Maternal and child nutrition

Evidence statement 1.2

One systematic review (+) included studies on interventions and media campaigns conducted in developed countries to promote the uptake of folic acid supplements using advertising, leaflets and promotional material. It reported that these campaigns were effective in increasing the proportion of women of child‑bearing age that regularly take folic acid supplements.

Evidence statement 1.3

A large proportion of women of child‑bearing age who are planning a pregnancy or may become pregnant do not regularly take folic acid supplements. Evidence from 1 systematic review (+), which included 30 studies that reported risk factors for low pre‑conception folic acid use, found that low levels of formal education, young maternal age, lack of a partner, immigrant status and unplanned pregnancy are associated with lower odds of using folic acid around the time of conception.

Evidence statement 1.4

Evidence from a randomised controlled trial (RCT; +) based on a southern population in the USA who received brief counselling from a physician about the benefits of folic acid along with free folic acid supplement tablets, found that this was effective in increasing weekly folic acid supplement use.

Evidence statement 1.5

There is evidence from a large survey (+) of health professionals working in England that folic acid advice is not perceived by them as being part of general health advice for women of child‑bearing age. The survey also found that many health professionals working in England have gaps in their knowledge about the appropriate dosage and timing of folic acid for women.

Evidence statement 2.2

A non‑randomised trial in London (+) compared intervention groups that received multiple episodes of nutrition counselling alone or with 2 different types of food supplement during the second and third trimester with a control group. It found no significant differences among the groups when measuring maternal weight gain, length of gestation, babies head size or babies length. The study found a small but statistically significant increase in the mean birthweight of babies born to women in all intervention groups combined, compared to women in the control population.

Evidence statement 2.5

One randomised control trial (+) found a statistically significant maternal weight gain among pregnant women recruited into the USA's WIC programme at mid‑pregnancy, compared with women in a control population who received no free dietary supplements.

Evidence statement 2.6

One randomised trial (+) of pregnant women attending a welfare clinic in Finland, found that an intervention involving the provision of healthy foods and advice delivered 3 times during pregnancy improved some nutritional outcome measures. It did not lead to a significant difference in pregnancy outcomes between the intervention and control groups.

Evidence statement 3.1

There is evidence from 4 RCTs (all -) that diet and exercise programmes are effective in enabling some postpartum women to lose weight gained during pregnancy. This finding is based on US studies of women not noted to be from disadvantaged groups and who appear to be highly motivated to lose weight.

Evidence statement 3.2

There is evidence from 2 RCTs (both -) that a combination of diet and physical activity results in more effective and preferable weight loss than diet or physical activity alone.

Evidence statement 3.3

There is evidence from an RCT (-) that physical activity as part of a combined diet and physical activity intervention to promote weight loss, is more effective when frequent and regular, than when vigorous and less frequent.

Evidence statement 3.4

There is evidence from 2 RCTs (both -) that integrated programmes of activity, which support participants in combining diet and regular physical activity in order to promote weight loss in the postpartum period, are more effective than interventions which provide information alone.

Evidence statement 3.5

There is evidence from 2 RCTs (both -) that the characteristics of programmes which are effective in enabling some women to lose weight in the postpartum period are those which: combine diet and physical activity; include strategies for behaviour change; tailor the intervention to individual or group needs; include some group sessions and written materials; provide ongoing support and contact with programme staff; and are of a sufficient duration to make sustained lifestyle changes.

Evidence statement 3.6

There is evidence from 1 RCT (-) that short‑term weight loss of 1 kg a week achieved through a combination of diet plus physical activity in healthy postpartum women has no detrimental effect on milk quantity or quality and does not appear to affect infant weight gain.

A second RCT (-), combining diet and physical activity in healthy postpartum women (body mass index [BMI] 25 to 30) over a longer time period and resulting in a mean weight loss of 0.5 kg a week, did not appear to affect infant weight or length. However the study may not have been sufficiently powered to demonstrate such effects.

Evidence statement 4.1

Three (++) non‑randomised control trials evaluated peer support programmes. The interventions included training of peer supporters, antenatal and postnatal support (telephone, home visits group or contact at clinic that was initiated by the peer supporter). The studies found a statistically significant increase in the initiation and/or duration of breastfeeding among women from low-income groups who intended to breastfeed.

