The aim of the present paper is to consider: (a) the context for nutrition interventions amongst women living in poverty (and some of the difficulties in undertaking trials in this population subgroup) in the UK; (b) the rationale for why nutrition interventions might be important; (c) intervention approaches in the UK and USA (and limitations of these approaches); (d) implications of the evidence base for policy and practice.
The context for nutrition interventions amongst women living in poverty
In 2003–4 twelve million individuals in Britain were living in poverty, of whom 3·5 million were children, approximately 2·5 million were adults living with those children, approximately 3·5 million were working-age adults living without any dependent children and two million were pensioners (Glennerster et al. Reference Glennerster, Hills, Piachaud and Webb2004). In 2003–4 levels of poverty were defined as <60% of the contemporary mean, e.g. the threshold for a two-adult household was £180/week (Palmer et al. Reference Palmer, Carr and Kenoway2005).
Within this very heterogeneous population group there will be specific subgroups such as women with disabilities, single mothers and women from ethnic minorities. Recent decreases in the number of individuals living in poverty have been helped by increasing rates of employment. However, being in employment does not mean being rescued from poverty. Approximately 3·7 million female and 1·8 million male employees aged ≥22 years were paid <£6·50/h in 2004 and approximately half these were part-time workers, mainly women. The number of working households who are in receipt of in-work benefits has more than trebled since 1996 (Palmer et al. Reference Palmer, Carr and Kenoway2005). Furthermore, one-third of women who find work do not have that work 6 months later. Over the last decade the employment rate among lone parents has risen from approximately 45% to approximately 55%. It is becoming increasingly clear that to make further decreases in the number of individuals living in poverty in the UK that low pay (and other conditions at work) and higher out-of-work benefits will need to be seriously explored.
From a ‘determinants of health’ perspective (Dahlgren & Whitehead, Reference Dahlgren and Whitehead1991), living in low-income households will be relevant for every level of health influence, ranging from household to community to the wider environment. Living in conditions of relative poverty can exert a lifelong (and often intergenerational) effect on economic opportunities, life circumstances, individual health behaviours (such as smoking and dietary intake) and subsequent morbidity and mortality. Individuals from relative poverty have lower educational attainment. Unemployed individuals are more likely to be the victims of violent crime and lone parents are more than twice as likely as average to be burgled, and more likely not to have home insurance. Subsequently, fear of crime is also greater for individuals with lower incomes. Among those individuals aged >60 years, for example, 36% of women from low-income households report that they feel unsafe out at night (Palmer et al. Reference Palmer, Carr and Kenoway2005).
In terms of health, under-age pregnancy, low birth weight and infant death are classic indicators of poverty and health outcomes. However, there is a much wider range of lifestyle-related disorders that are more evident in poor women throughout the lifespan. Poorer women have worse levels of mental health and coping capacity to deal with life circumstances, without considering the efforts required to attain and consume a healthy balanced diet for themselves. Among the poorest 20% of the population, approximately 25% are judged to be at risk of developing mental health problems, just about twice the rate for individuals on average incomes.
The rationale for why nutrition interventions might be important
In the UK approximately one in five men and one in six women die from CHD (British Heart Foundation, 2006). In 2003 CHD caused approximately 114 000 deaths. The premature death rate from CHD for female manual workers was more than twice as high as that for female non-manual workers. South Asian women have a higher premature death from heart disease that is 51% greater than average (Petersen et al. Reference Petersen, Peto, Scarborough and Rayner2005). Data from the Scottish Executive (Bromley et al. Reference Bromley, Sproston and Shelton2005a) show that both IHD and stroke are greater amongst women from poorer areas compared with those from more affluent backgrounds.
Poorer women are at higher risk of many chronic conditions, including incidence, mortality and survival for lung cancer. Lung cancer is the second most common cancer in women in Scotland and has a clear social-class gradient (McLaren & Bain, Reference McLaren and Bain1998). Historical patterns of smoking in Scotland account for current trends in the incidence of lung cancer, which is now falling in men but not in women. However, there are also other factors that may increase the risk of lung cancer, including diet.
Causes of morbidity such as obesity follow a similar pattern and are notably higher in women compared with men and highest in poorest areas; in Scotland almost one-third of all adult women have a BMI of >30 kg/m2 (Bromley et al. Reference Bromley, Sproston and Shelton2005b). Dental caries, constipation and a range of other gastrointestinal symptoms (Bytzer et al. Reference Bytzer, Howell, Leemon, Young, Jones and Talley2001) related to diet and lifestyle are known to be more commonly reported amongst low-income households.
