Breakfast represents the first opportunity in the day to consume foods to provide sustenance to perform daily activities. Evidence indicates that breakfast skipping may be an indicator of poor dietary habits and behaviours in general(Reference Keski-Rahkonen, Kaprio and Rissanen1–Reference Horikawa, Kodama and Yachi4). Beyond the associations with unhealthy dietary patterns(Reference Howarth, Huang and Roberts5) and dietary-related disease(Reference Takagi, Hari and Nakashima6), breakfast skipping is a concern as it is known to impair daily functioning, reducing concentration and cognitive performance(Reference Martens, van Assema and Brug7), decreasing energy and negatively affecting mood(Reference Akeredolu, Okafor and Mbah8,Reference Boschloo, Ouwehand and Dekker9) , which can compound any problems children and adults experience during their school and work commitments. Breakfast skipping is a known risk factor for health, having been positively associated with non-communicable diseases(Reference Smith, Breslin and McNaughton3,Reference Szajewska and Ruszczyński10) and overweight and obesity(Reference Chowdhury, Richardson and Holman11–Reference Aanesen, Katzmarzyk and Ernstsen14). Furthermore, establishing breakfast habits when young can lead to lifelong habits including healthy eating behaviours(Reference Smith, Breslin and McNaughton3,Reference Martens, van Assema and Brug7,Reference Szajewska and Ruszczyński10) .
In Australia, 19 % of adolescents and 12 % of adults do not consume breakfast indicating that as many as one in ten adults is not meeting ideal breakfast recommendations(Reference Fayet-Moore, McConnell and Kim15,Reference Fayet-Moore, McConnell and Cassettari16) . Adolescents and young adults commonly engage in unhealthy habits such as breakfast skipping, certainly in western countries(Reference Smith, Breslin and McNaughton3,Reference Martens, van Assema and Brug7,Reference Nelson, Story and Larson17,Reference Dykstra, Davey and Fisher18) . Recommendations for an ideal breakfast propose that it should provide 15 %–25 % of total daily energy(Reference O’Neil, Byrd-Bredbenner and Hayes19). In addition to providing energy and nutrients to commence the day’s activities, research demonstrates that young people who eat breakfast demonstrate better food behaviours throughout the day(Reference Leidy, Armstrong and Tang20).
There is a critical time when adolescents transition to adulthood that it is thought to be the catalyst for unhealthy eating and behaviour(Reference Nelson, Story and Larson17). This period is important for developing healthy eating behaviours due to the development of autonomy for emerging adults(Reference Nelson21,Reference Fayet-Moore, Kim and Sritharan22) . As adolescents move into adulthood, not only do eating issues arise but physical activity also slows which declines further as the adult moves through their lifespan. Over 33·6 % of Australian adolescents do not meet physical activity guidelines(Reference Nelson, Story and Larson17). As adolescence marks a critical time when behavioural changes occur, including eating behaviours, change from adolescence to adulthood is an optimal time to encourage healthy eating to allow new habits and behaviours to form(Reference Koehn, Gillison and Standage23,Reference Fielding-Singh24) , including healthy breakfast habits that can then endure across the lifespan.
The reasons why adolescents and young adults do not eat breakfast have included a lack of self-efficacy, perceived barriers, other established behavioural patterns (such as late nights and disorganised mornings) and conflicts between competing demands and preferences(Reference Dehdari, Rahimi and Aryaeian25). Other studies have found absence of hunger, low enjoyment of breakfast and lack of time to contribute, and use breakfast skipping as a weight loss strategy(Reference Hearst, Shanafelt and Wang26,Reference Mullan, Wong and Kothe27) .