Evidence statement 4.2

Seven RCTs in a (++) systematic review evaluated peer support programmes. Six studies found that lay support resulted in a significant reduction in the cessation of exclusive breastfeeding, which appeared to be predominately during the first 3 months. However, 3 of the studies were in countries not considered relevant to NICE reviews and neither of the 2 contributing UK studies individually gave significant results. (These 2 UK studies were of populations containing a mixture of all social classes.)

Seven studies showed a similar but less significant reduction in the cessation of any breastfeeding, but subgroup analysis did not give a significant effect at any time point. Overall, the effect of incorporating an antenatal element of breastfeeding support into a study was not significant but those studies incorporating postnatal support alone significantly reduced the cessation of any breastfeeding up to 6 months.

Six studies using lay support contributed to the analysis and their results were compatible with the conclusion. Similarly, face‑to‑face support appeared to be more effective than telephone support in preventing the stopping of breastfeeding up to 6 months and all 7 studies which used lay support contributed to the analysis.

Evidence statement 4.4

Two (++) RCTs evaluated volunteer breastfeeding counsellors. The first found telephone support instigated by the supporter within 48 hours of hospital discharge significantly increased the duration of any and exclusive breastfeeding at 4, 8 and 12 weeks. This was compared with conventional care in relatively well‑educated mothers who were breastfeeding at study recruitment. The other study found that 1 antenatal visit, at which the offer of postnatal support was made along with a contact card and leaflets, had no effect on breastfeeding initiation or duration rates.

Evidence statement 4.7

Four RCTs (2 [++] and 2 [+]) evaluated health professional support. One (++) RCT included frequent postnatal visits and telephone support from a skilled, knowledgeable midwife and found breastfeeding duration rates increased significantly in women who had planned to breastfeed. One (+) RCT evaluated intrapartum visits in hospital and postnatal home visits with telephone support from a community nurse and peer counsellor. It found this to be effective in increasing the duration of exclusive breastfeeding among minority women on a low income. One (+) RCT evaluated structured support from a health professional (1 intrapartum and postnatal visit, and 1 phone call). It found no significant increases in breastfeeding rates at 6 weeks in women from the US armed forces. An Australian (++) RCT evaluated a series of structured postnatal home visits for teenage mothers starting at 1 week postnatal that included discussions on infant feeding by a midwife, in addition to routine hospital services. No increases in any breastfeeding rates were demonstrated.

Evidence statement 4.8

Eighteen RCTs in a (++) systematic review evaluated professional support programmes and found them to be effective overall. Twelve studies found that professional support gave a significant reduction in the cessation of 'exclusive' breastfeeding at all time points (except 4 months, for which it was marginally significant). The effect was greatest in the first 3 months.

The overall reduction in the cessation of 'any' breastfeeding found in the 16 relevant studies was not significant, but subgroup analysis found it was significant at 4 and 9 months and only 2 studies had an antenatal element. Face‑to‑face support appeared to be more effective than telephone support in preventing the stopping of 'any' breastfeeding up to 6 months. Four studies were set in low-income countries not considered to be relevant to NICE reviews.

Evidence statement 4.10

One (+) US‑based RCT evaluated the effect of a lactation consultant conducting 2 educational antenatal visits, weekly antenatal telephone contacts, a hospital intrapartum contact and postnatal home visits. It compared this with standard care for women on low incomes who were primarily Hispanic and black. The study found that the intervention significantly increased breastfeeding duration rates up to 20 weeks.

Evidence statement 4.11

One (+) RCT looked at antenatal breastfeeding education and postnatal lactation support for women who intended to breastfeed. Both these interventions (based in a hospital in Singapore) significantly improved rates of exclusive breastfeeding up to 6 months after delivery. Participants were chiefly Chinese or Malay. The postnatal support consisted of 2 one‑on‑one lactation consultant visits in hospital and was marginally more effective than the antenatal breastfeeding education. This consisted of a 16‑minute video which showed correct positioning, latching on, breast care and common problems and an opportunity to talk with a lactation consultant for 15 minutes. (Only postnatal support had a significant effect on breastfeeding rates and then only at 6 weeks after delivery.)