Thus, throughout adult life there are compelling reasons why dietary interventions might be targetted at low-income women to reduce total disease burden. In terms of nutrient intakes and food choices associated with the aetiology and progression of chronic diseases, the UK National Diet and Nutrition Survey (Hoare & Henderson, Reference Hoare and Henderson2004) highlights several areas of concern. One of the main nutritional concerns is actually in relation to lower reported energy intake in poorer households despite higher levels of obesity. In part this finding reflects some of the difficulties of obtaining valid and reliable data from adults who may have competing time demands and worries, fragmented irregular lives and highly-dependent family and friends around them. The dilemma of trying to disentangle under-reporting and true dietary intake will hopefully be progressed by the results of the Food Standards Agency (2006) Low Income Diet and Nutrition Survey, which are eagerly awaited later this year.
Returning to the UK National Diet and Nutrition Survey data published in 2003–4, the nutrient intakes of women living in households ‘in receipt of benefits’ when compared with those of households not in receipt of benefits show lower daily intakes of energy, protein, fat, alcohol and fibre (NSP). In addition, vitamin A, riboflavin, folate, vitamin C, Fe and Ca (and most micronutrients estimated) are also lower. For several nutrients >10% of this vulnerable group fail to achieve the lower reference nutrient intake. In terms of food choices, women living in households in receipt of benefits are more likely to report consumption of table sugar, whole milk, burgers and kebabs, meat pies and pastries and are less likely to have consumed wholegrain and other high-fibre breakfast cereals, oily fish, fruit juice, fruit and vegetables and low-energy carbonated beverages than their more affluent counterparts. Most notable of all these foods choices is the social patterning of vegetables, fruits and fruit juices, which has been echoed in a range of surveys including the UK National Diet and Nutrition Survey and smaller-scale work in Scotland. For example, data for North Glasgow from the Multinational Monitoring of Trends and Determinants in Cardiovascular Disease Project (Wrieden et al. Reference Wrieden, Connaghan, Morrison and Tunstall-Pedoe2004) show time trends in fruit and vegetable consumption in women, with the least affluent consuming least and the gap in intakes between the most affluent and the least affluent increasing progressively.
Clearly, women's own health is of fundamental importance. Not all women go on to become mothers (or minders of children) and the postmenopausal period can last several decades. However, it is clear that the greatest emphasis on nutrition interventions should be directed at women's reproductive and nurturing roles. There are, of course, sound reasons for this approach both in terms of focusing on a very vulnerable period when nutrient demands are high (e.g. pregnancy) and utilising opportunities to influence the health of the next generation (James et al. Reference James, Nelson, Ralph and Leather1997). Additionally, women are often seen as key agents for dietary change and gatekeepers of family health, with responsibilities for organising, attaining, preparing and delivering domestic food for small armies of disagreeable, and often rebellious, children and adolescents.
In terms of nutritional interventions in women most attention has been devoted to maternal, child and family health rather than focusing on adult health in women themselves. Birth weight is frequently used as an indicator of individual and population health and is known to be strongly correlated with adult CVD. Data from the Millenium Cohort (Mayhew & Bradshaw, Reference Mayhew and Bradshaw2005) show, however, that whilst poverty increases the odds of low birth weight by 61%, it is in fact the combined effects of unemployment, family type, educational level, young motherhood, being the firstborn and ethnicity that increase the odds of low birth weight. When all the factors are taken into account poverty no longer makes a separate contribution to low birth weight.