Given the detrimental effects breakfast skipping has on daily functioning and long-term health, there is a need for interventions or programmes to improve breakfast consumption. Furthermore, the transitional stages of adolescence to adulthood suggest that behaviour change strategies need to be cognisant of the changing individual, social and environmental circumstances(Reference Stok, de Vet and de Ridder28,Reference Arnett29) . Social marketing is a widely recognised behaviour change discipline(Reference Rundle-Thiele30) that integrates knowledge from psychology, sociology, anthropology and more to help build programmes that can change people’s behaviours to benefit themselves and the society in which they live, work and play(Reference Gordon, McDermott and Stead31). The effectiveness of social marketing in the area of healthy eating has been demonstrated(Reference Stead, Gordon and Angus32,Reference Carins and Rundle-Thiele33) , and the role of the environment on eating decisions is widely recognised(Reference Swinburn, Sacks and Hall34). Breakfast, as a specific healthy eating behaviour, has received less attention in the social marketing field(Reference Askelson, Golembiewski and DePriest35); therefore, opportunities exist to improve breakfast consumption using this approach.
Theories guide practitioners during the development of programmes ensuring that strategies are based on previous established knowledge that helps to explain relationships that are likely to affect programme outcomes and provide the basis for comprehensive evaluation of programme effectiveness(Reference Dehdari, Rahimi and Aryaeian25). However, low utilisation or underreporting of theory has been observed in previous social marketing interventions(Reference Luca and Suggs36,Reference Truong37) , including poor theory description and measurement(Reference Pang, Kubacki and Rundle-Thiele38,Reference Willmott, Pang and Rundle-Thiele39) , stating that deepening and broadening theory use in social marketing remain a key area for improvement. Given social marketing is known to underutilise theory(Reference Rundle-Thiele, David and Willmott40), the development of breakfast behaviour change programmes using social marketing should be theoretically informed. Furthermore, the nature of breakfast consumption behaviour requires utilisation of a theory that acknowledges the role that social and environmental factors exert on food decisions, such as Social Cognitive Theory (SCT). SCT is a well-established theoretical framework, in both nutrition interventions and social marketing(Reference Truong37), and may be an important starting point for extending understanding the role theory can play in delivering behaviour change, particularly in regard to breakfast consumption.
This paper aims to understand how programmes have attempted to improve breakfast consumption, and whether theory contributed to programme success, thereby responding to the call to advance understanding of theory application in behaviour change programmes(Reference Rundle-Thiele, David and Willmott40). Systematic literature reviews are a scientific tool to understand and examine a particular issue or topic in a methodical way to synthesise evidence(Reference Petticrew and Roberts41). Therefore, the aim of this study was to conduct a systematic literature review of the last 20 years of studies on breakfast eating programmes to determine programme effectiveness, analyse their programme features and their use of theory. SCT was used as a lens to examine theory use, to enable synthesis and to compare constructs across studies.
Methods
Search strategy
This study undertook a systematic literature review to examine interventions that aimed to change breakfast behaviour. The search was performed using eleven databases, namely Medline, Psych INFO, Inspec, NTIS, Web of Science, ProQuest G1 & G2, EBSCO, Emerald and ScienceDirect. The search protocol used the following key terms, grouped into two sets of terms: ‘breakfast+eating’ OR ‘breakfast+skipping’ OR ‘breakfast+consumption’ AND ‘intervention*’ OR ‘Randomi#ed+Controlled+Trial’ OR ‘evaluation’ OR ‘trial’ OR ‘campaign*’ OR ‘program*’ OR ‘experiment’ OR ‘study’ OR ‘studies’. Results were limited to those published after the year 2000 to capture the past 20 years of research on this topic.
Exclusion criteria
All results from the database searches were combined and ordered, and duplicates were removed. After the initial ordering of papers, each title and abstract were reviewed by two independent reviewers to ensure that all relevant interventions were included. The following exclusion criteria were applied: (i) papers that were disease-related or not nutrition-focused, (ii) papers not in English and (iii) reviews or conceptual papers. Full articles for any remaining records were examined by two independent reviewers, and articles were only retained if the intervention aimed to change breakfast behaviour. Backward and forward searching was conducted to identify any further papers associated with the identified studies. These related papers were used to obtain further study details not reported in the papers sourced during the initial search process. Screening of each record during the exclusion and inclusion process was conducted independently, and all differences were resolved through discussion between two researchers.