Evidence statement 4.12

Four RCTs (3 [+] and 1 [++]) included trained skilled, knowledgeable health professionals delivering breastfeeding interventions during pregnancy. Of these, 1 (+) found a group antenatal education specifically on positioning and attachment significantly increased exclusive breastfeeding rates at 6 weeks among low-income women who intended to breastfeed. One (+) included 2 to 4 individual antenatal sessions (lasting 10 to 15 minutes), training of health professionals and early frequent postnatal support that continued throughout the first year in a population of mostly white women on low income. It found a significant increase in the breastfeeding initiation and duration rates up to 2 months postpartum. One (++) included group antenatal education at 24 to 28 weeks; support in hospital; postnatal contact at 2 to 3 weeks and 3 months and found no difference in exclusive breastfeeding duration rates in women intending to breastfeed. One (+) included 5 to 8 home visits lasting up to an hour during the first 2 months with telephone support. Visits were concentrated in the first 2 weeks. The study found a significant increase in breastfeeding duration rates at 2 months postnatal.

Evidence statement 4.13

One (++) RCT evaluated a single, 30‑minute, one‑to‑one discussion and a leaflet on 'breastfeeding and employment' by a midwife or intern. The intervention did not significantly increase exclusive, or any, breastfeeding at 17 weeks postpartum. This study was conducted in France on a relatively affluent group of women.

Evidence statement 4.14

One (++) US‑based RCT evaluated a single discussion at registration for the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC; mean 12 minutes) and discharge packs at delivery. The study found breastfeeding duration was highest among mothers who had planned to breastfeed but had low breastfeeding knowledge.

Evidence statement 4.15

A US‑based (+/-) RCT compared an antenatal, breastfeeding session (50 to 80 minutes, led by the researchers) with a single one‑to‑one breastfeeding session (15 to 30 minutes) and standard care. It found significantly higher breastfeeding initiation rates in both intervention groups among US black women on low incomes.

Evidence statement 4.17

A (+) systematic review included studies in both developed and underdeveloped countries and older studies that had not been included in the other reviews. The main conclusions were: the most effective interventions in extending duration of breastfeeding combined information, guidance and support and were long term and intensive. During prenatal care, group education was the only effective strategy. During the postnatal period or both periods (antenatal and postnatal), home visits used to identify mother's concerns with breastfeeding, assist with problem solving and involve family members in breastfeeding support, were effective. Individual education sessions were also effective in these periods, as was a combination of 2 or 3 of these strategies in interventions involving both periods. Strategies with no effect had no face‑to‑face interaction, gave contradicting messages or were small‑scale interventions.

Evidence statement 4.18

A systematic review (+) included studies of breastfeeding support in developed countries all of which were included in more recent reviews. The main conclusions were that educational programmes were the most effective intervention and had the greatest effect on both initiation and short‑term duration of breastfeeding (up to 3 months). Support programmes conducted by telephone, in person, or both, increased short‑term and long‑term duration (up to 6 months). Written materials alone did not significantly increase breastfeeding. There was insufficient data to determine whether a combination of education with support was more effective than education alone.

Evidence statement 4.21

One (+) RCT examined a 1‑hour group, antenatal, breastfeeding session on positioning and attachment given by a lactation consultant. Most participants were from a low‑income group. The study demonstrated significantly higher rates of exclusive breastfeeding at 6 weeks compared to women who received standard antenatal care.

Evidence statement 4.22

One (+/-) Australian RCT evaluated a small, informal group antenatal, breastfeeding session in immigrant Vietnamese woman on low incomes. It found significantly higher breastfeeding initiation and duration rates among women who received the intervention as opposed to a leaflet alone.

Evidence statement 4.23

A Canadian based (+) RCT utilised a single 2.5‑hour antenatal breastfeeding workshop designed using Bandura's theory of self‑efficacy and adult learning principles at 34+ weeks gestation (with optional attendance by fathers). Using actual workshop attendance, the study found a significant increase in exclusive breastfeeding at 8 weeks postpartum, compared with standard care – but the result was not significant using intention to treat analysis. The study population was relatively well‑educated with a reasonable income.