It is recognised that using the variable ‘low income’ in itself does not necessarily reflect the complex interaction of factors associated with living in poverty, including area effects. Indeed, it has become more common to use deprivation indices of areas of residence as a proxy for a range of relevant issues. For example, in terms of breast feeding in Scotland, current targets (Warren, Reference Warren2005) are set at >50% feeding at 6 weeks. Current rates for the population overall are 37·7%, but only 7·2% of mothers from the most deprived areas in the youngest age-group are still feeding at this important life stage (Bolling, Reference Bolling2006). Breast-feeding is associated with a reduced infection rate in early months, but also has longer-term health benefits, including a reduced risk of obesity in childhood and adolescence, and it is likely that this factor has a dose–response effect (Harder et al. Reference Harder, Bergmann, Kallischnigg and Plagemann2005). Similar findings have been observed for age of introduction of complementary foods (Wilson et al. Reference Wilson, Forsyth, Greene, Irvine, Hau and Howie1998). Work from the University of Dundee (Anderson et al. Reference Anderson, Guthrie, Alder, Forsyth, Howie and Williams2001; Alder et al. Reference Alder, Williams, Anderson, Forsyth, Florey and Van der Velde2004) has reported the importance of the deprivation category of the area of mother's residence and the influence of maternal mother and peers, as well as being in receipt of free samples of solid foods, as factors that influence early weaning. Even within the preschool stage when infants have been fully weaned the socially-patterned food choices and dietary intakes shown in adult surveys clearly emerge. Whole grains, fruit, vegetables and fish are consumed less frequently (Gregory et al. Reference Gregory, Collins, Davies, Hughes and Clarke1995) and correspondingly higher levels of teeth decay by deprivation category are seen (Newton et al. Reference Newton, Williams and Bower2005).
In summary, the woman as family's health gatekeeper must steer a course influenced by: life circumstances; available money; knowledge and skills about appropriate food choices; social pressure to conform to what friends, neighbours and family do and meet the culture norms of husbands and partners. In addition, actions will be influenced by neighbourhood effects of available stores, transport to larger supermarkets and the safety of their immediate area to support and facilitate physical activity and energy balance.
Intervention approaches in the UK and USA
To achieve reductions in chronic diet-related diseases the World Health Organization/Food and Agriculture Organization (2003) highlights the importance of both top–down (e.g. policy approaches) and bottom–up approaches, and local strategies and initiatives. Programmes and policy should be integrated and work hand-in-hand for maximum effect, employing a range of complementary actions (Stockley, Reference Stockley2001). Interventions to improve dietary changes can act by impacting on modifiable factors at an individual level, such as dietary knowledge, beliefs and attitudes and by improving psycho-social components (e.g. self efficacy). However, the long-term impacts of these dietary interventions will ultimately be enhanced and facilitated by societal interventions that tackle the context and situation of the living environment, and the balance between health promotion and food industry marketing. Dietary interventions cannot tackle the non-modifiable demographic characteristics such as socio-economic status of women, but available income will both influence and be influenced by dietary interventions.
Few interventions to reduce chronic diet-related diseases in low-income women in the UK have been systematically undertaken with dietary and health outcomes reported. Since low-income households tend to cluster geographically, a number of community-based programmes have been designed and implemented with the aim of reducing coronary risk factors. However, such programmes as ‘Good Hearted Glasgow’ (Hanlon, Reference Hanlon and William1992), ‘Have a Heart Paisley’ (Blamey et al. Reference Blamey, Ayana, Lawson, Mackinnon, Paterson and Judge2005) ‘Heart Beat Wales’ (Tudor Smith et al. Reference Tudor Smith, Nutbeam, Moore and Catford1998) and ‘Action Health’ (Baxter et al. Reference Baxter, Milner, Wilson, Leaf, Nicholl, Freeman and Cooper1997) rarely report specific results for low-income women or have been inadequately powered to show subgroup analysis. In some cases the response rates for adults living in the most-deprived communities have been particularly low, and more recently gaining ethical approval has added to the evaluators’ burden (Blamey et al. Reference Blamey, Ayana, Lawson, Mackinnon, Paterson and Judge2005). Such programmes have also raised a range of issues relating to the evaluation of community programmes, as illustrated by the Heart Beat Wales programme. This intervention used a coordinated range of activities, including public education campaigns along with supportive policy and infrastructure change. The resources that were developed included a BBC television series, food labelling and nutrition education with a major grocery retailer and Heart Beat restaurant awards. However, outcome measures (including the use of butter, fruit, vegetables, chicken, low-fat milk, wholemeal bread, smoking, frequency of exercise and body weight) in intervention and reference areas did not differ and it was concluded that these measures had been influenced by secular trends in lifestyle behaviours. Similarly, the analysis of results from independent samples in the USA (Minnesota and Pawtucket heart health programmes) has shown little difference in intervention effects between the test and control areas (Farquhar et al. Reference Farquhar, Fortmann, Flora, Taylor, Haskell and Williams1990), which were ‘modest in size and duration and generally within chance levels’ (Luepker et al. Reference Luepker, Murray, Jacobs, Mittelmark, Bracht and Carlaw1994) and ‘very limited’ (Carleton et al. Reference Carleton, Lasater, Assaf, Feldman and McKinlay1995). These findings raise a number of issues relating to evaluation.
