Maternal and child obesity is a matter of concern in the United Kingdom (UK) with 1 in 5 women of childbearing age (16–44) categorised as obese (body mass index >30 kg/m²) [1]. Throughout the UK, 1 in 20 pregnant women have been classed as obese (body mass index >35 kg/m²) with a higher proportion found in Wales as 1 in 15 women [2]. Evidence suggests that infants (aged 0–12 months) born to obese mothers are over nourished in the womb, potentially leading to changes in metabolism, behaviour and appetite regulation. These infants are often larger at birth, show increased adipose tissue mass and obesity, and are likely to develop insulin resistance in later life [1,3]. Further data indicates that infant weight gain during the first year of life and even weight status at age 6 months can also have implications for future obesity status, diabetes and other cardiovascular risk factors [4]. Overweight infants are more likely to become overweight children and continue on this trajectory into adulthood [5,6]. Although UK national measurement data on infant weight is not available, data for pre-school aged children suggests an increasing incidence of obesity among young children. The 2010 Health Survey England (HSE) found that 13 percent of English toddlers aged 2–3 years were considered overweight and 10 percent were classed as obese [7]. In Wales the first National Child Measurement Programme (NCMP) published in 2013 found that 22 percent of reception class pupils (4–5 year olds) were found to be overweight or obese [8].

Research evidence indicates that pregnancy presents an opportune time for early prevention of obesity during the life course [9,10]. Infant diet is dependent on choices made by its parents [11,12]. Observational studies suggest that exposure to an ‘obesogenic’ environment often begins in the first 2 years of life. This represents a critical window of opportunity to establish healthy taste preferences that will impact on later health [4,13]. WHO/Europe [14] have restated in a recent report that mothers should aim to exclusively breastfeed their infants until 6 months of age. There is evidence that formula-fed infants consume more calories and are likely to be 2.5 times more obese by age 2 years compared to those who are breastfed for the first 6 months of life [15]. WHO/Europe also recommend that after 6 months of age infants are introduced to a wide variety of foods, including vegetables, to support healthy taste development early on. In the UK the National Institute for Clinical Excellence [10] recommend that mothers maintain a healthy weight before, during and after pregnancy to promote health in their offspring.

Many studies indicate that high consumption of fruit and vegetables by children correlates positively with parental modelling and consumption of fruit and vegetables [16-19]. The National Diet and Nutrition Survey [20] and the Welsh Health Survey [21] found that approximately only a third of adults were meeting the ‘Eat 5 a Day’ recommendation of fruit and vegetable consumption, and indicate that fruit and vegetable consumption is socially patterned with families from the most deprived neighbourhoods consuming less fruit and vegetables compared to those living in affluent neighbourhoods. A closer inspection of the prevalence of maternal and child obesity and overweight data presented above show that childhood obesity also correlates significantly with neighbourhood deprivation. In 2013, European ministers of health recognized the disproportionate impact of obesity on those who are socioeconomically disadvantaged, in the Vienna Declaration [14,22]. They acknowledged the need to: promote access to healthy and affordable food, intervene across the life course, and address gaps in food governance system.

For more than a decade the UK Government has raised concerns about the overconsumption of food high in salt, fat and sugar and the long term negative health consequences for children and health inequality [23]. Attempts to change dietary behaviours have ranged from providing information and healthy lifestyle incentives, to regulatory measures such as the school meal guidelines in England, Wales and Scotland [24-26].

The UK government dietary recommendations [27] advise parents to provide a diet high in complex carbohydrates (whole grains), fruit and vegetables and low fat dairy products, which are more satiating than fat and sugar based products [28]. The Change4life [29] social marketing campaign has been running in the UK since 2009 offering advice to families on how to make small changes to improve their dietary and lifestyle habits. Similarly the government’s Responsibility Deal [30] has initiated voluntary agreements with the food industry to reduce the salt [31] and fat content in food [32]. Recently, campaigners have now added calls for the reduction of sugar found in processed food aimed at children to this list [33]. One study indicates that brands of baby food sold in the UK contain total sugar content higher than 10 percent [34] (government guidelines recommend that no more than 10 percent of average total energy intake should be consumed as sugar) [35]. Critics of the Responsibility Deal have claimed that the policy remains is vague as there are no sanctions or timescales indicating when targets on food reformulation are to be met [36,37].