Study approach
The following data were extracted from each finally selected study: details of the study (author, year of publication and country), research/evaluation design, theory used (if reported), participants’ (sample size, characteristics), details of the intervention, constructs or measures used in the study, duration of the study and study findings. Given the aims of this study included the effectiveness of the programme and examination of theory use, the authors also extracted data, or coded data extracted above to examine (i) theory, did the paper discuss the type of theory used and was their theory explanation weak, moderate or strong; (ii) behavioural focus (breakfast or healthy eating), whether the study targeted a single (breakfast) behaviour or multiple behaviours (including breakfast); (iii) constructs, did the paper report constructs or measures involved in changing behaviour, or measuring the change in behaviour that was equivalent to Social Cognitive Theories three main constructs of behavioural, cognitive or environmental; (iv) effectiveness of the programme, on breakfast consumption frequency; (v) effectiveness of the programme for dietary quality and (vi) other changes observed. The approach to coding was as follows. Theory was coded as none, weak, moderate or strong through study design, measures, providing key points on how data collection was obtained and analysis that could be reproduced. Studies were coded as having included SCT constructs (even if they did not mention SCT factors within their interventions) if equivalent behavioural, personal and environmental intervention strategies were included. Behavioural constructs were skills, practice and self-efficacy; personal factors were knowledge, expectations and attitudes, and environmental constructs were measured in social norms, access in community and influence on others (ability to change their own environment). Two authors thoroughly reviewed the studies and resulting in summary tables to ensure accuracy and relevance. Variation in outcome measures was expected, which limits the capacity to perform meta-analysis without substantial data transformation and assumptions. Therefore, this study conducted a critical narrative synthesis of interventions, focusing on the behavioural outcomes sought, theory use, the reporting of strategies targeting socio-cognitive constructs and the effectiveness of interventions.
Quality assessment
The National Health and Medical Research Council(Reference Merlin, Weston and Tooher42) quality assessment framework was used to assess the level of evidence provided by each study in support of intervention efficacy with ratings from I (highest) to IV (lowest). Each intervention was assessed and rated using this framework.
Results
The search produced 1052 records. After duplicates were removed, and the inclusion and exclusion process was applied, nineteen studies remained, reported in twenty-one papers. A PRISMA flow chart details the search process in Fig. 1.
Fig. 1 Prisma flow diagram
The majority of studies (n 14) were school-based and involved primary schoolchildren (n 6), middle school (n 2), high school (n 5) or University students (n 1). The remaining studies were educational classes held outside schools (n 2), a free breakfast programme (n 1) and nationwide interventions (n 2). The studies were conducted in the USA (n 6), Canada (n 1), Australia (n 5), Iran (n 1), Turkey (n 1), Israel (n 1), Gom City (n 1) and Europe (n 3). The length of the studies varied greatly, from brief interventions conducted over 48-h durations to those programmes that lasted for a full year. The papers reporting the studies were published between 2005 and 2018, and these are summarised in Table 1.
Table 1 Studies included in the systematic literature review
Assessment quality
The quality assessment rating for National Health and Medical Research Council shown above in Table 1 has rated ten studies at level II, two studies level III-2 and 1 at level III-3 and five studies level IV. There was one case study that was not applicable as it only provided breakfast to schools and could not be assigned a level of evidence under this framework. Most interventions (n 10) where randomised control studies but had minimal strategies to prevent bias were reported, and the study design was described in full. Many articles did not provide data collection methods to the full extent to be reproduced nor did they provide evidence that the collection tools were reliable or valid. Thereby, most interventions were rated as weak according to National Health and Medical Research Council guiding frameworks. In future research, designs need to fully report programme design, methods and results for reproducibility.