Smaller intervention programmes specifically targetted at low-income groups in the UK include the Food Standards Agency ‘Cookwell’ trial, the aim of which was to develop a transferable-food-skills initiative that would increase consumption of fibre-rich starchy carbohydrates, fish, vegetables and fruit, and decrease consumption of fat in adults (including men) living in areas of deprivation. The intervention consisted of a 10-week programme of cooking skills classes in eight communities across Scotland, using a community development approach and informed by formative research with the target group (Stead et al. Reference Stead, Caraher, Wrieden, Longbottom, Valentine and Anderson2004). The study had a quasi-experimental design with an intervention group and a delayed intervention group. The effectiveness of the project was evaluated quantitatively: pre- and post-intervention measures of budget, diet and body weight; a questionnaire-based cooking skills assessment. Qualitative research was also included in the study design (focus groups and interviews; Wrieden et al. Reference Wrieden, Anderson, Longbottom, Valentine, Stead, Caraher, Lang, Gray and Dowler2007).
Of ninety-three participants, fifty completed food diaries at baseline and immediately post-intervention and forty of the fifty subjects completed a further 7 d food diary at 6 months post-intervention. There was an increase in fruit consumption in the intervention group (one extra portion of fruit per week) post-intervention, but this increase was not sustained after 6 months. Despite the small success in improving diet, the project coordinators also reported that the subjects’ confidence in following a recipe and cooking from basic ingredients increased (Wrieden et al. Reference Wrieden, Anderson, Longbottom, Valentine, Stead, Caraher, Lang, Gray and Dowler2007).
The limitations on the evaluation of community programmes add to the lack of evidence base to inform policy directives within the UK. A review published in the USA on cardiovascular health interventions in women (Krummel et al. Reference Krummel, Koffman, Bronner, Davis, Greenlund, Tessaro, Upson and Wilbur2001) suggests that ‘cardiovascular health interventions geared toward women are scant and intervention research to improve women's cardiovascular health is sorely needed’. The same authors recommend that community-based dietary interventions might usefully:
conduct qualitative research to determine the kinds of interventions women want;
examine relapse prevention, motivation and maintenance of behaviour change;
tailor programmes to the stage of the life cycle;
tailor programmes to a woman's readiness to change;
tailor programmes to subgroups, e.g. minority, low socio-economic and obese women;
evaluate policy and environmental interventions.
Women have been the focus of primarycare interventions, including cardiovascular risk, obesity and general healthy eating. Many individuals of all ages come into contact with health professionals on a regular basis, particularly women, who are more likely to be responsible for family planning requirements and attend for cervical screening. A wide range of cardiovascular preventions studies has been undertaken in primary care, including the British Family Heart Study (1994), but again few of these studies report findings by gender and socio-economic group. The approaches used in primary care vary, but they include one-to-one nutrition counselling, behavioural counselling, self-help materials and computer-tailored messages.
The focus of dietary change has included cardiovascular risk factors, obesity and fruit and vegetables intake. The European Prospective Investigation into Cancer and Nutrition Study (Khaw et al. Reference Khaw, Bingham, Welch, Luben, Wareham, Oakes and Day2001) has demonstrated that small increases in fruit and vegetable intake are likely to make a difference in cardiovascular mortality. One of the most notable trials that focused on low-income subjects (Steptoe et al. Reference Steptoe, Perkins-Porras, McKay, Rink, Hilton and Cappuccio2003) has demonstrated that behavioural counselling could increase fruit and vegetable consumption more than nutritional counselling (by 1·5 servings/d v. 0·9 servings/d respectively at 12 months from a baseline of 3·60 servings/d v. 3·67 servings/d). The subjects included men and women aged 18–70 years (n 271) recruited from a primary health centre in a deprived ethnically-mixed inner-city area in the UK who were randomly assigned to either behavioural or nutrition counselling. The behavioural counselling was based on social-learning theory and the stages-of-change model. The interventions included two 15 min consultations 2 weeks apart that were tailored to the individual, with personalised specific advice and setting of long- and short-term goals. Recruitment of the target group was, however, challenging.