Government regulatory measures to reduce obesity have included improved food labelling to support consumers to make informed dietary choices at the point of purchase [38] and restricted advertising of unhealthy food products aimed at children on television [39]. However food industry lobbying has led to the failure of the European Union to implement a uniform method of ‘Traffic Light’ labelling system (preferred by consumers [40,41]) across Europe [42], and ‘junk food’ advertisements continue to be viewed by children during family viewing times [43].

The current levels of obesity demonstrate that education on how to make ‘healthy choices’ about food does not necessarily guarantee the adoption of improved dietary habits by the population. Despite the large amount of dietary advice and information found in the media, in schools, on food packaging, in health centres and in doctors’ surgeries, obesity levels remain high. A systematic review of nutritional education interventions aimed at reducing obesity in children less than 2 years of age found limited effect on family diets or parental attitudes towards healthy eating [4]. The review concluded that it was difficult to determine if this lack of improvement was due to small sample sizes, short intervention periods, weak dose effects, or ineffective interventions. Another review [44] to assess the effectiveness of prenatal and postnatal interventions in extending duration of breastfeeding found that long term combined face-to-face information, guidance, and support from a varied range of professionals was most effective. A comprehensive review of recent systematic reviews and meta analyses that looked at interventions to prevent childhood obesity [45] identified that interventions were modestly effective when professional support, education and behaviour change theory was applied concurrently. The review concluded that individual behaviour change interventions would be far more effective if they were developed alongside population-based strategies (to improve the built environment for physical activity; to improve government and industry policy on food quality, availability and pricing; to improve government policy to tackle socioeconomic disadvantage and reduce advertising of unhealthy food).

The theory of reciprocal determinism states that an individual shapes his/her social environment and can also be shaped by it. Thus individual pre-existing behaviours can become entrenched in their social environment and render them resistant to change [46,47]. The failure of health messages to make an impact on rising obesity levels could be related to the poor application of theoretical models or frameworks during the development of such interventions.

Previous research suggests that perception and understanding of risk and the ability to change varies according to socio-demographic factors such as income, education, gender, and ethnicity [48] and the lived environment [49]. Most interventions on obesity have relied on individual-level motivation and education to make good lifestyle choices, rather than identifying ways to improve peoples’ ability to effectively make that change [50]. A study of cancer patients, illustrates that people diagnosed with cancer failed to stop smoking and were unable to change their behaviour to improve their life expectancy, despite being aware of the long-term adverse effects on their health and wellbeing [40].

Current developments in health promotion have acknowledged the importance of broader social and environmental influences on lifestyle choices [49,51-53]. This is reflected in the socio-ecological paradigm that encompasses several different research perspectives, and advocates multilevel interventions both at the individual and community/population-level. The socio-ecological paradigm suggests that the social, community and built environment, along with personal factors are all likely to predict health behaviours [49,53-56]. Multilevel community and population-wide interventions are found to work successfully if programmes are designed to take on a longer-term perspective on health improvement, with the goal to ultimately change the wider social environment so that healthy behaviours can be facilitated within the population [57-60]. Merzel and D’Afflitti [58] conclude that population or community-wide interventions work most effectively when implemented on three levels: 1) individual or family-targeted support that can be offered to high risk populations, 2) community or population-wide social marketing campaigns that can facilitate changes in social norms, and 3) policy-level levers that can be employed to modify the social and political environment, so that people’s ability to make lifestyle changes is enhanced. Parallels can be drawn from the successful implementation of the antismoking ban in the UK which adopted multi-level population wide intervention strategies and has shown year–on-year reductions in the number of people smoking in the UK [61].

