The influence of socioeconomic factors and family context on energy ...

European Journal of Clinical Nutrition (2015) 69, 47–54 © 2015 Macmillan Publishers Limited All rights reserved 0954-3007/15 www.nature.com/ejcn

ORIGINAL ARTICLE

The influence of socioeconomic factors and family context on energy-dense food consumption among 2-year-old children S Vilela1, A Oliveira1,2,3, E Pinto1,5, P Moreira1,4, H Barros1,2,3 and C Lopes1,2,3 BACKGROUND/OBJECTIVES: Adverse effect on health has been described for a high consumption of energy-dense food, among children and adults. Limited research has been performed among pre-school children. The objective of this study is to evaluate the association between socioeconomic characteristics and family structure, and the consumption of energy-dense food among 2-yearold children. SUBJECTS/METHODS: The study sample includes 808 2-year-old children from the Portuguese birth cohort Generation XXI with information on food consumption. Data were obtained from questionnaires administered by interviewers to parents. Based on a food frequency questionnaire, four groups of energy-dense food were defined: soft drinks (sweetened drinks), sweets (chocolate and candies), cakes (creamy and not creamy cakes and sweet pastry) and salty snacks (crisps, pizza and burger). Multinomial logistic regression models (odds ratio and 95% confidence intervals) were fitted to estimate the associations. RESULTS: Intakes of energy-dense food were much lower than in similar aged children in other Westernized countries. Maternal age and education, grandparents’ education, household income and maternal occupation were inversely associated with the consumption of energy-dense food, particularly soft drinks and sweets. Children with older siblings were more likely to have a daily consumption of any energy-dense food. Few significant associations were found between socioeconomic characteristics and family structure and consumption of cakes and sweets less than once a week. CONCLUSION: High socioeconomic characteristics were associated with lower consumption of energy-dense food by 2-year-old children, mainly soft drinks and sweets. This influence is not only from parents’ background but also from the preceding generations. European Journal of Clinical Nutrition (2015) 69, 47–54; doi:10.1038/ejcn.2014.140; published online 23 July 2014

INTRODUCTION Adverse effects of energy-dense food on health, in both children and adults, were previously described,1–3 in particular the link between consumption of sugar sweetened beverages (SSB) and overweight and obesity in children and adults, the association between fast-food intake and weight gain and obesity, or the higher risk of developing type 2 diabetes and metabolic syndrome among those in the upper categories of SSB intake. Specifically, greater intake of salty snacks, SSB and increased portion size of snacks have been observed as potential contributors to the daily energy intake,4 and consequently, may have an important role in childhood obesity.5 Moreover, snacking habits (sugar-containing snacks and chips/crisps) have also been associated with caries in pre-school children living in a low-socioeconomic status.6 A person’s socioeconomic position has a well-recognised relationship with health inequality that may influence their disease burden in the life course.7,8 It is possible that such influence is the result of dietary practices that vary according to socioeconomic position.9,10 Moreover a child’s diet is likely to reflect the socioeconomic position of the parents. Parental low income, unfavourable working conditions and low levels of 1

maternal education, have been associated with higher consumption of energy-dense food,11–15 while children of non-smoking and more educated mothers tend to have a healthier diet.16 Also, the family structure, the primary caregiver or the size of the family influence children’s diet: having older siblings or families with more than two children resulted in a higher intake of energydense food.13–15 Most studies dealing with the influence of socioeconomic position and food habits have focused on school-aged children and limited research has been conducted among younger children. However, it is likely that determinants of eating habits vary across age groups.17 In early life, parents have a more central role in shaping the family food environment and the eating experiences,18 while older children have greater autonomy over their food intake and progressively start making their own choices.19 Thus, different determinants of energy-dense food consumption at different ages probably operate throughout childhood, making their identification at each age essential to design appropriate interventions, directed to the early induction of meaningful changes with a long lasting effect on health. This study aimed to identify how socioeconomic characteristics and family structure impact on the consumption of energy-dense food among 2-year-old children.

Institute of Public Health, University of Porto, Porto, Portugal; 2Department of Clinical Epidemiology, Predictive Medicine and Public Health, University of Porto Medical School, Porto, Portugal; 3Cardiovascular Research & Development Unit, University of Porto Medical School, Porto, Portugal and 4Faculty of Food and Nutrition Sciences, University of Porto, Porto, Portugal. Correspondence: Professor C Lopes, Department of Clinical Epidemiology, Predictive Medicine and Public Health, University of Porto Medical School, 4200-319 Porto, Portugal. E-mail: [email protected] 5 Current address: CBQF—Centro de Biotecnologia e Química Fina—Laboratório Associado, Escola Superior de Biotecnologia, Universidade Católica Portuguesa/Porto, Porto, Portugal. Received 6 December 2013; revised 3 June 2014; accepted 10 June 2014; published online 23 July 2014

Influences on energy-dense food consumption in children S Vilela et al

48

METHODS Study design The present study is based on Generation XXI—a Portuguese populationbased birth cohort, previously described.20–22 In brief, 8647 children were enroled during 2005/2006 at the five public hospital maternities covering the metropolitan area of Porto, Portugal. Of the invited mothers 91.4% accepted to participate. At 2-years-of-age a subsample of 855 children was re-evaluated. The follow-up took place in two phases: children completing their second anniversary between April and August 2007 were invited to participate in the first follow-up; later, all children born in January 2006 were invited to a similar evaluation during January 2008. In the present study, these two groups were combined and 808 singleton children with complete information on food consumption were included.

