comprise the Motor and Social Development (MSD) scale. Although some of the child behaviours assessed by the MSD are dependent on the behaviour of the parent, this scale has been used in other studies of child development (Hediger, Overpeck, Ruan, & Troendle, 2002; Pevalin, Wade, & Brannigan, 2003). Overall, children had a mean standardized MSD score of 100.2 at ages 0 to 1, and 101.1 at ages 2 to 3 (Table 9).

Maternal age. Comparing scores by maternal age group (Table 9), children of middle and older mothers had significantly lower MSD scores at ages 0 to 1 compared to children of reference mothers. Conversely, children of teenaged and young mothers had significantly higher scores at ages 0 to 1 than reference mothers. These findings are consistent with other studies examining the motor and social development of children (Hediger et al., 2002; Pevalin et al., 2003).

After controlling for the effects of other factors, being a child of a middle or older mother was no longer associated with MSD scores at ages 0 to 1 (Table 10). At ages 2 to 3, controlling for other factors revealed an association with older maternal age that had been suppressed in univariate analyses, indicating that children of older mothers had significantly lower MSD scores at ages 2 to 3 compared to children of reference mothers.

Other factors. Other characteristics were also significantly associated with MSD scores in these models (Table 10). Boys had significantly lower MSD scores than girls at ages 0 to 1 and 2 to 3. Children from low-income families had significantly lower MSD scores at ages 2 to 3. Children who were breastfed for more than 6 months had significantly higher MSD scores at both time points compared to children who were never breastfed. Lastly, children who were born both preterm and at low birth weight had significantly lower MSD scores at both ages 0 to 1 and ages 2 to 3.

Children's behaviour and cognitive development

Behavioural outcomes

The relative importance of early childhood behaviour in predicting later outcomes has been the subject of numerous studies (Broidy et al., 2003; Koko & Pulkkinen, 2000; Nagin & Tremblay, 1999; Stevens & Pihl, 1987). Given these long-term implications, it was of interest to examine the relationship between advanced maternal age at birth and a child's behaviour during early childhood.

Along with maternal age, other factors that have been linked to child behaviour were examined including: mother's education (Lipman, Offord, Dooley, & Boyle, 2002), household income (Lipman et al., 2002), single-parent status (Lipman et al., 2002), parenting practices (Chao & Willms, 2002; Thomas, 2004; Pettit, Bates, & Dodge, 2007), family functioning (Racine & Boyle, 2002), maternal depression (Somers & Willms, 2002) and low birth weight (Japel, Normand, Tremblay, & Willms, 2002). Each of these factors was considered with respect to four types of behaviour: physical aggression, emotional disorder and anxiety, hyperactivity and inattention, and positive behaviour.

Physical aggression

Children had an average score of 4.3 out of 16 on the measure of physical aggression at ages 2 to 3, and an average score of 1.5 out of 12 at ages 4 to 5 (Table 11).

Maternal age. Children of older mothers did not differ in their average aggression scores from children of reference mothers at either time point. Children of teenaged mothers, however, scored significantly higher in physical aggression at ages 2 to 3 compared to children of reference mothers (Table 11). However, once all other factors were controlled for, maternal age at birth was not associated with this outcome at either time point (Table 12).

Other factors. Children whose mothers reported more ineffective parenting practices scored higher in physical aggression at both time points, which is consistent with other research (Chao & Willms, 2002; Thomas, 2004). Children whose mother reported more depressive symptoms scored higher in physical aggression at ages 2 to 3 but not later at ages 4 to 5. At ages 4 and 5, boys had higher scores than girls, and children whose mother had a high school diploma or less also had higher scores than those whose mother had more education. Children born preterm and at low birth weight, however, had significantly lower average scores in physical aggression (Table 12).

Emotional disorder and anxiety

Average scores on the emotional disorder/anxiety scale were 1.4 (out of 12) at ages 2 to 3, and 2.3 (out of 14) at ages 4 to 5 (Table 11).

Maternal age. At ages 2 to 3, there was no significant variation in average score by maternal age group (Table 11). At ages 4 to 5, while children of older mothers saw no difference in their average scores compared to the reference group, children born to middle mothers had significantly higher emotional disorder and anxiety scores compared to children of reference mothers (Table 11).

Once other factors were taken into account (Table 13), children of older mothers had similar emotional disorder and anxiety scores as children of reference mothers. However, children of teenaged and young mothers scored significantly lower in emotional disorder and anxiety at ages 2 to 3, but not at ages 4 to 5, compared to the reference group (Table 13).

Other factors. Higher emotional disorder scores were associated with more ineffective parenting practices and greater symptoms of maternal depression at both time points (Table 13). At ages 2 to 3 only, lower emotional disorder scores were observed for children with higher positive parenting scores as well as for children born both preterm and at low birth weight.

Hyperactivity and inattention

Average scores on the hyperactivity/inattention scale were 3.6 (out of 12) at ages 2 to 3, and 4.3 (out of 14) at ages 4 to 5 (Table 11).