Behavioural focus
The behavioural focus of the programmes varied. Eight studies focused on healthy eating as a broad concept and included breakfast eating as one aim in the programme. Each of these programmes was educative in nature, most adopting a lesson-based format(Reference Bessems, van Assema and Martens43–Reference Karadağ, Seviğ and Kılıç48), although one focused on practical activities(Reference Hossain, Yuginovich and Lambden49). The other eleven studies focused on changing breakfast behaviour alone. Eight of these were educative or informative in nature, focusing on increasing understanding of the importance of breakfast, the health consequences of skipping breakfast, which foods are healthy breakfast foods, and encouraging healthy breakfast habits(Reference Au, Whaley and Rosen50–Reference Watson, Kwon and Nichols57). One of these programmes combined a mass media campaign within school education to reinforce the importance of breakfast(Reference O’Dea and Wagstaff54). Another intervention tested messaging to improve attitudes towards breakfast(Reference Kothe, Mullan and Amaratunga55). The remaining four breakfast-focused interventions made changes to breakfast programmes, by introducing free breakfasts in schools(Reference Crepinsek, Singh and Bernstein58,Reference Murphy, Moore and Tapper59) , changing the frequency of school breakfast provision(Reference Leatherdale, Stefanczyk and Kirkpatrick60) or altering the way school breakfast was provided(Reference Olsta61). Most interventions included strategies within the intervention targeting personal factors, such as knowledge and self-efficacy; however, these appeared to be used less frequently in breakfast-only interventions (4 without v. 7 with personal strategies) compared with healthy eating interventions (0 without v. 8 with personal strategies).
Theoretical focus
Theory was rarely reported in the studies included in this review. Of the nineteen programmes examined, only six reported theory. The theories reported were SCT (two studies), Health Promotion Model (two studies) and Theory of Planned Behaviour (two studies). Three articles only made mention of the theory, two mentioning SCT(Reference Fairclough, Hackett and Davies45,Reference Foley, Shrewsbury and Hardy46) and one mentioning Health Promotion Model(Reference Olsta61). Two studies could be considered as reporting weak application of theory, in that they described the behaviour in terms of the theory. Those were an application of Health Promotion Model(Reference Dehdari, Rahimi and Aryaeian51) and an application of Theory of Planned Behaviour(Reference Gharlipour, Ghaffari and Hoseini53). One article reported strong application of theory, having designed the intervention and measured programme performance using key Theory of Planned Behaviour constructs(Reference Kothe, Mullan and Amaratunga55).
Despite not showing evidence of strong theoretical application, the two studies that self-nominated SCT as a theoretical basis did provide intervention descriptions showing evidence of strategies targeting each SCT construct (personal, behavioural, environmental). However, these interventions did not provide clear links to SCT constructs such as observational learning; goal setting; feedback and specific environmental facilitators. The failure to more comprehensively link theory to programme implementation was not unique to these interventions. Other studies not mentioning theories reported some SCT constructs with the most commonly reported including personal strategies such as knowledge from educative programmes, behavioural constructs (e.g. self-efficacy and practice) and environmental constructs of decreasing barriers to breakfast consumption and social influences from peers, teachers and parents (see Table 2).
Table 2 Effectiveness of behavioural change interventions using Social Cognitive Theory
Studies reporting theory more often targeted multiple behavioural constructs. One study that reported using theory targeted a single construct, whereas the other five targeted multiple constructs. Studies that did not report using theory tended to target fewer constructs – nine of these studies targeted a single construct, whereas four targeted multiple constructs. Studies reporting theory use also targeted behavioural constructs more often (one not targeting behavioural v. five targeting behavioural constructs) compared with studies not reporting theory use (ten not targeting behavioural v. three targeting behavioural constructs).