A US review (Whitlock & Williams, Reference Whitlock and Williams2003) of evidence-based recommendations for physical activity promotion, dietary improvement and tobacco cessation from the US Preventive Services Task Force and the Task Force on Community Preventive Services (CTF) has found insufficient evidence in support of or against recommending behavioural counselling to promote a healthy diet in unselected patients in primarycare settings. The US Preventive Services Task Force recommends intensive behavioural dietary counselling (medium- to high-intensity counselling was defined as multiple contacts, generally more than six, lasting >30 min) by specialists (e.g. dietitians) for high-risk CVD patients, but found insufficient evidence to recommend routine healthy diet or physical activity promotion in primary care. At most, brief low- to medium-intensity (single or limited multiple contacts lasting <30 min) behavioural counselling in primary care can produce small to medium changes in self-reported dietary components. Whitlock & Williams (Reference Whitlock and Williams2003) also note that ‘the evidence base for these recommendations generally applies to women. Better reporting of gender and minority subgroup outcomes will assist more in-depth understanding of potential differences in either the processes or outcomes of behavior change interventions’.
In the USA two trials specifically aimed at women's health have recently been undertaken. WISEWOMAN (see Centers for Disease Control and Prevention, 2006) is a Centers for Disease Control and Prevention-funded programme targetted at women aged 40–64 years from low-income backgrounds (e.g. those with little or no health insurance), in which they are given access to screening and lifestyle interventions aimed at disease prevention. The WISEWOMAN Program was established in 1993 through the expansion of services offered through the National Breast and Cervical Cancer Early Detection Program to include interventions for chronic disease risk factors. In 1995 the Centers for Disease Control and Prevention launched WISEWOMAN demonstration projects in three states, Massachusetts, North Carolina and Arizona (see Centers for Disease Control and Prevention, 2006). These projects offered lifestyle interventions designed to change risk factors for chronic diseases, especially physical inactivity and diet. Specific interventions included structured counselling, physical activity classes and walking groups. A major emphasis was placed on smoking cessation.
Between January 2000 and June 2005 more than 45 000 women were screened for risk factors for heart disease and stroke, and about 119 000 lifestyle intervention sessions were provided. The women who enrolled in the programme during this period were at high risk for heart disease and stroke; 74% were overweight or obese, 27% were smokers, 24% had high blood pressure and 22% had high cholesterol levels. Reports (Rosamond et al. Reference Rosamond, Ammerman, Holliday, Tawney, Hunt, Keyserling, Will and Mokdad2000; Stoddard et al. Reference Stoddard, Palombo, Troped, Sorensen and Will2004) suggest that the reduction in CVD risk among WISEWOMAN participants after 1 year ranged from 6% in white women to 8·6% in black women. A review of the lessons learned from the WISEWOMAN project undertaken in the three US States (Viadro et al. Reference Viadro, Farris and Will2004) has concluded that ‘WISEWOMAN projects faced challenges of integrating clinical and lifestyle interventions, reaching beyond a focus on individuals, marshaling substantial resources, and introducing complex interventions into stretched healthcare environments. The three Phase One projects were deemed successful in reaching underserved women, developing a more comprehensive women's health model, strengthening linkages to primary healthcare, experimenting with innovative behavioural interventions, and tapping into women's roles as social support providers and family/community gatekeepers’. The overall effects that have been reported to date when comparing the minimal-intervention and enhanced-intervention groups are, however, modest.
The Women's Health Initiative (see Department of Health and Human Services, National Institutes of Health, National Heart, Lung, and Blood Institute, 2006) is the largest clinical investigation of strategies for the prevention and control of some of the most common causes of morbidity and mortality among postmenopausal women, including cancer, CVD and osteoporotic fractures. The Women's Health Initiative was initiated in 1992. Postmenopausal women aged 50–79 years were enrolled at one of forty Women's Health Initiative clinical centres across the USA into either a clinical trial (64 500 women) or an observational study (100 000 women). The clinical trial was designed to allow evaluation of three distinct interventions: a low-fat eating pattern (in which only 20% energy were to be consumed from fat and fruit, vegetable and grain consumption was high), hormone-replacement therapy and Ca and vitamin D supplementation. The average follow-up for women in both the clinical trial and the observational study was approximately 8 years.
Participants had an average intake of 35% energy from fat when the trial began in 1993, which decreased to an average intake of 24% energy from fat by the end of the first year and by the end of the study the level had stabilised at 29% energy from fat (Howard et al. Reference Howard, Van Horn, Hsia, Manson, Stefanick and Wassertheil-Smoller2006). Fruit and vegetable intake increased but not consumption of grains. At the end of the study the difference in fat intake between the low-fat-diet group and the usual-diet group was 8% instead of the projected 11%.
