4. BREASTFEEDING OF BABIES
4.1 At the first interview of the Millennium Cohort Study (age 9-10 months) mothers were asked whether they had ever breastfed the cohort baby. Rates of ever breastfeeding varied considerably by UK country. The highest rate of ever breastfeeding was 72.2% in England, and the lowest was in Northern Ireland at 51.2%. In Scotland 64.7% of mothers said they had breastfed their baby for at least one day, a rate that is lower than the rest of the UK average (67.3%). Clearly the much higher rate of ever breastfeeding among England mothers has pulled up the average substantially. Scotland mothers' rate of breastfeeding for at least one month, 42.2%, was also lower than the UK average of 44.8%. The England sample of the Millennium Cohort Study contained sizeable numbers of mothers from minority ethnic groups, and minority ethnic mothers are known to have higher rates of breastfeeding than white mothers in England. This difference may be responsible in part or in whole for the much higher rates of breastfeeding in England and the higher average for the rest of the UK.
4.2 This analysis set out to examine whether the lower rate of breastfeeding among the MCS Scotland sample of mothers could be explained by the varying characteristics of MCS mothers in Scotland compared with the rest of the UK. The analysis examined whether a range of factors in combination, either known or suspected to influence mother's breastfeeding decisions, explained the difference between Scotland and the rest of the UK. Earlier studies suggested the following factors separately influenced mother's decisions to start breastfeeding their baby; whether the pregnancy was planned, attending an antenatal class, the mode of delivery, whether the mother was accompanied at the birth, whether the baby was admitted to special care and the duration of hospital stay, socio-economic status, ethnicity, mothers' educational qualifications, lone mother status, age at birth, age at first ever live birth, and parity (Bartington et al, 2006; Beresford, 2002; Griffiths et al, 2005; Hamlyn et al, 2002).
4.3 The factors that were analysed (in a logistic regression model) to examine whether they explained breastfeeding behaviour included: the mother's marital status, her highest educational qualification, whether she worked in pregnancy, mother's age at the birth of the cohort child, mother's ethnicity, the number of siblings to the cohort child and a selection of experiences around the birth. These variables were all collected at the first interview of MCS, and relied in most cases on mothers' recall about a period 9-10 months earlier. We also considered whether the sex of the child had any explanatory value. Other potential explanatory factors were analysed but dropped due to their lack of statistical significance, sometimes because they were highly correlated with existing variables. 10
4.4 The odds ratio of a mother in Scotland ever breastfeeding the cohort child was lower at 0.834 (p=0.000) compared with a mother in the rest of the UK ever breastfeeding.
4.5 After considering the range of explanatory factors in combination using the whole UK sample, the effect on ever breastfeeding of living in Scotland continued to be significant, and the odds ratio was reduced further (Table A4.1). A mother in Scotland is two-thirds as likely to breastfeed as a mother in the rest of the UK, after all of the other factors in combination were taken into consideration. So differences between UK countries in the MCS mothers' characteristics did not explain the lower rate of breastfeeding among Scotland's mothers.
4.6 The other predictors in the model were largely similar to those found in other studies of mother's breastfeeding behaviour. However, some of the predictors found to be significant in other studies were not significant in this analysis, (eg. the baby going to a special care unit after birth), largely through having correlations with other entered variables.
4.7 Factors that increased the chance of ever breastfeeding included:
- Being married over being cohabiting;
- Having higher educational qualifications compared with none;
- Being an older mother compared with the youngest mothers (14-21);
- Being the first child;
- The mother attending antenatal classes before the birth;
- Having a boy child.
4.8 Mothers from all of the minority ethnic groups were far more likely than white mothers to breastfeed their babies.
4.9 Factors that decreased the chance of ever breastfeeding included:
- Being a lone parent;
- Having a caesarean section delivery;
- Being alone at the birth.
4.10 In all of these cases mothers in Scotland had similar odds of breastfeeding as mothers in the rest of the UK with the same characteristics. However, due to the smaller sample size for Scotland, the odds ratios were not always significant in the Scotland model.
4.11 In conclusion, the lower rate of breastfeeding among mothers in Scotland compared to the rest of the UK is still a puzzle. It is not explained by the differences in characteristics between MCS mothers in Scotland and those in the rest of the UK, as far as we have been able to investigate. The lower rates of breastfeeding in Scotland are robust to the adjustment for a wide range of confounders. In fact, they were lower once these other factors are taken into account which suggests that there is a Scottish effect at work. This may have implications for policy and research. Attendance at antenatal class appears to have a stronger positive effect on mothers taking up breastfeeding in Scotland than in the rest of the UK. This suggests that policy should seek ways of encouraging attendance at antenatal classes in order to boost breastfeeding rates in Scotland.