Intervention effectiveness
Positive change in the frequency of breakfast consumption was observed in eight of the nineteen studies(Reference Evans, Massey-Stokes and Denson44,Reference Karadağ, Seviğ and Kılıç48–Reference Dehdari, Rahimi and Aryaeian51,Reference O’Dea and Wagstaff54,Reference Watson, Kwon and Nichols57,Reference Olsta61) . A further two studies observed positive effects for subgroups, such as increases in frequency of breakfast consumption in only those from high socio-economic backgrounds(Reference Fairclough, Hackett and Davies45), and increased frequency for females only(Reference Heo, Irvin and Ostrovsky47). No change was observed in seven studies(Reference Bessems, van Assema and Martens43,Reference Foley, Shrewsbury and Hardy46,Reference Eilat-Adar, Koren-Morag and Siman-Tov52,Reference Kothe, Mullan and Amaratunga55,Reference Radcliffe, Ogden and Welsh56,Reference Crepinsek, Singh and Bernstein58,Reference Murphy, Moore and Tapper59) , one was inconclusive(Reference Leatherdale, Stefanczyk and Kirkpatrick60) and one study observed a decrease in breakfast frequency(Reference Gharlipour, Ghaffari and Hoseini53). There were five studies that did not observe an increase in frequency of breakfast consumption but positive changes in the dietary quality of breakfast were observed(Reference Bessems, van Assema and Martens43,Reference Eilat-Adar, Koren-Morag and Siman-Tov52,Reference Radcliffe, Ogden and Welsh56,Reference Crepinsek, Singh and Bernstein58,Reference Murphy, Moore and Tapper59) . Three studies found that healthier food options were being consumed even though breakfast frequency did not increase(Reference Eilat-Adar, Koren-Morag and Siman-Tov52,Reference Radcliffe, Ogden and Welsh56,Reference Murphy, Moore and Tapper59) , one found more substantial breakfasts were being consumed(Reference Crepinsek, Singh and Bernstein58) and one found an increase in fruit consumption(Reference Bessems, van Assema and Martens43). Thirteen studies measured changes in psychological constructs, such as knowledge, attitudes or perceptions. Of these, six observed positive changes to psychological constructs paired with increases in breakfast consumption(Reference Karadağ, Seviğ and Kılıç48–Reference Dehdari, Rahimi and Aryaeian51,Reference O’Dea and Wagstaff54,Reference Watson, Kwon and Nichols57) , six saw positive changes to psychological constructs without resulting changes in behaviour (Reference Foley, Shrewsbury and Hardy46,Reference Heo, Irvin and Ostrovsky47,Reference Eilat-Adar, Koren-Morag and Siman-Tov52,Reference Gharlipour, Ghaffari and Hoseini53,Reference Radcliffe, Ogden and Welsh56,Reference Murphy, Moore and Tapper59) and one saw no changes in either behaviour or psychological constructs(Reference Kothe, Mullan and Amaratunga55). There appeared to be no pattern between intervention effectiveness (increased breakfast consumption frequency) and a singular breakfast behavioural focus (four effective, seven not effective) v. a broader healthy eating behavioural focus (four effective, four not effective). No pattern was observed between effectiveness and reported theory use (two effective, four not effective) v. no theory use (six effective, seven not effective); nor between effectiveness and strategies targeting a single construct (four effective, six not effective) v. interventions targeting multiple constructs (four effective, five not effective).
Discussion
Breakfast consumption is a protective factor against obesity and overweight(Reference Ma, Chen and Pu62), and food consumption early in the day is required to support physical and cognitive performance(Reference Martens, van Assema and Brug7,Reference Clayton, Barutcu and Machin63) ; however, many adolescents, younger and older adults do not consume breakfast. This indicates the need for behaviour change programmes to increase breakfast consumption. This study aimed to analyse previous breakfast programmes to determine the effect of those programmes on breakfast consumption. This study also sought to evaluate the extent of theory use within the programmes and to examine the features of the programmes to inform future programme development.
This review found mixed evidence of the effectiveness of interventions in increasing the frequency of breakfast consumption. Some studies failed to find a change to frequency, but found positive changes to the foods consumed, or to psychological measures that may, with a longer duration or more exposure result in behaviour change. Few patterns emerged that might explain the mixed success. The variation in intervention focus, intervention breadth and study design may have obscured any patterns.