In the UK, the Medical Research Council (MRC) provides comprehensive guidance for developing and evaluating complex multilevel interventions. The 2007 updated guidance [62] highlights the importance of understanding local contexts, attitudes, and beliefs using an iterative approach to inform intervention pathways. Similarly, the socio-ecological paradigm advocates that both content-specific and contextual data should be taken into account when informing the development of appropriate interventions [53,55]. A qualitative approach to data collection provides the necessary tool to achieve this iterative process of inquiry. Qualitative research has the potential to contribute to a better understanding of the real-world user perspective (for whom interventions and policies are planned) within the context of wider socio-economic and environmental determinants of health [63]. This kind of formative research is used to inform and design interventions, models or messages with the target audience [64].

There have been many examples of interventions to prevent childhood obesity that have included settings outside the home such as schools, local communities and ‘early year’s’ settings, usually focused on behaviour-change outcomes, for example The Healthy and Sustainable Pre-school Scheme [65]. However Summerbell et al. [66] highlight that there have been few studies which have focused on multi-level intervention for childhood obesity, and to date very few studies have explicitly used the MRC framework for developing and evaluating interventions [67,68]. Furthermore, there are even fewer effective intervention strategies aimed at children under 5 years of age, especially for infants aged 0–3 years [4,69].

This current study used qualitative interviews with parents to elicit evidence on the main barriers and facilitators to dietary choice, and to inform the development of interventions that they would like to see put in place to promote a healthier food environment for their children. We focused specifically on priorities identified by parents at the community/population-level, thus addressing a significant gap in the literature.

Participants for this qualitative study were purposively sampled for a diversity of opinions from those participating in an existing birth cohort study “Growing Up in Wales” [70]. The cohort study aims to examine how the environment in which a child lives influences their health in terms of obesity, asthma and injury. Participants were recruited for the ‘Growing Up in Wales’ cohort study when they attended maternity appointments in hospitals and clinics within Abertawe Bro Morgannwg (ABM) University Health Board. Participants who agreed to take part in the cohort study were provided with an information sheet detailing areas of the study. The study protocol and consent forms were approved by the South East Wales Research Ethics Committee for Wales (09/WSE02/37). Recruitment and subsequent research for the birth cohort study and for this current study was conducted according to the ethical principles for medical research set by the Declaration of Helsinki [71]. All participants were aged 16 years and older. They gave their own written informed consent to take part in the ‘Growing up in Wales’ cohort study and again prior to taking part in this qualitative study. Participants were interviewed for this study either during pregnancy or when their child(ren) was/were 12 months old.

Data collection for this present study was conducted according to the guidelines for inductive qualitative research, which includes semi-structured interviews as a means of data collection [72,73]. An inductive approach was chosen as it permits research findings to be influenced by both research objectives and raw data, and provides in-depth rich descriptions. Parental decisions on infant diets are influenced by many factors, therefore we selected face-to-face semi-structured interviewing rather than group data capture methods for this study. Semi-structured interviews offered the opportunity to build rapport [74] with the interviewees, to explore socio-cultural and environmental influences on family dietary choices, thus allow reflection on parents’ described ‘lived experiences’ [75].

A semi-structured interview schedule (see Interview schedule below) was developed underpinned by the socio-ecological paradigm which suggests that the social, community and built environment, and personal factors are likely to predict health behaviours [49,53,54]. The interview topics focused on: (a) parents’ knowledge and views about living a healthy lifestyle and if/how this influenced their current dietary choices for their families; (b) multi-level barriers and facilitators to healthy dietary choices; and (c) population-level interventions and policy changes that parents would like to see put in place to facilitate healthy dietary behaviour for their families. The interview schedule evolved in keeping with the research aims to explore emerging concepts. Interviews took place in the family home (with the exception of two interviews, which were conducted in the workplace).