Evidence statement 4.25

One (++) RCT evaluated the effect of an outpatient appointment 2 weeks after the birth with a physician/paediatrician (who had received 5 hours lactation training) on well‑educated women on high incomes. The study found significant increases in exclusive breastfeeding at 4 weeks and extended overall duration of breastfeeding.

Evidence statement 4.29

One (+) RCT, 1 (++) RCT, and 1 (+) before‑and‑after study suggest that post‑registration or update training for healthcare professionals to increase knowledge or skills in breastfeeding can be effective.

Evidence statement 4.30

Two (+) before‑and‑after studies evaluated a breastfeeding training programme for hospital health professionals and found a significant increase in breastfeeding duration rates.

Evidence statement 4.31

Two before‑and‑after studies evaluated the UNICEF Baby Friendly Hospital Initiative (BFI) training for health professionals in hospital settings. One study found significant increases in breastfeeding rates at 6 months, where initial breastfeeding rates were low. The BFI training did not increase breastfeeding rates at hospital discharge, where breastfeeding rates were relatively high. These conclusions are supported by a UK cross‑sectional study of BFI‑trained health visitors.

Evidence statement 4.32

One (+) RCT evaluated education and support, including: individual education that was given to all women in both groups (mostly white on low incomes); support in the ante‑, intra‑ and postpartum period and into the first year of infancy. This included training of health professionals, daily inpatient visits, a telephone call 48 hours after discharge, lactation clinic at 1 week and lactation consultant present at all health clinics up to 1 year after the birth. Significant increases were found in the initiation and duration of breastfeeding.

Evidence statement 4.33

One (++) before‑and‑after study conducted among American Indian women, evaluated the adoption of hospital policy and practices which were culture‑specific together with a media campaign. The latter included the 10 steps in the Baby Friendly Hospital Initiative, a peer‑support programme and a public health campaign. The study found a statistically significant increase in breastfeeding initiation rates.

Evidence statement 4.35

A (+) systematic review found the reconstitution of infant formula milk from powder may be associated with errors with a greater tendency to over‑concentrate feeds than under‑concentrate them.

Evidence statement 4.36

One (+) RCT compared a specific brand of mini‑electric breast pump with a specific brand of manual breast pump. No significant differences were found in the volume of milk expressed or its fat content.

Evidence statement 4.37

One (+) RCT compared pumping each breast sequentially with both breasts simultaneously. Women preferred simultaneous pumping which also produced a greater volume of milk. No significant differences were found in milk fat concentrations.

Evidence statement 5.1

There is evidence from 1 RCT (reported as moderate quality in a [+] systematic review) that intensive home visiting by a health professional significantly improved daily milk intake, self‑feeding, fruit or fruit juice and meat intake in children under 3 years, whose mothers were unmarried, low income, black schoolgirls (aged 15 to 18 years).

Evidence statement 5.2

One study (graded as 'moderate' in a [++] systematic review) found monthly visits by 'community mothers' significantly improved dietary intake of animal protein, non‑animal protein, wholefoods, milk, fruit and vegetables in infants under 1 year, from low‑income families in Dublin.

Evidence statement 5.6

There is no evidence from a systematic review (++) to support feeding with a hydrolysed formula for the prevention of allergies, compared to exclusive breastfeeding. In high‑risk infants who are unable to be completely breastfed, there is limited evidence that prolonged feeding with a hydrolysed formula compared to a cow's milk formula reduces infant and childhood allergy and infant cow's milk allergy. In view of methodological concerns and inconsistency of findings, further large, well designed trials comparing formulas containing partially hydrolysed whey, or extensively hydrolysed casein with cow's milk formulas are needed.

Evidence statement 5.12

A systematic review (+) based on poor quality studies found evidence that the duration of bottle use beyond age 12 months was not significantly associated with caries risk, though sweetened milk or juice given in a bottle increased the risk of early childhood caries before the age of 6 years.

Evidence statement 6.1

There is evidence from 2 RCTs (graded as 'moderate' in a [+] systematic review) and 2 other studies (graded as 'moderate' in a [++] systematic review) that nutrition education interventions that focus on skills development in the mothers of young children can be effective in improving the diets of the family. This is in terms of increasing the amount of fruit and vegetables consumed and in improving the quality and diversity of the diet.