Data collection At birth and at 2-years-of-age, data were collected by trained interviewers using structured questionnaires. Determinants and co-variables. At baseline, maternal data were collected within 72 h after delivery, during the hospital stay, through face-to-face interviews. The following characteristics were obtained at baseline and used for the present analysis: mothers’ birth date, maternal education level (categorized into ⩽ 6, 7–9, 10–12, ⩾ 13 schooling years), maternal working condition (defined as ‘superior professional and intermediate occupations’—corresponding to the class I and II of Standard Occupational Classification 2010,23 ‘skilled (non-manual and manual) occupations’, ‘semi-skilled and unskilled occupations’ and ‘others’—housewife, student, unemployed) and the household income (⩽1000, 1001–1500, 41500 euros per month). Maternal grandparents’ education was evaluated at baseline, recorded as the highest level of education achieved and categorized in this analysis into ⩽ 5, 45 schooling years. Maternal prepregnancy weight was obtained by recall. At baseline, maternal height was measured (without shoes) to the nearest 0.1 cm. When measurement was not possible, height was registered as in the identity card. Maternal pre-pregnancy body mass index (BMI) was categorized using the WHO (World Health Organization)24 reference data and re-categorized into non-overweight o 25.0 kg/m2 and overweight/obese ⩾ 25.0 kg/m2. The evaluation at 2-years-of-age was also conducted using a previously designed protocol by face-to-face interview. The following variables were used in the present analysis: child’s current caregiver (parents, other family, kindergarten and babysitter/others, defined as the current child’s caregiver), siblings (none, younger, older; children with young and older siblings were grouped with the older siblings) and food consumption variables. The family structure definition includes the type of caregiver and siblings. The child’s current caregiver was defined based on the question, made to the parents: ‘Who, regularly, takes care of the child during the day?’ Child’s weight and height were measured by a team of experienced investigators, at 2-years-of-age. Weight was measured in light clothing and without shoes using a TANITA UM-018 digital scale (Tanita Corporation, Tokyo, Japan), and was recorded to the nearest 0.1 kg. Height was measured as the distance from the top of the head to the bottom of the feet without shoes using a fixed SECA 214 (SECA, Chino, CA, USA) stadiometer, to the nearest 0.1 cm. Child’s BMI was defined as weight in kg divided by height in metres squared. This variable was then categorized using specific cut-offs for sex and age from WHO 25 and re-categorized into underweight/normal (BMIo2 s.d.) and overweight/obesity (BMI ⩾ 2 s.d.).

sweets (two items: ‘chocolate’ and ‘candies’, including lollipop, gum and chewing gum). For each of the four energy-dense food groups the frequency of consumption was categorized as ‘at least once a week’, ‘less than once a week’ and ‘never’. The four food groups were combined into the same group ‘energy-dense foods’ and potential determinants were evaluated for a weekly and daily consumption, in comparison to a monthly consumption. The caregivers were asked to complete 2 day food records. Pearson’s correlation coefficients were calculated for key groups comparing the responses from the FFQ and those from the food records to assess the validity of the FFQ. With the exception of sweets (r = 0·531), a weak correlation was found at 2-years-of-age.

Ethics The project Generation XXI was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects/patients were approved by the Ethical Committee of the São João Hospital/ the University of Porto Medical School. The project was approved by the Portuguese Authority of Data Protection. Written informed consent was obtained from parents for the collection of information at baseline and follow-up evaluations.

Statistical analysis Mean (standard deviation) and frequency differences were compared through student’s t-test and Χ2-test, respectively. Multinomial logistic regression models (odds ratio (OR) and 95% confidence intervals (95%CI)) were fitted to estimate the associations between potential determinants and food consumption. All potential determinants were examined to estimate the independent associations between determinants and food consumption. Collinearity effects between education, maternal occupation and household income were tested. Parents’ education is described in literature as an important determinant of child’s consumption and it was included in the final model. Maternal occupation and household income were entered separately in the final model, after removing the education variable. Statistical analysis was performed using SPSS 20.0 software (SPSS INC. 2011, Chicago, IL, USA).

RESULTS Table 1 shows the comparison between the study sample characteristics and the remaining cohort. No statistical significant

Table 1. Comparison between characteristics of eligible participants and the remaining cohort evaluated at baseline Sample (n = 808)

Maternal age (years) Maternal education (years)

European Journal of Clinical Nutrition (2015) 47 – 54

s.d.

30.4 11.1

5.10 4.25

Mean 29.4 10.4

P

s.d. 5.51 o0.001 4.25 o0.001

n 399

% 49.4

n 3843

% 49.0

142 370

17.7 46.2

1721 3706

22.2 47.8

223 66

27.8 8.2

1771 560

22.8 7.4

0.001

Household income(euros per month) ⩽ 1000 248 1001–1500 211 41500 229

36.0 30.7 33.3

2776 1894 2078

41.1 28.1 30.8

0.035

Child’s sex (girl) Food consumption at 2-years-of-age. A food frequency questionnaire (FFQ) was used to gather information on food groups not expected to be consumed on a daily basis by children (for example, sweetened beverages, chocolate and pizza). Parents and caregivers were asked how often the child was currently consuming 17 food groups and six categories of frequency were considered, ranging from ‘never’ to ‘every day’. Daily consumption frequencies were calculated using the following conversion: ‘every day’ as 1 time per day, ‘3–6 per week’ as 0.643 times per day, ‘1–2 per week’ as 0.214 times per day, ‘1–3 per month’ as 0.067 per day, ‘less than once a month’ as 0.017 times per day and ‘never’ as 0. Four energy-dense food groups were defined, gathering in the same group different items: soft drinks (two items: ‘sweetened carbonated drinks’ and ‘other sweetened drinks’), salty snacks (three items: ‘crisps’, ‘pizzas’ and ‘burgers’), cakes (two items: ‘creamy cakes’ and ‘not creamy cakes’, including sweet pastries) and

Mean

Remaining cohort (n = 7839)

Maternal occupation Semi-skilled/unskilled Nonmanual/manual skilled Superior/intermediate Others

0.847

Note: in each variable, the total may not add to 808/7839 due to missing data.