Completion of data analysis produced three key themes: ‘community’, ‘family and individual’, and ‘policy-level influences’ on dietary choices. Finally recommendations from parents on ways to improve dietary choices for young families were sought.

This study examined factors influencing infant diet and included parents’ recommendations on ways to improve family and infant dietary choices. Parents who participated in the study understood the pre-requisites of a healthy diet, and information such as ‘Eat 5 a Day’ [29] or the ‘Traffic Light’ system of food consumption, but were often unsure how to make key changes in their own lives. There were various barriers to providing a healthy family diet, these included: the ready availability of fast food; difficulty accessing fresh fruit and vegetables; peer pressure from others in the community to conform to typical diets, or to not deny the family affordable (unhealthy) treats.

There was a fatalistic, or apathetic approach to the long term health consequences of poor dietary choices among some families living in deprived neighbourhoods. Parents’ lives were evaluated in the context of the environment they lived in [83] and the dangers associated with a diet high in fat, salt and sugar were considered a less risky lifestyle choice, compared to living in a deprived neighbourhood.

The findings highlight the significance of past and present dietary experiences of parents. Parents’ own childhood food experiences, their own parents’ ability to cook, the parents’ individual food preferences, and their personal views on the best way to feed children influenced the food choices they made for their infants.

Interestingly we found that 77 percent of mothers in this study hoped to or had initiated breastfeeding. The data is comparable to Welsh national breastfeeding data from 2010 which shows breastfeeding initiation rates at 71 percent. Welsh data also indicates that 23 percent of mothers were breastfeeding their children up to the age of 6 months. In our study 41 percent (n = 25) of mothers planned to or were exclusively breastfeeding their infants. Whilst 36 percent (n = 22) chose to breast and bottle feed and 23 percent (n = 23) had decided to just bottle feed. Only one mother (from an affluent neighbourhood) in this study had commented that there was too much pressure placed on women to breastfeed. She chose to forego breastfeeding as it would be more convenient for her when she returned to work, as a teacher after the birth of her baby.

The Welsh government has had a breastfeeding strategy in place since 2001 [84] and has supported local breastfeeding initiatives which complement its strategy to reduce obesity levels in Wales. The UK government also has policies in place to support women who return to work and would like to continue to breastfeed [85]. Whilst policies and strategies are in place, widespread sociocultural barriers to breastfeeding remain [86].

The main policy level recommendations from parents focused on the accessibility and affordability of good quality, healthy food: they wanted improved ease of access to low cost healthy foods (in their local area and further afield), reduced promotion of unhealthy food (in their local area and in supermarkets), provision of targeted advice to fathers and minority ethnic parents, and tailored and practical advice and information on how to manage time, purchase, prepare, store and cook food for themselves and their children. There was a strong message to policy makers to work with commercial companies (food manufactures) as they have the necessary skills and resources to lower costs, and target messages at different groups of people. Parents were of the opinion that food providers should be subject to legislation, to improve the quality of their food and they should restrict access to and the sale of unhealthy food. Similar recommendations were made for healthy food provision in public facilities such as hospitals, workplaces and schools.

Opinions expressed by parents’ in this study are supported by existing literature. It is recognised that family food choices are influenced by complex factors outside their control. Cultural beliefs, values and local availability of healthy food have been identified as barriers by many others [87-89]. Furthermore, family income, convenience, and food preferences of family members [13,16] such as the father [90], are additional barriers to ‘healthy choices’.

Research indicates that low cost, energy dense, and low nutrient foods high in salt, sugar and fat, are consumed largely by low income families [91,92], highlighting a link between poor dietary choices and deprivation. These families have adapted to structural inequality by developing irregular meal consumption patterns [79]. There was a perception among interviewees in this study that healthy food costs more and this coincided with the general low consumption of fruit and vegetables described overall [13,19]. Supporters of healthy eating have called on the government to do more. They have suggested greater regulation of the food industry, clearer labelling and introduction of taxes on unhealthy food [93,94]. The parents in our study made similar requests and desired the provision of ‘good quality’, fresh produce at affordable prices [95]. They questioned why processed food was cheaper than fresh produce, but on the question of taxing unhealthy food they held mixed views [96].

