Dietary data in the Avon Longitudinal Study of Parents and Children

Learn about the Avon Longitudinal Study of Parents and Children (ALSPAC) and its dietary measurements

Summary of cohort
Dietary data collection
Estimation of nutrient intake
Response
Key findings

Summary of cohort

The Avon Longitudinal Study of Parents and Children (ALSPAC), also known as Children of the 90s, is an ongoing birth cohort study of a sample of the population from Bristol and the surrounding area [88]. The main aim of ALSPAC is to understand how genetic and environmental factors influence the health and development of parents and children.

During initial recruitment all pregnant women who were resident in the former county of Avon, an area around Bristol in South West England, with an expected delivery date between 1^st April 1991 and 31^st December 1992 were invited to participate [88, 89]. ALSPAC initially enrolled a cohort of 14,541 pregnancies. When the oldest children were approximately 7 years old, additional eligible participants were invited to join the study. Therefore, the total sample for the child-based data collected at 7 years is 15,589 with 14,901 alive at 1 year of age. All of these children have been regularly followed up using parental and self-completion questionnaires, medical records, educational and clinical assessment and through linkage. A proportion of children born in the last 6 months of the recruitment phase (equivalent to 10% of the whole cohort) was selected to take part in a sub-study known as ‘Children in Focus’ (CiF). These children attended clinics between 4 months and 5 years of age (n=1432 ever attended).

In addition to studies of the children, ALSPAC has also followed up the mothers (‘Focus on the Mothers’), fathers (‘Focus on the Fathers’) and the children of the Children of the 90s.

Compared with the whole of Great Britain in 1991, the population of mothers with infants under one year of age resident in Avon were more likely to live in owner occupied accommodation, to have a car and less likely to have one or more persons per room and be non-white [90]. Similarly when comparing the ALSPAC participants to the whole eligible Avon population, less affluent people and ethnic minorities were less likely to be represented [90]. Ethical approval for the study was obtained from the ALSPAC Ethics and Law Committee and the Local Research Ethics Committees.

Dietary data collection

ALSPAC collected dietary information from both mothers, their partners and children at various time points (see the table below) using FFQs and diet diaries.

Diet-related questions in ALSPAC

	Mother’s FFQ	Partner’s FFQ	Child’s FFQ	Child’s diet diary‡
32 weeks gestation (1991/1992)	✔	✔ (partial)
4 weeks (1991/1992)			✔ (infant feeding)
4 months (1992/1993)				✔*
6 months (1991/1993)			✔ †
8 months (1992/1993)				✔*
15 months (1992/1994)			✔ †
18 months (1994)				✔*
2 years (1994/1995)			✔
3 years (1995/1996)			✔
3½ years (1996)				✔*
4 years (1996/1997)	✔	✔	✔
5 years (1997)				✔*
6/7 years (1997/99)			✔
7 years (1998/2000)				✔
8/9 years (2000/02)	✔	✔	✔
10 years (2002/2003)				✔
12/13 years (2004/2006)	✔	✔	✔ (part parent, part child)
13 years (2004/2006)				✔

Notes.
* CIF – 10% subsample.
† Infant FFQ not detailed enough to estimate energy intake.
‡ One day diet diary at 4 months. 3 day diet diary for other years.

The ALSPAC study website contains details of all the data that are available through a fully searchable data dictionary and variable search tool through the following webpage: http://www.bristol.ac.uk/alspac/researchers/our-data/. Since a number of FFQs were conducted we provide links to relevant questionnaire sources in the table below. Details of the dietary collection methods have been discussed in a previous publication and will be briefly outlined here [91].

Questionnaire sources for FFQs

Child’s Age	Questionnaire source	File	Link to questionnaire
32 weeks gestation	Carer questionnaire	C	Study source / CLOSER Discovery
4 years	Carer questionnaire	J	Study source / CLOSER Discovery
9 years	Carer questionnaire	N	Study source / CLOSER Discovery
12 years	Carer questionnaire	S	Study source / CLOSER Discovery

32 weeks gestation	Partner questionnaire	PB	Study source / CLOSER Discovery
4 years	Partner questionnaire	PG	Study source / CLOSER Discovery
9 years	Partner questionnaire	PL	Study source / CLOSER Discovery
12 years	Partner questionnaire	PQ	Study source / CLOSER Discovery

4 weeks	Child-based questionnaire	KA	Study source / CLOSER Discovery
6 months	Child-based questionnaire	KB	Study source / CLOSER Discovery
15 months	Child-based questionnaire	KC	Study source / CLOSER Discovery
2 years	Child-based questionnaire	KE	Study source / CLOSER Discovery
3 years	Child-based questionnaire	KG	Study source / CLOSER Discovery
4 years	Child-based questionnaire	KK	Study source / CLOSER Discovery
7 years	Child-based questionnaire	KQ	Study source / CLOSER Discovery
9 years	Child-based questionnaire	KT	Study source / CLOSER Discovery
12-13 years	Child-based questionnaire	TA	Study source / CLOSER Discovery

12-13 years	Child-completed questionnaires	CCM	Study source / CLOSER Discovery

Note. http://www.bristol.ac.uk/alspac/researchers/our-data/ contains full details of all the data that are available through a fully searchable data dictionary and variable search tool.