© 2015 Macmillan Publishers Limited


49 differences were found regarding child’s sex ratio, while significant differences were found for the socio-demographic characteristics. Mothers included in the present sample were slightly older (mean = 30.4 years, s.d. = 5.10 vs 29.4 years, s.d. = 5.51) and more educated (11.1 years, s.d. = 4.25 vs 10.4 years, s.d. = 4.25). The consumption of energy-dense foods according to the three frequency categories considered is shown in Figure 1. Sweets and soft drinks were consumed at least once a week by 50 and 37.5% of the participants, respectively. Most of the children consumed cakes and salty snacks less than once a week (52.2% and 56.3%, respectively). Thirty-two per cent of children had a daily consumption of any energy-dense food (7.8% of soft drinks, 14.1% of sweets, 1.4% of cakes and 0.2% of salty snacks). Tables 2 and 3 present the crude and adjusted OR between potentials determinants and consumption of soft drinks and sweets. Increasing maternal age and education, increasing grandparents’ education and household income and also more specialized maternal occupations were significantly associated with a lower consumption of both soft drinks and sweets. The effect sizes associated with maternal education are greater than those associated with grandparent’s education but quite similar to those found for more specialized occupations and high income. Having older siblings was positively associated with a higher frequency of soft drinks (OR = 2.11, 95%CI: 1.29–3.47) and sweets (OR = 2.94, 95%CI: 1.60–5.40; OR = 1.88, 95%CI: 1.02–3.24; ⩾ 1 week vs never and o 1 week vs never, respectively), comparatively with children with no siblings. Children with kindergarten as main caregiver were less likely to have a consumption of at least once a week of soft drinks, compared with children with parents as main caregivers (OR = 0.38, 95%CI: 0.18–0.79). Being a girl was positively associated with a higher frequency (⩾1 per week) of sweets consumption (OR = 1.85, 95%CI: 1.10–3.11) (Tables 2 and 3). For the consumption of cakes and salty snacks, fewer significant associations were found with the potential determinants. Only maternal age and education were inversely associated with consumption of cakes (OR = 0.38, 95%CI: 0.20–0.73; OR = 0.42, 95%CI: 0.19–0.91, respectively) and salty snacks (maternal age, 435 years vs o 30 years: OR = 0.44, 95%CI: 0.20–0.95) (data available in online supplementary Table). Table 4 presents the associations between socioeconomic and family structure characteristics and a weekly and daily consumption of any energy-dense food, compared with a monthly consumption. Higher maternal age and education, household

% in energy-dense food consumption category

100% Never

90%

Less than once a week

80%

At least once a week

70% 60%

56.3% 52.2%

50.9%

50% 40%

37.5% 34.2%

32.8%

30%

28.3%

29.7%

26.6%

19.4%

20%

17.1%

15.0%

10% 0% Soft drinks

Sweets

Cakes

Salty Snacks

Energy-dense food groups consumption

Figure 1. Proportion of 2-year-old children consuming each energydense food group according to three categories of consumption (never, less than once a week and at least once a week). © 2015 Macmillan Publishers Limited

income and maternal grandparents’ education were associated with a lower consumption of this type of food. Presence of older siblings was associated with a daily intake of energy-dense food (OR = 1.97, 95%CI: 1.11–3.51). DISCUSSION In the present study, high levels of maternal age, education and occupation and also a high household income were inversely associated with the consumption of most of the energy-dense food studied, which corroborates findings of previous studies among older children. Research carried out within the British birth cohort the Avon Longitudinal Study of Pregnancy and Childhood (ALSPAC)26 showed that 3-year-old children with higher scores in a dietary pattern termed ‘junk’ (characterized by high positive loading for snack foods and high-fat foods) were the ones with younger and less educated mothers and with lower family income.14 Additionally, at 4- and 7-years-of-age, a component describing a diet based on junk-type foods was also associated with low levels of maternal education and age.15 Moreover, a study 12 conducted among 2 to 5-year-old US children showed an inverse association between household income and education level of the female head of the household and children’s sugar intake. Low maternal education and unfavourable working conditions were all associated with a higher intake of energydense foods among Flemish children.13 These results underline the prominent role of mothers in shaping the children’s food habits in early life. As only a few characteristics of the father were included in the present study, it was not possible to discuss fathers’ influence on children’s food consumption. Nevertheless, as mothers usually spend significantly more time in direct interactions with children in diverse familial situations, including mealtime, they seem to influence more children’s eating habits than fathers.18 The observed association of grandparents’ education on grandchildren’s food consumption supports a long-term influence of socioeconomic circumstances on children’s food consumption. This life-course approach is in favour of a social environment influence on parents when they were a child and then on their own child. Few studies have examined the influence of social background of grandparents on children’s food habits, focusing only on the nutritional needs of the grandparents and not on the children’s consumption 27 or carried on in cultures greatly family oriented.28 However, due to the colinearity between education of mothers and grandparents we could not confirm that the effect of the association of grandparent’s education with children’s consumption is independent of maternal education. Overall the effect sizes associated with maternal education and consumption of soft drinks and sweets were greater than those associated with grandparent’s education but similar to the association found for maternal occupation and household income. Having older siblings was a risk factor for a higher consumption of soft drinks, sweet and salty snacks. These findings are in accordance with previous studies,14,15 which also showed a positive association between having older siblings and the consumption of food high in fat and sugar. The older sibling could influence the youngest relative by bringing into the household ‘kid-appreciated’ foods such as sweetened beverages, crisps and nutrient-poor snacks. Furthermore, bearing in mind that these foods require little preparation time and culinary knowledge and skills, mothers who have more than one child might find difficulties in having time to prepare healthy meals. Children with a kindergarten as the main caregiver were less likely to consume soft drinks compared with children with parents as caregivers. A recent study29 described that the food offered to children at home is less healthy than those offered at child-care centres in the US and among the 16 child-care centres included in the study, none reported providing soft drinks to children. European Journal of Clinical Nutrition (2015) 47 – 54