Early reports from the Women's Health Initiative for breast cancer (Prentice et al. Reference Prentice, Caan, Chlebowski, Patterson, Kuller and Ockene2006), colo-rectal cancer (Beresford et al. Reference Beresford, Johnson, Ritenbaugh, Lasser, Snetselaar and Black2006) and CVD (Howard et al. Reference Howard, Van Horn, Hsia, Manson, Stefanick and Wassertheil-Smoller2006) show no reduction in disease by dietary intervention. However, in the subgroup analyses breast cancer rates were reduced by 22% among women who started with the highest fat intake (>36·8% energy from fat) and reduced their fat the most (to 23·6% after 1 year; Stein, Reference Stein2006). In addition, analysis by food records v. FFQ shows a positive association between fat and breast cancer risk (Freedman et al. Reference Freedman, Potischman, Kipnis, Midthune, Schatzkin, Thompson, Troiano, Prentice, Patterson, Carroll and Subar2006). For CVD there were small improvements in body weight, LDL cholesterol, diastolic blood pressure and Factor VIIC (Howard et al. Reference Howard, Van Horn, Hsia, Manson, Stefanick and Wassertheil-Smoller2006). Furthermore, HDL, triacylglycerols, glucose and insulin levels did not increase. Despite the lack of apparent effect on colo-rectal cancer, adenomas were reduced among the low-fat-diet group compared with the usual-diet group (Beresford et al. Reference Beresford, Johnson, Ritenbaugh, Lasser, Snetselaar and Black2006). The analysis of results by income status has not yet been published.
Implications of evidence base for policy and practice
Interventions to address diet inequalities in women have tended to focus on the most vulnerable periods of the female life cycle, i.e. during pregnancy and lactation when there are specific opportunities within healthcare settings to provide educational and other interventions. The nutrient-dense diet recommended for optimal nutrition appears to provide a good opportunity for affecting lifetime health, but many women may be unable to attain this diet because it is not affordable or accessible. Acheson (Reference Acheson1998) has recommended that ‘further reductions in poverty in women of childbearing age, expectant mothers, young mothers, young children and older people should be made by increasing benefits in cash and kind to them’. However, dietary interventions aimed at increasing access, acceptability (within norms of current cultural eating habits) and awareness of appropriate food choices are also required.
In the UK there have been ranges of policy-driven initiatives directed at reducing health and nutrition inequalities in mothers and young children. These initiatives have included The Welfare Food System, now ‘Starting Well’, which has committed to ‘using the resource more effectively to ensure children in poverty have access to a healthy diet and increased support for breast feeding and parenting’ (Scientific Advisory Committee on Nutrition, 2002). In addition, a wide range of national schemes driven through local initiatives have operated in England (through activities within Sure Start, Health Action Zones, Healthy Living Centres and local action to deliver National Service Frameworks) and within Scotland (e.g. Health Improvement Plan, Scottish Community Diet Project) and Wales (Community Food Initiative). In addition, schemes such as the national School Fruit Scheme in England, breakfast clubs and school meal provision have supposedly contributed to improvements in diet in children from socially disadvantaged areas. There are, however, three important issues to consider here. The first is that current policy work lacks robust evidence that such initiatives have changed dietary intakes and subsequent health. Second, an analysis of how successful such initiatives might be in an environment that promotes excess consumption has never been undertaken. Third, the target of dietary improvement has commonly been to improve the health of children, with women often acting as the conduit for children's health, rather than the woman's health.
In the UK there has been a long history of dietary interventions (mainly counselling) aimed at pregnant women, which indicates that nutrient intake can be improved during pregnancy, but there is no robust evidence that nutrition counselling impacts on rates of low birth weight, gestational age or length of birth. There have been very few evaluations undertaken on practical interventions aimed at improving access to healthy diets, improving food affordability and availability (e.g. practical food skills classes. One published study from University of Dundee (Wrieden & Symon, Reference Wrieden and Symon2003) has evaluated the feasibility of a cooking skills programme led by midwives in a community setting for teenage pregnant women. The programme incorporated seven informal food preparation sessions and opportunities for discussion of food and health matters (including food safety and well-being in pregnancy). Whilst the midwives found the package easy to follow and use, only sixteen (of the 120 invited) women attended the course and the authors concluded that alternative methods of delivering and evaluating such a package should be investigated. Further cooking skills interventions are currently being funded by the Food Standards Agency, including studies of dietary interventions aimed at improving the diets of girls and women at risk of having low-birth-weight babies.