Evidence statement 6.2

There is evidence from 2 RCTs (graded as 'moderate' in a [+] systematic review) and 2 other studies (graded as 'moderate' in a [++] systematic review) reported that effective nutrition education programmes aimed at the mothers of young children are those which: are multi‑faceted; include 'hands on' skills development; are tailored to the educational level and needs of the mothers and to family resources; include strategies for behaviour change; are intensive and ongoing; and are delivered by nutrition paraprofessionals and/or peer supporters.

Evidence statement 6.3

There is evidence from 2 studies among low‑income mothers, including teenage mothers (1 [+] RCT and 1 [+] before‑and‑after study) reported in 2 systematic reviews (1 [++] and 1 [+]), that interventions based on intensive and regular home visits by health professionals delivering tailored advice are effective in improving the diets of pre‑school children.

Evidence statement 6.4

There is evidence from 3 RCTs (all graded as 'moderate' in a [++] systematic review) reported that educational interventions which provide information through a variety of different media – such as storybooks, videos and audiotapes – can be effective in improving children's knowledge and understanding of healthy eating and their understanding of the relationship between nutrition and health. However, the provision of information alone does not appear to change eating behaviour.

Evidence statement 6.5

There is evidence from 2 RCTs (1 [+] and 1 graded as 'moderate' in a [++] systematic review) that the more frequently young children taste new or previously disliked foods, the more likely they are to accept those foods. One trial demonstrated that looking at the foods without tasting them was not effective in increasing the acceptance of those foods.

There is evidence from 1 systematic review (-) and 2 RCTs (1 [+] and 1 graded as 'moderate' in a [++] systematic review) that interventions which provide the opportunity for children to handle and repeatedly taste foods, are more likely to be successful in changing eating behaviour than interventions that provide information alone.

There is evidence from 1 non‑RCT (graded as 'sound' in a [+] systematic review) which compared 5 different actions intended to encourage pre‑school children to taste new fruits and vegetables. Offering a choice of whether to try the fruit or vegetable was more effective than offering a reward, which was effective only in the short term. An RCT (graded as 'moderate' in a [++] systematic review) also found that the use of rewards was not effective in bringing about dietary change.

Evidence statement 6.6

There is evidence from 1 RCT (graded as 'moderate' in a [++] systematic review) that foods should be positively presented in interventions which aim to encourage young children to eat healthily. A systematic review (+) concluded that: children consider taste, not health, to be the key influence on food choice; interventions should promote children's favourite fruit or vegetables, or target the ones they do not like; the emphasis on health messages should be reduced, particularly those concerning future health; and fruit and vegetables should not be promoted in the same intervention.

Evidence statement 6.8

There is evidence from 3 systematic reviews (1 [+], 1 [-] and 1 [++]) that classroom‑based interventions can be effective in increasing pre‑school children's nutrition knowledge and their consumption of particular foods.

Effective interventions appear to be those which are multi‑faceted and which include characteristics such as: teaching based on behavioural approaches; teaching levels which are developmentally appropriate; training for teachers in delivering the intervention; activity‑based teaching; opportunities to taste and handle foods; and reinforcement of learning from the classroom in the cafeteria and at home by parents.

Evidence statement 6.10

Evidence from a Brazilian study (++) reported in a systematic review (+) found children attending nurseries which restricted the consumption of sugar and who consumed lower amounts of sugar at lower frequencies, had a substantially lower risk of dental caries.

Evidence statement 6.11

A systematic review (+) based on 36 studies, found that the relationship between sugar consumption and caries is weaker in the modern age of fluoride exposure than it used to be, but controlling the consumption of sugar remains a justifiable part of caries prevention.

Evidence statement 6.13

A large US prospective study (+) of 642 children from birth living in a fluoridated water area, found an association between sugared drinks intake at age 1 to 4 years and dental caries at age 4 to 7 years – with the highest risk associated with sweetened drinks intake in the first year. Milk had a neutral association with caries. Total water intake at age 1 to 4 years was highly protective against dental caries at age 4 to 7 years. Total non‑water drinks consumption in the first year (including cow's milk) was the highest risk factor; while total water consumption was highly protective, suggesting that some of the adverse effect of sugary drinks may be because they reduce consumption of (fluoridated) water.