50 Table 2.

Associations from a multinomial model of socioeconomic and family’s structure characteristics with consumption of soft drinks in 2-year-old

children Soft drinks o1 week vs never OR (95%CI)

⩾ 1 week vs never OR (95%CI)

n (%)

Crude

Adjusteda

n (%)

Crude

Adjusteda

58 (24.2) 126 (52.5) 56 (23.3)

1b 0.79 (0.51–1.25) 0.59 (0.36–0.99)

1b 0.61 (0.34–1.10) 0.51 (0.26–1.00)*

134 (44.5) 110 (36.5) 57 (18.9)

1b 0.30 (0.20–0.45) 0.26 (0.16–0.42)

1b 0.24 (0.13–0.42)*** 0.16 (0.08–0.32)***

(16.7) (26.1) (29.9) (27.4)

1b 1.20 (0.66–2.20) 0.90 (0.51–1.59) 0.47 (0.27–0.81)

1b 1.60 (0.76–3.35) 1.03 (0.50–2.10) 0.55 (0.27–1.13)

(26.6) (30.0) (29.3) (14.1)

1b 0.87 (0.50–1.49) 0.55 (0.33–0.92) 0.15 (0.09–0.26)

1b 0.63 (0.32–1.26) 0.44 (0.22–0.84)* 0.20 (0.10–0.41)***

Maternal grandmother’s education (years)c ⩽5 153 (81.4) 45 35 (18.6)

1b 0.58 (0.36–0.95)

1b 0.46 (0.26–0.83)**

196 (81.7) 44 (18.3)

1b 0.57 (0.36–0.90)

1b 0.51 (0.28–0.92)*

Maternal grandfather’s education (years)c ⩽5 132 (75.0) 45 44 (25.0)

1b 0.52 (0.33–0.83)

1b 0.42 (0.24–0.74)**

177 (75.3) 58 (24.7)

1b 0.48 (0.31–0.74)

1b 0.48 (0.28–0.84)**

Maternal occupationc Semi-skilled/unskilled Nonmanual/manual skilled Superior/intermediate Others

(19.6) (46.7) (26.7) (7.1)

1b 0.54 (0.31–0.93) 0.32 (0.18–0.56) 0.94 (0.38–2.36)

1b 0.65 (0.34–1.26) 0.40 (0.20–0.81)* 0.98 (0.34–2.86)

69 143 48 39

(23.1) (47.8) (16.1) (13.0)

1b 0.47 (0.28–0.78) 0.16 (0.09–0.29) 1.47 (0.64–3.36)

1b 0.44 (0.23–0.81)*** 0.22 (0.11–0.43)*** 1.12 (0.43–2.96)

Household income (euros per month)c ⩽ 1000 67 (33.0) 1001–1500 68 (33.5) 41500 68 (33.5)

1b 0.65 (0.39–1.06) 0.45 (0.28–0.74)

1b 0.69 (0.39–1.23) 0.53 (0.30–0.94)*

134 (51.7) 69 (26.6) 56 (21.6)

1b 0.33 (0.21–0.52) 0.19 (0.12–0.30)

1b 0.47 (0.27–0.83)** 0.34 (0.19–0.61)***

Child’s sex Boy Girl

1b 0.91 (0.64–1.30)

1b 0.96 (0.62–1.47)

152 (50.2) 151 (49.8)

1b 0.99 (0.71–1.37)

1b 1.08 (0.70–1.65)

(10.4) (34.2) (38.8) (16.7)

1b 0.64 (0.34–1.20) 0.87 (0.46–1.65) 1.42 (0.67–3.00)

1b 0.73 (0.34–1.57) 0.89 (0.41–1.92) 1.44 (0.58–3.58)

59 112 84 48

1b 0.37 (0.21–0.64) 0.33 (0.19–0.59) 0.72 (0.37–1.42)

1b 0.57 (0.28–1.161) 0.38 (0.18–0.79)** 0.76 (0.33–1.76)

113 (54.3) 7 (3.4) 88 (42.3)

1b 0.53 (0.21–1.33) 1.26 (0.85–1.85)

1b 0.59 (0.21–1.62) 1.26 (0.21–1.62)

123 (45.9) 18 (6.7) 127 (47.4)

1b 1.25 (0.61–2.56) 1.66 (1.15–2.40)

1b 1.54 (0.64–3.73) 2.11 (1.29–3.47)**

Maternal age (years) o30 30–35 435 Maternal education (years) ⩽6 7–9 10–12 412

Main caregiver Parents Other family Kindergarten Baby-sitter/others Siblings None Younger Older

39 61 70 64

47 112 64 17

125 (52.1) 115 (47.9) 25 82 93 40

79 89 87 42

(19.5) (37.0) (27.7) (15.8)

Abbreviations: CI, confidence interval; OR, odds ratio. The bold represents the OR that are statistically significant. aP-values *0.05, ** o0.01 and *** o 0.001 adjusted for maternal age and education, maternal pre-pregnancy body mass index, children’s body mass index, children’s main caregiver and siblings. b Reference class. cAdjusted for maternal age, maternal pre-pregnancy body mass index, children’s body mass index, children’s main caregiver and siblings.