A major focus of health interventions has been to promote breast-feeding amongst low-income women. The merits of breast-feeding for the baby have been discussed, but it should be noted that lactation is also related to a reduced likelihood of maternal breast cancer. Considerable progress has been made in the arena of increasing breast-feeding but much more remains to be done. The conclusions of a recent Cochrane review (Dyson et al. Reference Dyson, McCormick and Renfrew2006) reports that the forms of breast-feeding interventions (health education, breast-feeding promotion packs, early infant–mother bonding) evaluated for systematic review were effective at increasing breast-feeding initiation rates among women on low incomes in the USA. In the review seven trials involving 1388 women were included. Five trials involving 582 US women on low incomes show that breast-feeding education has a significant effect on increasing initiation rates compared with routine care (relative risk 1·53; 95% CI 1·25, 1·88).
The improvement of nutrition and health is a major aim of the American Special Supplemental Food Program for Women, Infants and Children (see US Department of Agriculture Food and Nutrition Service, 2006), which is directed at low-income nutritionally-at-risk pregnant women and preschool children. The scheme provides supplemental food, nutrition education and coordination of health care (Rush et al. Reference Rush, Horvitz, Seaver, Aluir, Garbowski, Leighton, Sloan, Johnson, Kulka and Shanklin1988). The funds provided for the American Special Supplemental Food Program for Women, Infants and Children are divided between supplemental foods (75% of funds) and nutrition education (one-sixth of the administration funds), with food vouchers provided for Fe-fortified formula, cereal and (vitamin C-enriched) fruit juice for babies and infants and specific foods for children and pregnant and lactating women (e.g. milk, citrus juice). In addition, breast-feeding promotion and peer-counsellors have become an integral part of the programme. The impact of the programme has been demonstrated to increase maternal weight gain in pregnancy, increase intakes of energy and nutrient density, increase Hb and reduce anaemia, increase breast-feeding rates (Fairbank et al. Reference Fairbank, O'Meara, Renfrew, Woolridge, Sowden and Lister-Sharp2000) and decrease childhood anaemia (Yip et al. Reference Yip, Binkin, Fleshwood and Trowbridge1987).
Another US programme that has demonstrated the achievement of changes in dietary awareness and empowering participants to change dietary practices (Chipman & Kendall, Reference Chipman and Kendall1989) is the Expanded Food and Nutrition Education Program (see US Department of Agriculture, 2001). This programme is a federally-funded nutrition programme aimed at assisting low-income youth and families (with young children) and ethnic minorities to acquire practical food knowledge, skills, attitudes and behaviour change (including money management and getting the most from health assistance programmes) necessary to help achieve nutritionally-sound diets and the improvement of the total family diet and nutritional well-being. This programme is delivered through a series of lessons by paraprofessionals and volunteers adopting a hands-on, ‘learn by doing’ approach, often by individually-tailored home education sessions (Bowering et al. Reference Bowering, Morrison, Lowenburg and Tirado1976). The programme has been shown to influence a range of food practices (including food budgeting, food safety and food preparation; Arnold & Sobal, Reference Arnold and Sobal2000). Cost–benefit analysis suggests that for every US $1 invested in the programme, between US $2·48 (Tennessee) and US $10·64 (Virginia) in benefits from reduced healthcare costs can be expected (Rajopal et al. Reference Rajopal, Cox, Lambur and Lewis2002).
In the UK the health and well-being of millions of women are influenced by living in poverty. Food choices, dietary intake and feeding behaviour are far from optimal amongst poorer women, and this situation is reflected in higher rates of diet-related morbidity and mortality well beyond maternal and child health considerations. Interventions to change diet need to take account of the factors that influence household economic status and should be integrated at policy, community and individual level.
There is a dearth of research on effective interventions in this subgroup of the population from within the UK, although there are some encouraging US examples to draw on in terms of possible intervention approaches.
Lack of evidence does not mean that that policy work should be delayed. It is recognised that engaging with ‘hard to reach’ groups of women is a challenge for intervention implementation and evaluation, and it is essential that policy work should be evaluated for its ability to engage with target groups as well as for behavioural change and health outcomes.