« Previous | Contents | Next »
Listen
CHAPTER FOUR Resilience amongst disadvantaged mothers
Introduction
4.1 To explore resilience amongst mothers considered to be 'disadvantaged', two sets of analysis were undertaken - one was restricted to mothers aged under 25 at the time of birth, the other was restricted to lone parents 5. Those young mothers and lone parents who demonstrated 'positive' behaviours - breastfeeding, attendance at ante-natal classes, not smoking, - were compared, using the same set of socio-economic measures that have been applied above, with those young mothers and lone parents who demonstrated 'negative' behaviours.
Breastfeeding
4.2 Breastfeeding is associated with infant health outcomes, but remains very much patterned by socio-economic characteristics. Women in higher income households or with higher educational attainment are not only more likely to breastfeed in the first place, but to continue breastfeeding and delay weaning. Within the birth cohort, overall, 60% of mothers breastfed their baby but analysis for chapter three showed that, as well as age and family type, educational qualifications, household income, area deprivation and other socio-economic measures appeared to impact on the extent to which a mother had breastfed the cohort child. To identify the key independent factors which impacted on likelihood of breastfeeding amongst mothers aged under 25 and amongst lone mothers, logistic regression was undertaken. Table 15 displays the results of the under 25s regression.
Table 15 Logitistic regression model detailing factors related to breastfeeding amongst
mothers aged under 25 at the child's birth: birth cohort
Variable | Category | Significance | Odds ratio | 95% C.I. |
|---|
Lower | Upper |
|---|
Mother's education | (No qualifications) |
|---|
Standard grade or equivalent | 0.02 | 1.80 | 1.10 | 2.94 |
|---|
Higher grade or above | 0.00 | 2.72 | 1.69 | 4.38 |
|---|
Family type | (Lone parent living with other adults) |
|---|
Lone parent living only with child(ren) | 0.38 | 1.22 | 0.78 | 1.89 |
|---|
Couple family | 0.01 | 1.72 | 1.16 | 2.53 |
|---|
Tenure | (Rents from the local authority) |
|---|
Rents from a housing association | 0.48 | 1.15 | 0.78 | 1.68 |
|---|
Rents from a person or company | 0.00 | 2.37 | 1.58 | 3.56 |
|---|
Other rent arrangement or rent free | 0.70 | 1.13 | 0.60 | 2.14 |
|---|
Owns outright or buying with mortgage | 0.02 | 1.48 | 1.06 | 2.05 |
|---|
Nagelkerke's R 2 | 0.111 |
Hosmer & Lemeshow test | 0.953 |
4.3 Although significantly less likely to breastfeed than mothers aged 25 or older (66% of mothers aged 25 or older breastfed compared with 41% of those aged under 25), those younger mothers who did breastfeed were characterised by relative social and economic advantage: they were more likely than those who did not breastfeed to be in couple families (64% of breastfeeding young mothers were in couple families compared with 44% of non-breastfeeding younger mothers); to be owner occupiers and not social renters; and to have attained a higher level of education than non-breast-feeding younger mothers. Educational qualifications at Higher grade or above was the strongest predictor of breastfeeding within this group of younger mothers. The odds of younger mothers who had qualifications at Higher grade or above having breastfed the cohort child were almost three times as large as for mothers with no qualifications. The Hosmer and Lemeshow figure indicates good fit for this model, but the low value returned for Nagelkerke's R 2 suggeststhat measures of advantage and disadvantage only explain a small part of why younger mothers breastfeed or not.
4.4 Table 16 displays the results from the lone parents' regression model. Lone parents who breastfed were more likely than lone mothers who did not to be older, to be living in areas of lower deprivation and to be more highly educated. Being educated to Higher grade or above was the strongest predictor of breastfeeding amongst lone parents: being only slightly stronger than being aged 35 or older at the child's birth; the odds of lone mothers with Higher grades breastfeeding was 2.8 times higher than that for lone mothers with no qualifications. As with the younger mothers regression, the lone parents model parameters indicate that the particular measures of advantage and disadvantage considered here are perhaps not the most important factors influencing these behaviours.