In the present study, 32% of children at 2-years-of-age had a daily consumption of energy-dense foods (soft drinks, sweets, cakes or salty snacks), this number was low compared with those reported by studies in other developed countries.30,31 In 2008 a survey conducted in a national random sample of US children from birth up to age of 4 years, showed that the daily frequency of consumption among 21–23.9 month-children, was 38.2% for sweetened beverages, 31.8% for candies, 2.5% for cakes, 12.4% for pizza and 23.7% for salty snacks.30 Results from the National Diet and Nutrition Survey, designed to be representative of the UK population (2008–09), showed a high median intake, among children aged 4–10 years, of soft drink (86 g/day), buns, cakes, European Journal of Clinical Nutrition (2015) 47 – 54

pastries and fruit pie (16 g/day), sugar, preserves and confectionery (10 g/day) and crisps and savoury snacks (boys 7 g/day and girls 10 g/day).31 Even though the frequency of consumption was different in our sample, the same pattern was found as sweetened beverages and candies were the most consumed among the studied food groups. Moreover, in 2006 the American Heart Association highlighted that between 2- and 6-years-of-age, sweetened beverages and other snacks high in sugar were a major source of caloric intake.32 Data from the population-based birth cohort ALSPAC showed that 450% of the children at 18 and 43 months of age, during the 3-day recording period, were consuming savoury snacks, crisps, © 2015 Macmillan Publishers Limited


51 Table 3.

Associations from a multinomial model of socioeconomic and family’s structure characteristics with consumption of sweets in 2-year-old

children Sweets o1 week vs never OR (95%CI)

⩾ 1 week vs never OR (95%CI) n (%)

Crude

Adjusteda

1b 1.28 (0.62–2.66) 0.66 (0.29–1.49)

156 (38.1) 161 (39.2) 92 (22.5)

1b 0.52 (0.31–0.85) 0.46 (0.26–0.80)

1b 0.39 (0.20–0.78)** 0.19 (0.09–0.43)***

1b 0.86 (0.40–1.87) 0.89 (0.41–1.92) 0.55 (0.27–1.27)

1b 0.87 (0.32–2.33) 0.89 (0.33–2.38) 0.54 (0.21–1.37)

98 102 126 76

(24.4) (25.4) (31.3) (18.9)

1b 0.54 (0.26–1.12) 0.64 (0.31–1.32) 0.19 (0.10–0.37)

1b 0.40 (0.16–1.04) 0.50 (0.20–1.28) 0.29 (0.08–0.45)***

Maternal grandmother’s education (years)c ⩽5 166 (75.8) 45 53 (24.2)

1b 0.75 (0.43–1.30)

1b 0.70 (0.36–1.34)

64 (69.6) 28 (30.4)

1b 0.49 (0.29–0.84)

1b 0.52 (0.26–0.99)*

Maternal grandfather’s education (years)c ⩽5 141 (68.8) 45 64 (31.2)

1b 0.66 (0.39–1.10)

1b 0.62 (0.34–1.16)

236 (76.4) 73 (23.6)

1b 0.43 (0.26–0.72)

1b 0.47 (0.25–0.86)*

Maternal occupationc Semi-skilled/unskilled Nonmanual/manual skilled Superior/intermediate Others

1b 0.78 (0.39–1.55) 0.61 (0.31–1.23) 1.75 (0.51–5.98)

1b 0.35 (0.12–0.99)* 0.30 (0.10–0.87) 0.95 (0.19–4.69)

86 200 81 41

(21.1) (49.0) (19.9) (10.0)

1b 0.64 (0.34–1.21) 0.26 (0.14–0.51) 1.67 (0.52–5.39)

1b 0.30 (0.11–0.81)* 0.17 (0.06–0.47)*** 0.68 (0.15–3.16)

Household income (euros per month)c ⩽ 1000 75 (31.1) 1001–1500 77 (32.0) 41500 80 (36.9)

1b 0.75 (0.38–1.47) 0.43 (0.23–0.78)

1b 0.91 (0.42–1.97) 0.53 (0.26–1.06)

154 (44.1) 108 (30.9) 87 (24.9)

1b 0.51 (0.27–0.97) 0.20 (0.11–0.36)

1b 0.90 (0.43–1.91) 0.33 (0.17–0.66)**


140 (50.7) 136 (49.3)

1b 1.25 (0.81–1.92)

1b 1.59 (0.95–2.68)

201 (48.9) 210 (51.1)

1b 1.34 (0.89–2.02)

1b 1.85 (1.10–3.11)*

Main caregiver Parents Other family Kindergarten Baby-sitter/others

36 101 107 32

1b 0.51 (0.23–1.13) 0.57 (0.25–1.28) 0.53 (0.21–1.39)

1b 0.54 (0.20–1.48) 0.53 (0.19–1.43) 0.41 (0.13–1.31)

62 161 121 67

(15.1) (39.2) (29.4) (16.3)

1b 0.47 (0.22–1.01) 0.37 (0.17–0.80) 0.65 (0.27–1.59)

1b 0.67 (0.25–1.79) 0.38 (0.14–1.01) 0.74 (0.24–2.27)

Siblings None Younger Older

123 (51.7) 12 (5.0) 103 (43.3)

1b 1.15 (0.42–3.21) 1.80 (1.11–2.94)

1b 1.09 (0.37–3.24) 1.88 (1.02–3.24)*

179 (48.9) 23 (6.3) 164 (44.8)