Maternal age. Children of older mothers had similar hyperactivity and inattention scores as children of reference mothers, while children of teenaged mothers scored significantly higher in hyperactivity and inattention at ages 2 and 3 (Table 11). The relationship with maternal age disappeared, however, once other factors were taken into account (Table 14).

Other factors. In the multivariate models, parenting practices and maternal depression were significantly associated with hyperactivity and inattention at both time points (Table 14). Increases in ineffective parenting practices and mother's depressive symptoms scores were linked to increases in hyperactivity scores. Higher positive parenting practice scores, on the other hand, were associated with a decline in hyperactivity.

At ages 2 to 3, low-income children had higher average scores in hyperactivity compared to children with greater household income. At ages 4 to 5, boys scored significantly higher in hyperactivity compared to girls; this gender difference was not observed at ages 2 to 3 (Table 14).

Positive behaviour

Overall, children had an average score of 8.5 out of 12 on the positive behaviour scale, which assessed a child's perseverance and independence (Table 11).

Maternal age. Children of older mothers did not differ in their average scores from children of reference mothers (Table 11). In the multivariate model, the differences between children of older and reference mothers became statistically significant, with children of older mothers scoring significantly lower in positive behaviour compared to the reference group (Table 15).

Other factors. Boys had significantly lower positive behaviour scores than girls, as did children in single-parent households, and children whose mother had a high school diploma or less (Table 15). Ineffective parenting practices and higher family dysfunction were also linked to lower scores. Positive parenting practices, on the other hand, were significantly related to increases in children's positive behaviour scores.

Cognitive outcomes

Early cognitive abilities can have implications for social and academic success later in life (Feinstein & Duckworth, 2006; Stevenson & Newman, 1986). Three measures of cognitive ability at ages 4 to 5 were examined in this study: receptive vocabulary skills, number knowledge, and copying and symbol use.

Maternal age at birth, socio-demographic characteristics, parenting practices, family functioning, gestational age and birth weight, and whether the child was read to daily were each examined in relation to the child's cognitive abilities.

Receptive vocabulary

Children, on average, scored 103.3 in receptive vocabulary at ages 4 and 5 (Table 16).1

Maternal age. Children of older mothers did not differ significantly from the reference group, while children of teenaged and young mothers had significantly lower average receptive vocabulary scores (Table 16). Once other factors were taken into account, only children of young mothers persisted in having significantly lower scores in receptive vocabulary (Table 17).

Other factors. Lower scores were observed among children whose mother had a high school diploma or less, children with a single-parent mother, or children who were not read to daily. Higher scores in receptive vocabulary were observed among children whose mother reported more positive parenting practices (Table 17).

Number knowledge

The children's average score in number knowledge was 98.6 at ages 4 and 5 (Table 16).

Maternal age. Children of older mothers did not differ significantly from children born to reference mothers (Table 16). Children of teenaged and young mothers, on the other hand, performed significantly worse in number knowledge at ages 4 and 5 and this relationship persisted once other factors were taken into account (Table 18).

Other factors. As with receptive vocabulary, mother's education and daily reading were significantly linked to number knowledge. In addition, children born preterm and with low birth weight had significantly lower number knowledge scores compared to full term children (Table 18).

Copying and symbol use

Children's average copying and symbol use score at ages 4 and 5 was 100.5 (Table 16).

Maternal age. Children of older mothers did not differ significantly in copying and symbol use from children born to reference mothers (Table 16). Children of teenaged, young and middle mothers, however, all performed significantly worse than the reference group.

Once other factors were taken into account (Table 19), the significance of the relationship between maternal age at birth and copying and symbol use persisted for children of teenaged and young mothers. However, scores for children of middle mothers were no longer significantly different from those of children of reference mothers.

Other factors. Significantly lower copying and symbol use scores were observed among boys, children with less educated mothers, children born preterm (regardless of whether they were of low birth weight or not), and children who were not read to daily (Table 19).

Discussion

In Canada, it is increasingly common for women to delay childbearing. The proportion of Canadian women in their thirties and forties having a first child has increased in the past 20 years.

NLSCY data indicated that children of older mothers shared a similar socio-demographic profile with children of reference mothers. There were no significant differences between the proportions of children of older and reference mothers who had a mother with a high school diploma or less, lived in a low-income household or lived with a mother who was a single parent.

The children of older mothers were, however, more likely than children of reference mothers to be exposed to prenatal and perinatal risk factors. Higher proportions of children of older mothers had a mother who had suffered from hypertension during pregnancy, were born preterm, and had a caesarean delivery.

In summary, children of older mothers were born into similar socio-economic circumstances, but tended to experience more prenatal and perinatal risks compared to children of reference mothers. To effectively assess the link between late childbearing and children's outcomes, it was important to consider this information in the analysis. Consequently, the present study examined the relationship between late childbearing and children's developmental outcomes while taking these additional factors into account. In doing so, it was possible to isolate the relationship between maternal age at birth and the outcome, while also shedding some light on the relative importance of these additional factors.