Table 16 Logitistic regression model detailing factors related to breastfeeding
amongst lone mothers: birth cohort
Variable | Category | Significance | Odds ratio | 95% C.I. |
|---|
Lower | Upper |
|---|
Mother's education | (No qualifications) |
|---|
Standard grade or equivalent | 0.01 | 1.83 | 1.16 | 2.90 |
|---|
Higher grade or above | 0.00 | 2.80 | 1.82 | 4.31 |
|---|
Mother's age at child's birth | (Under 20) |
|---|
20 to 24 | 0.90 | 1.03 | 0.70 | 1.51 |
|---|
25 to 29 | 0.22 | 1.31 | 0.85 | 2.03 |
|---|
30 to 34 | 0.09 | 1.53 | 0.94 | 2.49 |
|---|
35 or older | 0.01 | 2.18 | 1.26 | 3.77 |
|---|
Area deprivation | Most deprived |
|---|
2 | 0.77 | 1.05 | 0.74 | 1.50 |
|---|
3 | 0.62 | 1.11 | 0.74 | 1.67 |
|---|
4 | 0.03 | 1.73 | 1.07 | 2.80 |
|---|
Least deprived | 0.00 | 3.18 | 1.53 | 6.63 |
|---|
Nagelkerke's R 2 | 0.078 |
Hosmer & Lemeshow test | 0.902 |
Attendance at ante-natal classes
4.5 Attendance at ante-natal classes is patterned by age, social position and particularly by parity: younger mothers, those in lower occupational grade households and those expecting a second or later child are much less likely to attend than older, more socially advantaged women and those expecting a first child. This pattern was reflected within the cohort of younger mothers as can be seen from the results of the regression model in Table 17. Again, those younger mothers who attended ante-natal classes were characterised by relative social and economic advantage. After controlling for parity, higher educational attainment, having a higher household income, and living in an area of lower deprivation were all positively and independently associated with ante-natal class attendance by mothers aged under 25. Whilst first-time motherhood was the strongest predictor of attendance, it is notable that having qualifications at Higher grade or above also had a strong association with attendance in this model. Although the explanatory power of this model, demonstrated by Nagelkerke's R 2,is betterthan for breastfeeding, it is still weak, and the model is also of poor fit.
Table 17 Logitistic regression model detailing factors related to attendance at
ante-natal classes amongst mothers aged under 25: birth cohort
Variable | Category | Significance | Odds ratio | 95% C.I. |
|---|
Lower | Upper |
|---|
Parity | (Multiparous) |
|---|
Primiparous | <0.01 | 7.68 | 5.29 | 11.14 |
|---|
Mother's education | (No qualifications) |
|---|
Standard grade or equivalent | 0.01 | 2.28 | 1.24 | 4.19 |
|---|
Higher grade or above | <0.01 | 3.09 | 1.71 | 5.56 |
|---|
Equivalised annual household income (quintiles) | (Less than £8410) |
|---|
Between £8411 and £13,750 | <0.01 | 1.66 | 1.17 | 2.36 |
|---|
Between £13,751 and £21,785 | 0.01 | 1.69 | 1.13 | 2.54 |
|---|
Between £21,786 and £33, 571 | <0.01 | 2.81 | 1.66 | 4.77 |
|---|
More than £33,572 | <0.01 | 3.43 | 1.56 | 7.53 |
|---|
Area deprivation | (Most deprived) |
|---|
2 | 0.05 | 1.45 | 1.01 | 2.09 |
|---|
3 | 0.03 | 1.58 | 1.06 | 2.35 |
|---|
4 | 0.01 | 1.80 | 1.14 | 2.86 |
|---|
Least deprived | 0.08 | 1.76 | 0.93 | 3.32 |
|---|
Nagelkerke's R 2 | 0.273 |
Hosmer & Lemeshow test | 0.010 |
4.6 Amongst lone parents, age was again a strong predictor of attendance - the odds of lone mothers aged over 25 having attended ante-natal classes were around twice as large as for younger lone mothers (Table 18). Working full-time was also a significant positive predictor of attendance for lone parents. The model parameters suggest that disadvantage is more associated with attendance at ante-classes for lone parents than for younger mothers, for whilst the predictive efficacy of the model is still weak, Hosmer and Lemeshow indicates good fit.
Table 18 Logitistic regression model detailing factors related to attendance at ante-natal classes amongst lone parents: birth cohort
Variable | Category | Significance | Odds ratio | 95% C.I. |
|---|
Lower | Upper |
|---|
Parity | (Multiparous) |
|---|
Primiparous | <0.01 | 6.98 | 4.66 | 10.46 |
|---|
Mother's age at child's birth | (Under 20) |
|---|
20 to 24 | 0.25 | 1.27 | 0.85 | 1.91 |
|---|
25 to 29 | 0.01 | 1.90 | 1.14 | 3.14 |
|---|
30 to 34 | 0.01 | 2.31 | 1.27 | 4.20 |
|---|
35 or older | 0.03 | 2.06 | 1.07 | 3.96 |
|---|
Maternal employment status | (Unemployed) |
|---|
Working part-time | 0.28 | 1.21 | 0.86 | 1.70 |
|---|
Working full-time | 0.01 | 2.19 | 1.22 | 3.95 |
|---|
Nagelkerke's R 2 | 0.200 |
Hosmer & Lemeshow test | 0.419 |
Smoking
4.7 Living in a couple household, being an owner occupier and having educational qualifications were all independently and positively associated with non-smoking amongst younger mothers (see Table 19). Maternal education was again the strongest predictor; the odds of those younger mothers having Higher grades or more not smoking were 3.5 times higher than those younger mothers who had no qualifications. The model parameters indicate this model is weak.