1b 1.52 (0.59–3.89) 1.97 (1.24–3.14)

1b 2.06 (0.73–5.85) 2.94 (1.60–5.40)***

Maternal age (years) o30 30–35 435

n (%)

Crude

57 (20.7) 153 (55.6) 65 (23.6)

1b 1.34 (0.78–2.32) 0.89 (0.48–1.63)

(14.9) (24.5) (26.4) (34.2)

Maternal education (years) ⩽6 7–9 10–12 412

40 66 71 92

42 119 92 21

(15.3) (43.4) (33.6) (7.7)

(13.0) (36.6) (38.8) (11.6)

Adjusteda

Abbreviations: CI, confidence interval; OR, odds ratio. The bold represents the OR that are statistically significant. aP-values *0.05, **o0.01 and ***o0.001adjusted for maternal age and education, maternal pre-pregnancy body mass index, children’s body mass index, children’s main caregiver and siblings. bReference class. c Adjusted for maternal age, maternal pre-pregnancy body mass index, children’s body mass index, children’s main caregiver and siblings.

chocolate confectionery and fizzy drinks.33 Although data is not directly comparable, our sample seems to have a lower consumption compared with UK children. It is possible to hypothesise that this lower consumption could have an effect on children’s BMI. In Europe the prevalence of overweight and obesity at 2-years-of-age, using the WHO standard criteria, ranged from 5.2% in Cyprus to 20% in Poland. Particularly in England this prevalence was described as 11.7% and for Portugal, at 3-years-of-age, it was described as 10.8%.34 Although the prevalence of overweight/obesity in our sample (8.5%) is lower than in many European countries, there is already a significant difference between children consuming daily © 2015 Macmillan Publishers Limited

energy-dense food and those with a lower consumption (11.8% vs 7.1%, P = 0.038). This might suggest an early negative effect of high consumption of energy-dense food that warrants further investigation using longitudinal approaches. Moreover, a higher consumption at early ages of these foods has been associated with a higher consumption a few years later and also with a worst diet quality.35 Limitations and strengths Our own results could be, at some extent, biased. The FFQ was mainly answered by parents in face-to-face interviews who might European Journal of Clinical Nutrition (2015) 47 – 54


52 Table 4. Associations from a multinomial model of socioeconomic and family’s structure characteristics with consumption of any energy-dense food in 2-year-old children Energy-dense foodsa Weekly vs monthly consumption

Daily vs monthly consumption

Crude OR (95%CI)

Adjusted ORb (95%CI)

Crude OR (95%CI)

Adjusted ORb (95%CI)

Maternal age (years) o30 30–35 435

1c 0.39 (0.24–0.65) 0.52 (0.30–0.92)

1c 0.33 (0.17–0.66)** 0.35 (0.16–0.76)**

1c 0.25 (0.15–0.41) 0.28 (0.15–0.50)

1c 0.21 (0.10–0.44)*** 0.13 (0.06–0.31)***

Maternal education (years) ⩽6 7–9 10–12 412

1c 1.43 (0.74–2.75) 1.07 (0.58–1.98) 0.50 (0.28–0.88)

1c 1.36 (0.59–3.14) 0.90 (0.42–1.95) 0.50 (0.24–1.06)

1c 0.89 (0.46–1.73) 0.67 (0.36–1.24) 0.16 (0.09–0.30)

1c 0.66 (0.28–1.57) 0.38 (0.17–0.86)* 0.14 (0.06–0.33)***

Maternal grandmother’s education (years)d ⩽5 1c 45 0.57 (0.36–0.91)

1c 0.53 (0.30–0.95)*

1c 0.43 (0.25–0.73)

1c 0.35 (0.17–0.70)**

Maternal grandfather’s education (years)d ⩽5 1c 45 0.57 (0.36–0.91)

1c 0.56 (0.32–0.96)*

1c 0.46 (0.28–0.77)

1c 0.48 (0.25–0.90)*

Maternal occupationd Manual/unskilled Nonmanual/skilled Superior/intermediate Others

1c 0.68 (0.33–1.43) 0.42 (0.19–0.90)* 1.18 (0.34–4.17)

1c 0.84 (0.47–1.51) 0.26 (0.14–0.48) 1.97 (0.68–5.15)

1c 0.57 (0.26–1.24) 0.23 (0.10–0.54)*** 1.17 (0.32–4.24)

Household income (euros per month)d ⩽1000 1c 1001–1500 0.41 (0.23–0.73) 41500 0.32 (0.18–0.54)

1c 0.58 (0.30–1.13) 0.42 (0.22–0.80)**

1c 0.30 (0.17–0.55) 0.13 (0.07–0.23)

1c 0.46 (0.23–0.93)* 0.19 (0.09–0.39)***


1c 1.10 (0.77–1.59)

1c 1.22 (0.78–1.91)

1c 1.26 (0.85–1.86)

1c 1.54 (0.92–2.57)

Main caregiver Parents Other family Kindergarten Baby-sitter/others

1c 0.58 (0.30–1.12) 0.56 (0.29–1.09) 0.71 (0.32–1.56)

1c 0.76 (0.34–1.72) 0.69 (0.31–1.56) 1.13 (0.42–3.03)

1c 0.51 (0.26–1.01) 0.37 (0.18–0.74) 0.81 (0.36–1.82)

1c 0.95 (0.40–2.29) 0.49 (0.20–1.20) 1.27 (0.44–3.65)

Siblings None Younger Older

1c 1.01 (0.44–2.32) 1.30 (0.87–1.96)

1c 1.61 (0.62–4.17) 1.44 (0.87–2.38)