The majority of studies in this review (eleven studies) focused on breakfast eating alone, whereas the remaining eight targeted healthy eating with breakfast eating as one component. Recommendations for intervention design often call for ‘narrowing in’ on a discrete behaviour, to more easily describe the steps individuals need to take and to facilitate measurement before and after intervention(Reference Atkins and Michie64). There is some evidence that interventions focusing on a single behaviour are more effective that those focusing on multiple behaviours(Reference Sweet and Fortier65); however, this can oversimplify the complex interaction between influences particularly in area of food consumption(Reference Parkinson, Dubelaar and Carins66) and confine intervention design to individually based strategies resulting in costly approaches. This review found no difference in effectiveness between interventions targeting singular v. multiple behaviours in healthy eating and breakfast interventions. Research carried out by Prochaska et al. (Reference Prochaska, Spring and Nigg67) also revealed inconsistent findings on whether interventions targeting multiple or singular behaviours work best. There is some evidence that associates multiple unhealthy eating behaviours together. Driskell et al. (Reference Driskell, Dyment and Mauriello68) and Kremers et al. (Reference Kremers, De Bruijn and Schaalma69) found that low fruit and vegetable intake was associated with lower physical activity, and Keski-Rahkonen et al. (Reference Keski-Rahkonen, Kaprio and Rissanen1) and Cohen et al. (Reference Cohen, Evers and Manske70) noted that breakfast skipping had a link to lower physical activity. This suggests that the effectiveness of healthy eating interventions may rely on targeting behaviours together in order to get behaviour change. Even so, breakfast-only interventions were no more or less effective in increasing breakfast frequency than the broader healthy eating interventions, indicating that more research is warranted. This review adds to the existing literature that breakfast eating behaviours changed just as much when singularly looking at that behaviour compared with a multiple healthy eating behaviour change intervention.
In this review, many interventions were education programmes delivered in school settings with a focus on individual factors such as knowledge and self-efficacy. This indicates a dominant focus on changing what people think, which overlooks a person’s ability and opportunity to perform the targeted behaviour. The studies targeting healthy eating in general frequently focused on delivering strategies to individuals, whereas programmes targeting breakfast often included behavioural or environmental strategies, such as student educative strategies(Reference Dehdari, Rahimi and Aryaeian51,Reference O’Dea and Wagstaff54,Reference Radcliffe, Ogden and Welsh56) , environmental strategies including parents and teachers(Reference Dehdari, Rahimi and Aryaeian51) and the wider community(Reference O’Dea and Wagstaff54,Reference Radcliffe, Ogden and Welsh56) . However, it is important to note the dominance of school-based interventions in this review, even though we did not restrict the search by age or setting, means few programmes are offering support to adolescents and emerging adults who are transitioning to adulthood, and as a result, are experiencing changes in behavioural routines and social and environmental influences. A broader focus on behavioural and environmental strategies is aligned to SCT(Reference Bandura71).
SCT is the most widely used social marketing theory (see Truong(Reference Truong37)) and was a focal theory of interest chosen to guide this review, which aimed to inform the development of a theoretically informed programme aimed at increasing breakfast eating. Two articles(Reference Fairclough, Hackett and Davies45,Reference Foley, Shrewsbury and Hardy46) used SCT, and a further two articles reported strategies targeting all three SCT constructs(Reference Dehdari, Rahimi and Aryaeian51,Reference Gharlipour, Ghaffari and Hoseini53) . Given theory is often not clearly reported in articles reporting inventions, and some theoretical constructs are common to more than one theory; this review sought to identify the presence of SCT constructs within intervention strategies even if the intervention authors did not claim SCT was the foundation for their work. SCT construct strategies most commonly identified were personal strategies (e.g. classroom interventions focusing on knowledge), followed by environmental strategies that sought to decrease barriers by providing breakfast and increasing social support through educating family and peers on the importance of breakfast. Behavioural strategies were also evident, and these sought to change outcome expectancies. Reported theory use was low, but strategies targeting SCT constructs were commonly used in many interventions. Prior research(Reference Young, Plotnikoff and Collins72,Reference Adhikari, Puri and Thapa73) has shown the use of one or more SCT constructs in healthy eating interventions is mostly effective(Reference Young, Plotnikoff and Collins72,Reference Adhikari, Puri and Thapa73) . Moreover, a review on the effectiveness of two or more healthy behavioural approaches saw a small but significant benefit when controlling both diet and physical activity compared with one behavioural approach(Reference Dombrowski, Knittle and Avenell74). SCT has been noted as a valuable theory to effectively change behaviour in healthy eating and breakfast interventions(Reference Zacarías, Shamah-Levy and Elton-Puente75), suggesting detailed application and testing are warranted. However, in this review, interventions containing strategies based on SCT equivalent constructs appeared no more or less likely to result in behaviour change.
