Overall, this randomised controlled study demonstrated that tailored, iterative, printed dietary feedback was more effective than small group nutrition education in increasing fruit intake over a 3-month period in adults at risk of CVD, and as effective as group education in reducing total saturated fat intake. The other areas of dietary behaviour focus in this study, namely intake of vegetables and cereals and grains, were not additionally changed by either dietary intervention, in comparison to the no-education control group who completed the baseline and 3-month post-test dietary measures only.
The increase in fruit intake of 16.7% seen in the TF group meant their mean intake of fruit at 3 months post-baseline was 2.1 serves/day and met the current Australian dietary recommendations for fruit of at least 2 serves/day [32, 33]. Furthermore, the greater effectiveness of the TF intervention was combined with a reduction in total saturated fat intake which was not found to be significantly different from that gained from the more resource-intensive GE intervention.
Such dietary improvements in fruit intake and saturated fat achieved from the TF intervention are meaningful in public health terms . These modest dietary changes, when reached by the population at large have been shown to be of substantial health benefit .
The findings of this study show areas of agreement with the mixed results published on the effectiveness of tailored, nutrition feedback. Other studies have found tailored feedback to be effective in reducing saturated fat intake, and able to effect both a dietary fat behaviour and intake of fruit and/or vegetable . This study's results are also consistent with a number of studies showing mixed effects in terms of the efficacy of a tailored intervention in improving multiple dietary behaviours [13, 20]. The inclusion of grain and cereal behaviours as a dietary target for a tailored dietary behaviour change intervention, in addition to saturated fat, fruit and vegetable intake behaviours has not been previously investigated. In terms of the between dietary intervention comparisons it is difficult to compare this study's results to those of others, as the comparison of tailored dietary feedback to small group nutrition education was in this study led by a dietitian.
Of all the dietary behaviours at baseline- mean fruit intake was closest to the dietary goal and thus on average, the fruit section of the tailored feedback reports would have identified the smallest dietary behaviour change required to meet the goal. It has been hypothesised that strength of tailored feedback is in providing objective individualised dietary feedback which provides a more accurate assessment of where a person's dietary intake is in comparison to goal and in this way tailored feedback incorporates an aspect of interpersonal dietary counselling that may not be part of small group nutrition education [19, 36]. It may be that when this objective feedback identifies that the dietary changes recommended are small, change processes supported by the non-face-to-face TF intervention shows effectiveness. The cognitive factors underpinning the processes of dietary changes will be a topic of further analysis and publication.
This dietary improvement in fruit intake by the TF group was achieved from a baseline mean intake of 1.7 serves per day which is within the population estimate of mean intake fruit intake of West Australian adults at that time, at 1.8 serves/day (95% CI 1.7-1.9) . It should also be noted that the improvement was made in the midst of extensive social marketing campaign for the increase of fruit and vegetable intake, the Go for 2 and 5 ™ campaign .
In this study small group nutrition education was found to be no more effective in producing dietary behaviour change than either tailored feedback or a control group. The reduction in total intake of saturated fat of 18.4% for the GE group between baseline and post-test was significant as a change within that group and approached significance (p = 0.077) in the between-group study effect. The GE group also demonstrated dietary behaviour improvements from baseline to post-test 3 months later in vegetable intake. Thus, the study results were consistent with the evidence base that GE is a successful dietary behaviour change technique at least for some dietary behaviours.
The positive impact of completing detailed dietary measures without further input on dietary behaviours, as seen in this study in the reduction of saturated fat in the C group, has been shown in previous research . The act of completing 7-day estimated food records and dietary questionnaires is an intensive act of dietary behaviour focus and is a form of dietary self-monitoring that can be considered low-level dietary change intervention in itself.
No dietary improvements were seen in the area of grain and cereal intakes for any of the study groups. Other RCT to date have not examined the ability of TF to alter grain and cereal intake (with a focus on wholegrain intake). The inclusion of grains and cereals in the nutrition education interventions (TF and GE) was seen to be important in providing a more complete dietary behaviour change message linked to the reduction of risk of CVD as there is increasing evidence that the increased consumption of grain and cereals, particularly wholegrain, reduces CVD risk . However, the public health nutrition messages on grain and cereal foods, with their provision of greater amounts of dietary energy, and messages about the role of carbohydrates in weight control, are more complex than consistently positive messages in regards to fruit and vegetable intake. Grain and cereal foods have also not had the same focus in health promotion social marketing that the increase in fruit and vegetable intake, and decreasing of fat intake (particularly saturated fat) has.
In this study recruitment was through advertisement and thus study participants would show volunteer bias and have an increased interest in dietary change. This limits the generalisability of the study's results. However, this volunteer group may represent well a group that accesses a tailored dietary feedback intervention, as applications of this intervention could be provided across the internet to person's with CVD risk who 'volunteer'. The participants were found to be representative of the demographic characteristics and dietary characteristics of Australian adults at CVD risk in the 40-65 year old age range. Also, this study with its assessment of change over 3-months looks at only short-term dietary behaviour change. As such it is not known whether the changes from the interventions would have been sustained. These questions would be answered by further research of longer duration.
This study focused on the efficacy of TF to promote dietary behaviour change in comparison to the more expensive face-to-face dietary intervention of GE, and a wait-listed C group. As such it was dietary behaviours that were the focus and the outcome measures for the study. The form of dietary measure used, the 7-day estimated record, is considered a robust tool for determining the food group based and saturated fat intake related dietary outcome measures used in this study . Furthermore, that this study included this measure as an independent measure of dietary change is an improvement on most previous studies. In most studies assessing the effectiveness of tailored feedback, dietary questionnaires only have been used to measure dietary change and often these questionnaires have been the intervention tool itself [13, 14].
This study with its comparison of tailored, iterative, printed dietary feedback to the commonly used nutrition education technique of group education sessions is the first of its kind. It has added to the knowledge base on the effectiveness of an intervention that does not include face-to-face education can have on dietary behaviour. This trial provides evidence that over the short-term that tailored, iterative feedback was more effective than group education in improving fruit intake, and was as effective in reducing total saturated fat intake in adults 40-65 years at risk of CVD. Neither tailored feedback nor group education demonstrated greater efficacy in changing the intake of vegetables, or grain and cereals in this subject group.