1c 1.12 (0.46–2.73) 1.58 (1.02–2.44)

1c 1.75 (0.58–5.35) 1.97 (1.11–3.51)*

1c 0.90 (0.52–1.58) 0.47 (0.27–0.83) 1.81 (0.67–4.88)

Abbreviations: CI, confidence interval; OR, odds ratio. The bold represents the OR that are statistically significant. aP-values *0.05, ** o 0.01 and *** o 0.001 includes soft drinks, salty snacks, cakes and sweets bAdjusted for maternal age and education, maternal pre-pregnancy body mass index, children’s body mass index, children’s main caregiver and siblings cReference category dAdjusted for maternal age, maternal pre-pregnancy body mass index, children’s body mass index, children’s main caregiver and siblings.

not be aware of all foods eaten by their children when being taken care of by others. Additionally, a social desirability bias could have occurred, when answering the FFQ, and implied a lower report of unhealthy food consumption. In this case, the associations found may be even stronger if food consumption was reported with more accuracy. The use of a qualitative FFQ for dietary data collection could be pointed out as another limitation of this study. We assumed only information from ‘frequency’ and not ‘quantity’, considering the same mean portion to all children, which could underestimate the European Journal of Clinical Nutrition (2015) 47 – 54

consumption of energy-dense food if the children usually eat more than one average portion each time. Nevertheless, in 2-year-old children this is less likely to occur compared with older children. Additionally, the study by Willet36 highlighted that portion sizes are positively correlated with frequency of use and for most food groups frequency is more important in evaluating food consumption of a group of individuals than portion sizes. The weak correlation found between 2 day food records and FFQ could be the result of a low intake of energy-dense food at © 2015 Macmillan Publishers Limited


53 this age, and a 2 day food record may not be a good method to validate low intakes. Compared with food records, the FFQ overestimated the consumption of energy-dense food, and as we did not expect to have an error different between groups, this would not compromise the conclusions of the present study regarding the associations found. The use of a sample of very young children, which is not so commonly described in previous studies, is a major strength of this study. Even though we performed a cross-sectional analysis at 2-years-of-age, due to the nature of socioeconomic characteristics and the fact that most of them were evaluated in the context of a cohort and therefore gathered at baseline, allowed us to establish a temporal sequence. Our sample derived from a populationbased cohort, with a balanced sex distribution, but with an unrepresentative proportion of older and higher educated mothers. We consider that these differences would not compromise the representativeness of our results as they do not seem to be relevant. In conclusion, an independent relation between socioeconomic characteristics and family structure, and consumption of energydense food, namely soft drinks and sweets, were observed in 2-year-old children. These findings support the influence of unfavourable socioeconomic characteristics, having older siblings and parents as main caregivers, on the consumption of unhealthy foods from the early stages of life. Moreover, a possible influence of long-term socioeconomic environment was shown through the grandparents’ education effect. Thus, an unfavourable socioeconomic environment has a long-term adverse influence on pre-school children’s food consumption, and this influence starts even before the child was born. Nutrition education and promotion of healthy food habits are needed and should be targeted at specific subpopulations. As preschool children are learning to eat properly and are acquiring family dietary behaviours, they represent a perfect age group to intervene. CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS We thank all members of the research team and the staff of Generation XXI for their essential and generous assistance. Generation XXI has been funded by the Operational Health Programme—XXI Health, Community support framework III (co-funded by feder), Administração Regional de Saúde do Norte, Fundação Calouste Gulbenkian and Fundação para a Ciência e Tecnologia (FCT—PTDC/SAU-ESA/108577/2008). The project Generation XXI was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects/patients were approved by the Ethical Committee of the São João Hospital/ the University of Porto Medical School.

REFERENCES 1 Malik VS, Schulze MB, Hu FB. Intake of sugar-sweetened beverages and weight gain: a systematic review. Am J Clin Nutr 2006; 84: 274–288. 2 Malik VS, Popkin BM, Bray GA, Despres JP, Willett WC, Hu FB. Sugar-sweetened beverages and risk of metabolic syndrome and type 2 diabetes: a meta-analysis. Diabetes Care 2010; 33: 2477–2483. 3 Rosenheck R. Fast food consumption and increased caloric intake: a systematic review of a trajectory towards weight gain and obesity risk. Obes Rev 2008; 9: 535–547. 4 Rolls BJ. The role of energy density in the overconsumption of fat. J Nutr 2000; 130: 268S–271SS. 5 Ebbeling CB, Pawlak DB, Ludwig DS. Childhood obesity: public-health crisis, common sense cure. Lancet 2002; 360: 473–482. 6 Johansson I, Holgerson PL, Kressin NR, Nunn ME, Tanner AC. Snacking habits and caries in young children. Caries Res 2010; 44: 421–430.