Evidence statement 7.1

Evidence from 7 studies (5 [+] RCTs and 2 [+] studies) show that antenatal vitamin D supplementation is effective in improving the vitamin D status of Asian and white women.

Evidence statement 7.2

No adverse effects were reported in any of the studies considering vitamin D supplementation to mothers or infants.

Evidence statement 7.3

Evidence from 1 (+) RCT indicates that infants of Asian mothers who received an antenatal vitamin D supplement achieved a higher body weight during the first year after birth than infants of mothers who received no antenatal vitamin D supplement.

Evidence statement 7.4

A (+) non‑RCT found that breastfed infants of supplemented (10 microgram/day) mothers had higher 25 hydroxyvitamin D levels 6 days after birth than breastfed infants of unsupplemented mothers. Vitamin D levels of all breastfed infants were lower than infants receiving infant formula.

Evidence statement 7.5

Evidence from a (+) RCT suggests that the infants of mothers given supplements during pregnancy (25 microgram/day during the third trimester) achieved a higher serum 25 hydroxyvitamin D levels than unsupplemented breastfed infants, at birth and at 4 days of age.

Evidence statement 8.3

A (+) cohort study found that anaemic children aged 5 years whose parents received individual counselling, group nutrition education and WIC food vouchers, achieved a higher mean haemoglobin level when compared with children whose parents who did not receive the intervention, at 6 months follow‑up.

Evidence statement 8.6

Evidence from 4 UK qualitative studies/surveys indicate that the introduction of solid foods is influenced by the mother's perceptions of the baby's needs, cultural beliefs and advice/encouragement from family members and friends. The most common reasons for early introduction of solid foods were the mother's perception that the infant was hungry and not settling (sleeping through the night). Infant weight was perceived as a marker of child health and successful parenting. There is an association between early introduction of solid foods and maternal smoking, non‑breastfeeding, male infants and low maternal educational level.

Evidence statement 8.7

The formation of children's food preferences and acceptance patterns are shaped by learning and repeated experience within the social context in which the food is consumed. Evidence from observational studies and surveys suggest that repeated exposure to a target food enhances the acceptance of same, similar and target foods in young infants.

Children's consumption of fruits and vegetables was positively associated with parental consumption of fruits and vegetables.

Women's own weight control attempt may influence their young daughters' emerging ideas, concepts and beliefs about dieting.

Evidence statement 8.8

A (+) UK‑based RCT suggests that a peer‑support intervention designed to improve infant feeding practices can increase feeding knowledge, confidence in following advice and was valued by recipients and volunteers providing the intervention. However, the intervention did not positively influence: vitamin C intake from fruits, growth parameters, use of NHS services and medication use among infants.

Evidence statement 8.9

Two UK‑based observational studies suggest that specially trained link workers can be effective in helping South Asian families to establish healthy weaning patterns and improve maternal knowledge, which may result in modest changes in children's diets, at least in the short term.

Evidence statement 8.10

Evidence from a UK‑based observational study suggests that a community‑based campaign to improve child feeding practices and oral health among the Asian children aged under 5 years was well received by the target populations. Long‑term outcomes were not reported.

Evidence statement 8.12

Evidence on the effectiveness of Sure Start was not identified. The St Philips Healthy Eating Project, which aimed to help families to develop healthy eating habits in a community setting, was well received and appreciated.

Evidence statement 8.13

There is evidence from 2 (+) non‑RCTs, 1 controlled before‑and‑after study and 1 observational study to suggest that interventions in day‑care centres improve the nutritional adequacy of the food provided and is associated with dietary improvements.

Expert reports

The expert reports were:

Cook P (2006) Handling and storage of expressed breast milk.

Goldberg G (2006) Nutrition and breastfeeding.

Sachs M, Dykes F (2006) Growth monitoring of infants and young children in the United Kingdom.

Cost‑effectiveness analysis

A model was developed, using data from the published literature, to determine the relationship between breastfeeding and later consequences. It also considered various interventions to increase breastfeeding rates, particularly peer support schemes. Sensitivity analysis was used to investigate how these different scenarios affect cost effectiveness.