Theory is important, providing valuable frameworks for the development and design of interventions to solve problems. Three different theories were mentioned or discussed by papers in this review – SCT(Reference Fairclough, Hackett and Davies45,Reference Foley, Shrewsbury and Hardy46) , Health Promotion Model(Reference Dehdari, Rahimi and Aryaeian51,Reference Olsta61) and Theory of Planned Behaviour(Reference Gharlipour, Ghaffari and Hoseini53,Reference Kothe and Mullan76) . In this review, studies reporting use of a theory (any theory) generally produced interventions targeting multiple constructs and were more likely to include strategies targeting the behavioural construct – linking the behaviour with a desired outcome. However, use of a theory (whether weak or strong) was not associated with effectiveness in these groups of studies, and neither was targeting all three SCT constructs. In this review, studies reporting theory use were just as likely to be effective as those not reporting use of theory. Furthermore, even when interventions were coded as using theoretical constructs (specifically SCT constructs), there appeared to be no link between the inclusion of theoretical constructs and effectiveness.
Transparent and clear reporting of theory can assist the research community to build robust and technical studies to evaluate and compare interventions(Reference Ward77). In health interventions, the use of theory helps researchers and practitioners to understand whether a behaviour change has occurred and to identify whether changes observed are a result of ‘attitudinal, normative, self-efficacy, environmental or social’ factors(Reference Luca and Suggs36). Theory is known to improve the effectiveness of interventions, but theory application is still under reported or not used at all to design and/or evaluate interventions(Reference Luca and Suggs36). Weak theory use is still commonly seen in social change programmes(Reference Rundle-Thiele, David and Willmott40,Reference Pang, Kubacki and Rundle-Thiele78–Reference Prestwich, Webb and Conner80) , and this was supported in the current review. Limited levels of theory reporting do not assist the research and practice community to understand what works, when, where and why it worked(Reference Rundle-Thiele, David and Willmott40). More rigour in theory application is needed. This is vital for the growth of behaviour change programmes and behavioural change disciplines(Reference Truong37).
This review examined interventions that aimed to improve breakfast consumption, from many countries around the globe. Some successful interventions included environmental changes (in isolation or combined with other strategies), suggesting a role for policymakers to create supportive environments on a broader scale, through legislation. Other interventions included strategies targeting behavioural or personal dimensions – those that are more commonly the focus for public health and social marketing researchers and practitioners. Despite mixed results, the findings should encourage those working in public health, social marketing and behavioural change more broadly to diligently consider which behaviours they are aiming to change, and which theoretical constructs or pragmatic behaviour change strategies are likely to lead to that change. Furthermore, meticulous reporting of these considerations during intervention development, and detailed reporting of the outcomes of evaluation using quality study designs will continue to build the evidence base to inform future practice.