7 Mackenbach JP, Stirbu I, Roskam AJ, Schaap MM, Menvielle G, Leinsalu M et al. Socioeconomic inequalities in health in 22 European countries. N Engl J Med 2008; 358: 2468–2481. 8 Marmot MG, Smith GD, Stansfeld S, Patel C, North F, Head J et al. Health inequalities among British civil servants: the Whitehall II study. Lancet 1991; 337: 1387–1393. 9 Mishra G, Ball K, Arbuckle J, Crawford D. Dietary patterns of Australian adults and their association with socioeconomic status: results from the 1995 National Nutrition Survey. Eur J Clin Nutr 2002; 56: 687–693. 10 Sanchez-Villegas A, Delgado-Rodriguez M, Martinez-Gonzalez MA, De Irala-Estevez J. Gender, age, socio-demographic and lifestyle factors associated with major dietary patterns in the Spanish Project SUN (Seguimiento Universidad de Navarra). Eur J Clin Nutr 2003; 57: 285–292. 11 Pabayo R, Spence JC, Cutumisu N, Casey L, Storey K. Sociodemographic, behavioural and environmental correlates of sweetened beverage consumption among pre-school children. Public Health Nutr 2012; 24: 1–9. 12 Kranz S, Siega-Riz AM. Sociodemographic determinants of added sugar intake in preschoolers 2 to 5 years old. J Pediatr 2002; 140: 667–672. 13 Vereecken C, Maes L. Young children’s dietary habits and associations with the mothers’ nutritional knowledge and attitudes. Appetite 2010; 54: 44–51. 14 North K, Emmett P. Multivariate analysis of diet among three-year-old children and associations with socio-demographic characteristics. The Avon Longitudinal Study of Pregnancy and Childhood (ALSPAC) Study Team. Eur J Clin Nutr 2000; 54: 73–80. 15 Northstone K, Emmett P. Multivariate analysis of diet in children at four and seven years of age and associations with socio-demographic characteristics. Eur J Clin Nutr 2005; 59: 751–760. 16 Rogers I, Emmett P. The effect of maternal smoking status, educational level and age on food and nutrient intakes in preschool children: results from the Avon Longitudinal Study of Parents and Children. Eur J Clin Nutr 2003; 57: 854–864. 17 Birch LL, Fisher JO. Development of eating behaviors among children and adolescents. Pediatrics 1998; 101: 539–549. 18 Scaglioni S, Salvioni M, Galimberti C. Influence of parental attitudes in the development of children eating behaviour. Br J Nutr2008; 99: S22–S25. 19 Patrick H, Nicklas TA. A review of family and social determinants of children’s eating patterns and diet quality. J Am Coll Nutr 2005; 24: 83–92. 20 Alves E, Lunet N, Correia S, Morais V, Azevedo A, Barros H. Medical record review to recover missing data in a Portuguese birth cohort: agreement with selfreported data collected by questionnaire and inter-rater variability. Gac Sanit 2011; 25: 211–219. 21 Larsen PS, Kamper-Jørgensen M, Adamson A, Barros H, Bonde JP, Brescianini S et al. Pregnancy and birth cohort resources in europe: a large opportunity for aetiological child health research. Paediatr Perinat Epidemiol 2013; 27: 393–414. 22 Alves E, Azevedo A, Correia S, Barros H. Long-term maintenance of smoking cessation in pregnancy: an analysis of the birth cohort generation XXI. Nicotine & Tob Res 2013; 15: 1598–1607. 23 Office for National Statistics. Standard Occupational Classification 2010. Volume 1: Structure and Descriptions of Unit Groups. The Stationery Office: London, 2010. 24 World Health Organization. Obesity: preventing and managing the global epidemic. World Health Organization, Geneva, Switzerland, 2000. WHO Technical Report Series 894. 25 WHO Multicentre Growth Reference Study Group. WHO Child Growth Standards: Length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: Methods and development. World Health Organization, Geneva, Switzerland, 2006. 26 Golding J, Pembrey M, Jones R. ALSPAC--the Avon Longitudinal Study of Parents and Children. I. Study methodology. Paediatr Perinat Epidemiol2001; 15: 74–87. 27 Kicklighter JR, Whitley DM, Kelley SJ, Shipskie SM, Taube JL, Berry RC. Grandparents raising grandchildren: a response to a nutrition and physical activity intervention. J Am Diet Assoc 2007; 107: 1210–1213. 28 Jingxiong J, Rosenqvist U, Huishan W, Greiner T, Guangli L, Sarkadi A. Influence of grandparents on eating behaviors of young children in Chinese three-generation families. Appetite 2007; 48: 377–383. 29 Erinosho TO, Dixon LB, Young C, Brotman LM, Hayman LL. Caregiver food behaviours are associated with dietary intakes of children outside the child-care setting. Public Health Nutr 2012; 10: 1–10. 30 Siega-Riz AM, Deming DM, Reidy KC, Fox MK, Condon E, Briefel RR. Food consumption patterns of infants and toddlers: where are we now? J Am Diet Assoc 2010; 110: S38–S51. 31 Whitton C, Nicholson SK, Roberts C, Prynne CJ, Pot GK, Olson A et al. National Diet and Nutrition Survey: UK food consumption and nutrient intakes from the first



54 year of the rolling programme and comparisons with previous surveys. Br J Nutr 2011; 106: 1899–1914. 32 Gidding SS, Dennison BA, Birch LL, Daniels SR, Gillman MW, Lichtenstein AH et al. Dietary recommendations for children and adolescents: a guide for practitioners. Pediatrics 2006; 117: 544–559. 33 Emmett P, Rogers I, Symes C. Food and nutrient intakes of a population sample of 3-year-old children in the south west of England in 1996. Public Health Nutr 2002; 5: 55–64.

34 Cattaneo A, Monasta L, Stamatakis E, Lioret S, Castetbon K, Frenken F et al. Overweight and obesity in infants and pre-school children in the European Union: a review of existing data. Obes Rev 2010; 11: 389–398. 35 Vilela S, Oliveira A, Ramos E, Moreira P, Barros H, Lopes C. Association between energy-dense food consumption at 2 years of age and diet quality at 4 years of age. Br J Nutr2013; 14: 1–8. 36 Willet W. Food-Frequency Methods. Nutr Epidemiol2nd ed. Oxford University Press: New York, 1998, pp 74–91.

Supplementary Information accompanies this paper on European Journal of Clinical Nutrition website (http://www.nature.com/ejcn)