Table 19 Logitistic regression model detailing factors related to non-smoking amongst mothers aged under 25: birth cohort
Variable | Category | Significance | Odds ratio | 95% C.I. |
|---|
Lower | Upper |
|---|
Mother's education | (No qualifications) |
|---|
Standard grade or equivalent | <0.01 | 2.17 | 1.29 | 3.63 |
|---|
Higher grade or above | <0.01 | 3.61 | 2.17 | 6.00 |
|---|
Family type | (Lone parent living with other adults) |
|---|
Lone parent living only with child(ren) | 0.77 | 0.93 | 0.58 | 1.50 |
|---|
Couple family | 0.04 | 1.58 | 1.02 | 2.46 |
|---|
Area deprivation | (Most deprived) |
|---|
2 | 0.01 | 0.64 | 0.45 | 0.91 |
|---|
3 | 0.53 | 1.13 | 0.77 | 1.67 |
|---|
4 | 0.22 | 0.76 | 0.48 | 1.19 |
|---|
Least deprived | 0.38 | 1.35 | 0.69 | 2.65 |
|---|
Tenure | (Rents from the local authority) |
|---|
Rents from a housing association | <0.01 | 1.96 | 1.33 | 2.88 |
|---|
Rents from a person or company | 0.97 | 0.99 | 0.67 | 1.47 |
|---|
Other rent arrangement or rent free | 0.14 | 1.38 | 0.90 | 2.12 |
|---|
Owns outright or buying with mortgage | <0.01 | 1.96 | 1.33 | 2.88 |
|---|
Receipt of benefits | (Not solely reliant on benefits for income) |
|---|
Solely reliant on benefits | 0.01 | 0.67 | 0.48 | 0.92 |
|---|
Nagelkerke's R 2 | 0.173 |
Hosmer & Lemeshow test | 0.915 |
4.8 The pattern for lone parents was slightly different with occupational classification and employment status featuring more prominently alongside educational attainment as predictors of non-smoking. Age was not significant but lone mothers with higher grades or above, those who worked full-time and those in managerial/professional or intermediate occupations were all less likely to smoke. Similar values are returned by the model parameters indicating that this model is also weak.
Table 20 Logitistic regression model detailing factors related to non-smoking amongst lone mothers: birth cohort
Variable | Category | Significance | Odds ratio | 95% C.I. |
|---|
Lower | Upper |
|---|
Mother's education | (No qualifications) |
|---|
Standard grade or equivalent | 0.55 | 1.14 | 0.73 | 1.78 |
|---|
Higher grade or above | 0.01 | 1.80 | 1.17 | 2.75 |
|---|
Maternal employment status | (Unemployed) | | | | |
|---|
Working part-time | <0.01 | 1.99 | 1.44 | 2.76 |
|---|
Working full-time | <0.01 | 2.64 | 1.41 | 4.94 |
|---|
Mother's NS- SEC | Semi-routine and routine occupations |
|---|
Lower supervisory and technical | 0.39 | 0.81 | 0.50 | 1.31 |
|---|
Small employers and own account workers | 0.72 | 1.21 | 0.42 | 3.48 |
|---|
Intermediate occupations | <0.01 | 1.87 | 1.23 | 2.84 |
|---|
Managerial and professional | 0.05 | 1.63 | 1.00 | 2.66 |
|---|
Nagelkerke's R 2 | 0.118 |
Hosmer & Lemeshow test | 0.919 |
Summary
4.9 The overwhelming impression is that within the vulnerable or disadvantaged groups - younger mothers and lone mothers (although there is inevitable overlap on account of association between the two) - the key predictors of more health promoting maternal behaviours and service use are largely linked to relative social position. Thus, those younger mothers and lone mothers who breastfeed, don't smoke and attend ante-natal classes are those living in more socially advantaged circumstances and, relatedly, are more highly educated.
« Previous | Contents | Next »