Mother’s and partners’ diets

At 32 weeks gestation, a self-competed FFQ with 43 food groups and 8 basic foods was used to assess diets of the pregnant women [92]. The food list was developed by nutritionists in Bristol in 1990. It aimed to cover all the main foods consumed in Britain based on those used in a study in South Wales [93] and modified according to a study which had recently collected weighed food intake data among adults in Avon [94]. A separate shorter version of the FFQ was sent to the women to pass onto her partner if she chose to do so.

Participants were asked to report the frequency of consumption of these foods with the following possible responses: (i) never or rarely (ii) once in two weeks (iii) 1-3 times a week (iv) 4-7 times a week (v), or more than once a day. Portion sizes were not reported. Further questions about more detailed aspects of the diet were asked such as: amount of fat on meat, type of bread, type of milk, and type of fat used for spreading and cooking. Participants were also asked if they were on a diet and if they were taking supplements.

Similar FFQs were used to assess the diet of mothers and their partners when the child was aged 4, 8 and 12 years. The food list was expanded to include 56 food groups and 12 drink groups based on experience gained when analysing the pregnancy FFQ and informed by foods and drinks recorded in the diet records collected on the 3.5 year olds. Questions about how many alcoholic drinks per day were also asked in more detail.

Children’s diets: FFQs

Parents completed information about the child’s diet at age 4 weeks, 6, 15 and 24 months as part of postal questionnaires. A range of questions were asked including: length of breastfeeding and/or formula, type of milk consumed, and age at which various solids were introduced. An infant FFQ was used at 6 and 15 months and 24 months. Responses to the infant FFQs included yes/no to if the baby has ever had the food along with the age started and the number of times per week. The questions cover ready-prepared baby foods, family foods and drinks. These questionnaires are not detailed enough to allow estimation of energy or nutrient intakes.

A full FFQ was completed by the main caregiver on behalf of the children when they were aged 3, 4, 7 and 9 years. This FFQ was adapted from the one used to assess maternal diet at 32 weeks gestation. There were some modifications over time; for example, from the age of 4 separate categories for coated poultry and coated fish were included [95]. From 7 years, separate questions were asked to establish what was eaten during school hours.

At 12/13 years, the child completed a 54-item FFQ which included items specific to school dinners, as well as other foods normally eaten foods outside the home. At the same time, parents completed an 80-item FFQ on behalf of the child covering foods provided by the parents including packed lunches but excluding school dinners and foods consumed outside the home. This FFQ was adapted from the maternal FFQ. These FFQ need to be used together to estimate energy and nutrient intakes and dietary patterns.

Children’s diets: Diet diaries

As part of the Children in Focus subsample, diet was assessed at 4, 8, 18 months and 3½ and 5 years using a three-day diet diary (one day diet diary at 4 months). These were not administered at the exact same ages as the FFQs. It was suggested that one weekend and two weekdays should be included in these diet diaries and they did not have to be consecutive days. The parents were asked to record all foods and drinks consumed by the child in household measures and bring the diary to the clinic visit where, if possible, any anomalies in the diary were clarified by a member of the nutrition team. From 3½ years onwards questions about vitamin supplements, types of spread normally used and types of bread and milk used were also asked in a separate short questionnaire accompanying the diary.

At 7 years, parents of the whole cohort were sent a 3-day diet diary to complete about their child. At 10 and 13 years, the diary was designed for the child to complete with the help of their parents.

Estimation of nutrient intake

Food Frequency Questionnaires

Estimated nutrient intake from the FFQs were calculated by multiplying the weekly frequency of consumption of each food item by the nutrient content (from McCance and Widdowson’s ‘The Composition of Food’s’ and it’s supplement reports) of a standard portion of that food item [24] and summing for all food/drinks in the questionnaire. Portion sizes were allocated according to the age of the participants. For children, there was no differentiation between boys and girls; however for adults the portions were larger for men than women [91]. Using information from the diet diaries collected when the children were 3½ years, Bristol-based nutritionists refined the proportion of individual foods to use when estimating the nutrient intake of each food group [91]. They also contacted schools to obtain copies of menus and serving sizes [91].

Diet diaries

At each age from 18 months, the diet diaries were coded using DIDO as described in the NSHD section of this guide [35]. Where appropriate, portion sizes were based on average portion sizes for children from previously published data [96-98] or based on manufacturers’ information or by adapting adult portion sizes [24].

Response

The table below provides the response to each dietary assessment as outlined in a previous publication [91]. The overall cohort consisted of, mothers and partners (n=14 541 pregnancies), children (n=14 062 live births, 13 988 alive at 1 year; clinic visits for whole cohort n=13 602 available at 7 years). The 10% CiF subsample included n=1432 attending at least once.