Limitations and further directions
This study is restricted by several important limitations, which should be considered when interpreting the findings. First, the study is limited by the search parameters utilised and the theory application frameworks applied. For example, the review only includes studies that empirically test interventions aiming to change breakfast eating that have been published in peer-reviewed English literature. Hence, studies that undertake experiments to examine breakfast eating, non-English and non-peer-reviewed studies were excluded. Grey literature may contribute important information, and future studies may benefit from examining these sources. Second, due to the heterogeneity in the identified programmes, study populations and reporting of results, a meta-analysis was not possible and a qualitative description of study outcomes was provided. Few studies included effect sizes and OR, limiting our ability to compare effectiveness of breakfast eating intervention. Finally, based on the theory assessments for included studies, there is a clear absence of rigorous theory application; hence, any conclusions drawn in the present review should be interpreted with caution given the absence of strong theory application in studies located in the present review.
Karadağ et al. (Reference Karadağ, Seviğ and Kılıç48) specifically noted that future research needs to go broader and involve environmental/social influences, supporting the recommendations of others(Reference Hoek and Jones81,Reference Wymer82) . Of the theories mentioned or used in this review, SCT is the only one that takes a broader view, but application of this theory within the studies was very weak. This suggests that there is a need for research that embeds and develops SCT to better understand how SCT can be used to positively change behaviour, particularly in the area of breakfast consumption. Future research should test if all three SCT constructs are required to create behaviour change, and whether all three increase the degree of behaviour change. This review is the formative stage of a larger research programme that seeks to embed SCT and test the use of all three constructs to increase breakfast consumption. This review was limited by large variation in evaluation design, measures and timeframe and future research that incorporates consistent outcome measures and intervention durations is recommended to permit meta-analytic studies to be undertaken. Many studies used pre- and post-surveys to record behaviour change, which is known to be impacted by social desirability bias(Reference Denscombe83) and future research that utilises more objective behavioural assessments (e.g. observing food selections) is recommended.
Conclusion
In summary, this review demonstrated that recent behavioural interventions targeting breakfast, and some extending to healthy eating, have had mixed success in increasing frequency of breakfast consumption. Theory, if proven effective, offers a roadmap that practitioners can confidently apply to receive intended outcomes, and the research community is working to identify effective theories that can be applied in public health and behaviour change practice(Reference Armitage, Conner and Prestwich84). This evidence review identified that many studies did not apply theory to develop, measure and evaluate their studies. This is not inconsistent with other research practice(Reference Truong37,Reference Rundle-Thiele, David and Willmott40) . Until theory is proven ineffective, research practice needs to deliver strong theory application ensuring consistent application of theoretical constructs and measures to build an evidence base outlining clearly what works, when, where and why to assure all interventions deliver positive intended outcomes. Moving forward, researchers are encouraged to apply theory using guiding application frameworks(Reference Rundle-Thiele, David and Willmott40). Utilisation of theory application frameworks delivers clear reporting demonstrating how interventions were theoretically informed. Strong theory application can deliver a more comprehensive understanding of behaviour change and the mechanisms that support programme effectiveness to emerge. Demonstrated evidence that theoretical constructs (e.g. individual, social and environmental mechanisms in the case of SCT) contribute to intended outcomes provides evidence supporting theories effectiveness over time. Furthermore, evaluation practice needs to be strengthened, and clear and consistent measurement of all theory constructs is needed to assess the contribution of each and every theory construct in achieving the intended change. Finally, there is a need for studies to be replicated to ensure reproducibility to permit a clear understanding of the role of theory in delivering behavioural change to emerge(Reference Michie and Prestwich85). Research that develops and tests theoretically developed interventions alongside similar testing of pragmatic interventions will shed light on the role of theory in behaviour change. Finally, further research can be applied to understand which behavioural strategies can help increase breakfast frequency in a broad sense and these need to be measured and evaluated using longitudinal methods and objective behaviour change measures.
Acknowledgements
Acknowledgements: Not Applicable. Financial support: No financial support was used on this paper. Conflict of interest: There is no conflict of interest within this paper. Authorship: All authors have read and agreed to the published version of the manuscript. The study was designed by J.A.H., with support from J.E.C. Analyses were conducted by J.A.H and J.E.C. All authors J.A.H, J.E.C and S.R-T contributed to the writing of the drafts and agreed to the form of the final paper. Ethics of human subject participation: Not Applicable.
