Diabetes and Healthy Eating: A Systematic Review of the Literature

Objective
The objective was to review the literature on healthy eating interventions within diabetes care as the state of evidence exists.

Data Sources
The databases PubMed, Cumulative Index to Nursing and Allied Health Literature, and PsycINFO were searched for the terms “(“healthy eating” or “dietary”) and intervention and diabetes” in the abstracts. In addition, electronic searches of the contents of the specific journals Diabetes, Clinical Diabetes, Diabetes Care, Diabetes Spectrum and Diabetic Medicine were conducted.

Study Selection
Papers were included in the review if the participants were diagnosed with type 1 or type 2 diabetes; results from an intervention to promote dietary change were reported; measurements were taken from two time points: pre- and post-intervention; at least one outcome measured eating behavior; and the articles had been published between 1990 and the present date. Articles were excluded if: they had not been peer reviewed; they were cross-sectional studies; they were descriptive articles such as reviews, advice, lectures or conference proceedings; they were not written in English; they did not use human participants; and the intervention was targeted towards a group which was so specific that the extent to which it could be applied to a general population with diabetes was reduced. Finally samples were not restricted to the United States only, or to any specific age group. After retrieval, one further criterion was added, which stated that any study which did not “include some type of individualized assessment, development of a plan, and periodic reassessment between instructor(s) and participant when directing the selection of appropriate education materials and intervention” was excluded.

Data Extraction
Data were extracted from eligible studies. Each paper included in the review was scrutinized thoroughly and methodically by two independent reviewers, and the relevant data extracted. During this process, two articles were removed, as under closer scrutiny it was found that they did not meet all the inclusion criteria. The total number of articles which were reviewed was 23.

Data Synthesis
Studies included a wide range of outcome measures, making direct comparison between studies difficult. To complicate the issue further, studies varied in how the outcome was assessed or reported. However, results from studies were compared by examining the characteristics of those studies which produced a statistically significant result, with those which did not for the different outcome measures. To make the task more manageable this process was limited to those outcomes which had been measured by at least six studies. Studies were compared in terms of sample types and sizes, duration, and type and content of intervention.

Conclusions
Comparisons between significant and non-significant studies revealed some interesting differences between them, although there did not appear to be any clear patterns across the different outcomes, making it difficult to identify any clear distinctions. However, for studies measuring the outcomes of weight, fat intake saturated fat intake, and carbohydrates, there was a tendency for successful interventions to include an exercise dimension, and group work. There was no clear pattern for sample size or duration of studies. However, it is interesting that some outcomes were more likely to show significant changes in studies of longer duration (e.g. serum cholesterol); whereas others were more likely to show significant changes in studies of shorter duration (e.g. weight, fiber), suggesting that certain behavioral outcomes may be more difficult to maintain. Future research would benefit by ensuring sample sizes are adequate to give sufficient power, and also by including outcomes in the four different areas of learning, behavior, clinical improvement and health status. In addition, interventions should be designed which focus on the maintenance, in addition to the initiation of eating behavior change.