Responses to dietary measures in ALSPAC

	Mother’s FFQ	Partner’s FFQ	Child’s FFQ	Child’s diet diary
	Response (n(%))	Response (n(%))	Response (n(%))	Response (n(%))
32 weeks gestation (1991/1992)	12,423 (85)	9,960 (68)
4 weeks (1991/1992)			12,353 (88)
4 months (1992/1993)				964 (67)*
6 months (1991/1993)			11,490 (82)
8 months (1992/1993)				1,131 (79)*
15 months (1992/1994)			11,077 (79)
18 months (1994)				1026 (72)*
2 years (1994/1995)			10,432 (75)
3 years (1995/1996)			10,145 (73)
3½ years (1996)				863 (60)*
4 years (1996/1997)	9,504 (65)	5,102 (35)	9,722 (70)
5 years (1997)				772 (54)*
6/7 years (1997/99)			8,512 (61)
7 years (1998/2000)				7,309 (54)
8/9 years (2000/02)	7,661 (53)	3,638 (25)	7,965 (56)
10 years (2002/2003)				7,474 (55)
12/13 years (2004/2006)	6,819 (47)	3,340 (23)	6,781 (48) (parent)
12/13 years (2004/2006)	6,819 (47)	3,340 (23)	6,780 (48) (child)
13 years (2004/2006)				6,113 (45)

Notes.
* CIF – 10% subsample.

Key findings

A number of review papers using the ALSPAC dietary data have previously been published [99-102]. These describe over 100 papers published about diet in ALSPAC covering 4 research areas: pregnancy, infancy, childhood and dietary patterns.

Maternal diet during pregnancy

One of the strengths of ALSPAC is its ability to capture maternal diet during pregnancy. Results from papers analysing maternal diet from FFQs in ALSPAC found that median intakes were above the recommended nutrient intakes for the majority of nutrients but not for iron, magnesium, potassium or folate and that not many women took supplements [92]. Since folate is an important nutrient to prevent neural tube defects, this finding strengthens the argument for fortification of stable foods with folate.

Five dietary patterns (DP) were described in this sample of women using PCA: 1) ‘health conscious’ characterised by high consumption of salad, fruit, rice, pasta, breakfast cereals, fish, eggs pulses and non-white bread; 2) ‘traditional British’ characterised by high consumption of all vegetables, meat and poultry; 3) ‘processed’ high consumption of high-fat processed foods; 4) ‘confectionery’ high consumption of snack foods and high sugar content; and 5) ‘vegetarian’ high consumption of meat substitutes, pulses, nuts and herbal tea [103]. These DPs were socially patterned e.g. higher consumption of a health conscious DP was associated with higher education and older age [103].

The effects of maternal diet during pregnancy on their children’s health was assessed in a number of papers. For example, higher maternal fish consumption during pregnancy was associated with higher verbal IQ at 6 to 8 months as well as higher development scores between 15 and 18 months [104, 105]. These papers concluded that the benefits outweigh the risks of fish consumption during pregnancy (e.g. there was no indication of fish consumption being associated with high total mercury concentrations).

Partner’s diet

When the children were 4 years old, the mother’s partners’ (men only) DPs were similar to the DPs identified during pregnancy: 1) ‘health conscious’ 2) ‘traditional’ 3) ‘processed/confectionery’, and 4) ‘semi-vegetarian’” [106]. As with the women, there were strong associations between these DPs and sociodemographic variables [106].

Children’s diet

Diet in infancy has been described [99, 107]. DPs of the children at 3, 4, 7 and 9 years have also been described [108-110]. Three main DPs were identified at each of these childhood ages: 1) ‘processed/junk food,’ 2) ‘health conscious’, and 3) ‘traditional British’. These DPs were socially patterned and while similar patterns were identified at each childhood point, stronger periods of change were apparent between 3 and 4 and 7 and 9 years. DPs in adolescents were also described by combining FFQ information from both the parents and the child: 1) ‘traditional/health conscious’, 2) ‘processed’ 3) ‘snacks/sugary drink’, and 4) ‘vegetarian”’. There were clear sociodemographic differences. It was also noted that capturing dietary intake among adolescences is a difficult task and that using sources from both the parents and children themselves increased accuracy [111].

ALSPAC has collected diet longitudinally over the first 13 years of the study child’s life and has the ability to assess dietary change. The stability over time was assessed by obtaining dietary patterns using cluster analysis of 3 diet diaries kept for the same children at ages 7, 10, and 13 years [112]. A healthy cluster was identified and was the most stable pattern, with half of the children starting in that cluster remaining in it at all three ages. A processed cluster was the next most stable, with approximately 40% retained. This suggests that children introduced to either of these types of dietary pattern by 7 years of age are likely to continue with this pattern into adolescence.

A strength of the dietary data in ALSPAC children is the fact that they capture school lunches in the context of overall dietary intake which can support public health guidelines. While the nutrient composition of both school dinners and packed lunches were below dietary guidelines, children eating packed lunches had poorer nutrient intakes overall than those eating school dinners [113].

Learn about the other studies covered by this guide and their dietary measurements:

Get background detail on this guide:

Learn about harmonisation in the context of dietary data:

Discussion & illustration of dietary data harmonisation

Further information:

This page is part of the CLOSER resource: ‘A guide to the dietary data in eight CLOSER studies’.