Maternal age and children's outcomes

The results of this study showed that children of older mothers were no different from children of reference mothers with respect to many developmental outcomes. For some outcomes, this similarity was immediately apparent from a descriptive perspective. For others, initial differences disappeared once additional factors were taken into account.

For example, children of older mothers were as likely as children of reference mothers to have received special care at birth, to be in excellent or very good health during early childhood, or to be diagnosed with asthma. They shared similar timing with respect to saying their first word and taking their first step and had similar averages scores in physical aggression, emotional disorder and anxiety, and hyperactivity and inattention. Children of older mothers also had similar scores in receptive vocabulary, number knowledge, and copying and symbol use as children of reference mothers.

However, advanced maternal age was significantly associated with other outcomes. Even after controlling for a number of characteristics, a higher proportion of children of older mothers were considered as late achievers in sitting up by themselves compared to children of reference mothers. In addition, children of older mothers scored lower on the motor and social development scale at ages 0 to 1 and 2 to 3. They also had lower positive behaviour scores at ages 4 and 5.

The relative importance of perinatal health and socio-demographic characteristics

Several factors other than maternal age showed significant associations with the developmental outcomes examined in this study. These included mother's education, single-parent status, household income and the child's gestational age and birth weight.

For example, mother's education was particularly important in terms of children's cognitive development. Children of mothers with a high school diploma or less had significantly lower receptive vocabulary, number knowledge and copying and symbol use scores compared to children whose mothers had a higher level of education.

Furthermore, regardless of maternal age at birth, children who were born preterm fared less well than children born at full term. This was particularly true for outcomes that were tied to a child's chronological age. For example, children born preterm were significantly more likely to be late in achieving developmental milestones and scored lower on the MSD scale compared to children who were not born preterm. Given that the attainment of milestones and the development of motor and social skills are possibly affected by differences between a child's chronological and biological age, observing a delay among preterm children (who are biologically younger than their full-term counterparts) may not be surprising. However, the clinical significance of these lags is not clear from the present study, nor is it clear whether they will be important as the child ages.

Study limitations

As with any study, this study had some limitations. The first was the size of the sample. Although there were more than 3,000 children in this study's baseline sample, due to sample and survey design (see the section About the National Longitudinal Survey of Children and Youth), the sample diminished at older ages (e.g. ages 2 through 5). Consequently, certain differences that may have appeared meaningful did not reach statistical significance. As the NLSCY continues to recruit and follow younger cohorts of children, these analyses may be re-run with a larger sample of children.

Sample size limitations precluded examining the outcomes of children whose mothers were aged 40 and over at the child's birth. Studies of fertility and fecundity suggest that, compared to younger women, women aged 40 and older experience significant declines in their ability to conceive and to bring a child to term (Fretts et al., 1995; Gougeon, 2005; Ziadeh & Yahaya, 2001), while little information exists regarding the outcomes for their children. By grouping children born to women in their forties together with children born to women in their late thirties, the current study may have masked differences in their respective children's outcomes.

The second limitation was the age period of analysis. Due to the survey design, this study could only examine children's development up to age 5. It is possible that distinctions between children of older and reference mothers may not emerge until later in childhood. For example, the majority of children in this study were not yet in school. Other research has shown that intellectual differences in children born to younger versus older mothers become more pronounced as the child ages and develops (Brooks-Gunn & Furstenberg, 1986). Thus, differences in various academic and school-related outcomes may emerge after age 5.

A further limitation is that many of the measures used in the NLSCY are based on maternal reports, which may result in error or bias. While certain biases are known to exist, such as the tendency for parents to underreport their child's height (Shields, 2006), the potential bias in other measures is unknown. Furthermore, the biases may differ between maternal age groups. These biases may act to either conceal a real difference or, conversely, may establish a false association. However, for many measures, such as developmental milestones, the best and perhaps only option is maternal report.

Despite these limitations, the present study offers novel insight into the relationship between late childbearing and children's development in the Canadian context.

Conclusions and future research

This study found that, in general, older maternal age was not significantly associated with children's physical, behavioural or cognitive outcomes measured between the ages of 0 and 5. However, this study demonstrated that higher socio-economic status was positively associated with positive outcomes, while birth-related risk factors were linked with poorer outcomes. Given that both sets of characteristics are relatively more common among children of older mothers, could they be interacting in some fashion to obscure a direct relationship between maternal age and child outcomes? Future research could explore the interrelationships between maternal age, socio-economic status and birth-related risks, and investigate indirect and direct links between each of these factors and children's outcomes.

This study also focussed on children's outcomes between the ages of 0 and 5. As more cycles of data become available, future research could examine the relationship between maternal age at birth and outcomes measured later on in childhood and during adolescence.


Footnotes

  1. Although it is beyond the scope of this paper to compare the outcomes of first-born children to higher birth order children, it is worth noting that first-born children scored significantly higher (p<.001) in receptive vocabulary at ages 4 to 5 compared to the non first-born children whose average score was 100.9.