It is well documented in the literature that an over-abundance of fast food outlets as opposed to local fruit and vegetable retailers in an area is conducive to an ‘obesogenic environment’ [97-100], that increases the opportunity to consume unhealthy food. We found that not only are there macro obesogenic environments but respondents also suggested that micro-obesogenic environments can be found in supermarkets. In-store marketing of unhealthy food and promotional offers on processed food encouraged unplanned meals and over-consumption [101]. Providing recipes on packages of fresh produce could be a possible way to mitigate against the effects of such environments as suggested by parents.

It has been argued that in an obesogenic environment the degree of influence parents can have on their growing child’s diet is questionable [102]. In fact, the social pressure on parents to provide fast food to their child echoes Bourdieu’s theory of ‘cultural capital’ [103], where similarities in taste and dietary behaviour is perpetuated by the community to create a sense of collective identity. This was most prevalent among families from deprived neighbourhoods. Consuming fast food was seen as a treat [104] and had taken on a “…‘market value’ in the struggle for privilege…” [105] that became an alternative way to achieve a designated status in the community and practised as normalised behaviour [80,106]. In contrast parents living in affluent areas mostly talked about the importance of eating healthy meals at home and outside the home, and negated the consumption of unhealthy processed or fast food.

Research has shown that infant dietary taste and preferences start to develop in the womb and continue to develop in early infancy (encouraged by breast milk consumption) [13,107,108]. Good dietary habits are developed in the first three years of a child’s life [109,110]. Our findings did not completely support this viewpoint, because we found dietary habits were flexible and greatly influenced by changing family circumstances and by life experiences (e.g. learning to cook, healthy peers promoting a healthy diet), highlighting the social influences on diet [111].

The findings here indicate that parental education and professional employment protect infants from an unhealthy diet [112]. Parents in our study who had had the opportunity of a university education (including those who had been bought up in deprived neighbourhoods) had more exposure to new food and improved or developed their cooking skills as a consequence of living independently. The ability to cook meals from fresh ingredients influenced dietary choices and this is supported by the literature [113,114]. However, parents were critical about cookery lessons they had received in school. They asked for improved cookery lessons in school, sought tailored advice aimed specifically at fathers, or ethnic minority families, or advice as to how they could change their socio-economic circumstances for the better. Marks et al. [115] have argued that health-orientated messages need to be simple, clear and consistent to be effective. Others have suggested that cogeneration of knowledge [116] regarding risk to health is required between policy makers and the target community/population, to increase adherence to health messages and prevent ill health.

Parental recommendations for improving infant diets focused on reducing barriers to accessing high-quality, low cost, healthy foods in local communities, workplaces and schools. Legislation was recommended to reduce access and affordability of unhealthy foods and to improve the obesogenic retail environment and family work-life balance. Recent policy initiatives complement some of the recommendations made by parents such as the mandatory guidelines in Wales and Scotland aimed at improving hospital food [118,119] and school meals in England, Wales and Scotland [24-26].

To enact the recommendations suggested by parents in this study will require health professionals, local and national policy makers, the media and the food industry to engage with both mothers and fathers, to challenge determinants of inequality in dietary intake, and to tailor information towards those most in need. In the current food environment, untailored support and advice was found to be particularly unsuitable for families in deprived neighbourhoods, who had a tendency to embrace a more fatalistic attitude towards their family’s health and wellbeing. They saw many aspects of their lives out of their control and understanding health risks associated with a poor diet was seen as less of a priority. Parents would benefit from a participatory approach to policy making that can create local environments that reduce inequality, meets their needs and supports them to make healthy dietary choices as recommended by the recent NHS England report ‘Five Year Forward View’ [120].