点击显示 收起
1 From the Institute of Human Nutrition, University of Southampton, Southampton, UK (RLT); School of Health and Social Care, University of Teesside, Middlesbrough, UK (CDS); MANDEC, University Dental Hospital of Manchester, Manchester, UK (LH); MRC Biostatistics Unit, University of Cambridge, Cambridge, UK (JPTH); Primary Medical Care, University of Southampton, Southampton, UK (PSL); Leicester Nutrition and Dietetic Service, Leicester, UK (DT); Department of Social Medicine, University of Bristol, Bristol, UK (SE).
2 Presented at the Third Heelsum International Workshop, held in Heelsum, the Netherlands, December 10–12, 2001. 3 Supported by the Systematic Reviews Training Unit, Institute of Child Health, University of London, and the British Dietetic Association. 4 Address reprint requests to RL Thompson, Institute of Human Nutrition, University of Southampton, Level B, South Academic Block, Southampton General Hospital, Southampton, SO16 6YD, United Kingdom. E-mail: r.l.thompson{at}soton.ac.uk.
ABSTRACT
Background: Dietary advice to lower blood cholesterol may be given by a variety of means. The relative efficacy of the different methods is unknown.
Objective: The objective was to assess the effects of dietary advice given by dietitians compared with advice from other health professionals, or self-help resources, in reducing blood cholesterol in adults.
Design: We performed a systematic review, identifying potential studies by searching the electronic databases of the Cochrane Library, MEDLINE, EMBASE, CINAHL, Human Nutrition, Science Citation Index, and Social Sciences Citation Index. We also hand-searched relevant conference proceedings, reference lists in trial reports, and review articles. Finally, we contacted experts in the field. The selection criteria included randomized trials of dietary advice given by dietitians compared with advice given by other health professionals or self-help resources. The main outcome was difference in blood cholesterol between the dietitian group compared with other intervention groups. Inclusion decisions and data extraction were duplicated.
Results: Eleven studies with 12 comparisons met the inclusion criteria. Four studies compared dietitians with doctors, 7 with self-help resources, and 1 with nurses. Participants receiving advice from dietitians experienced a greater reduction in blood total cholesterol than those receiving advice from doctors (-0.25 mmol/L, 95% CI -0.37, -0.12 mmol/L). There was no statistically significant difference in change in blood cholesterol between dietitians and self-help resources (-0.10 mmol/L, 95% CI -0.22, 0.03 mmol/L).
Conclusions: Dietitians appeared to be better than doctors at lowering blood cholesterol in the short to medium term, though the difference was small (about 4%), but there was no evidence that they were better than self-help resources or nurses.
Key Words: Dietitian • dietary advice • cholesterol • ischemic heart disease • health professionals • self-help resources
INTRODUCTION
Differences in the average levels of blood cholesterol between communities or populations are largely determined by differences in their diets. Countries with high dietary saturated fat intake and a low ratio of polyunsaturated to saturated fatty acids have high average cholesterol levels (1). The generally small changes in blood cholesterol that are produced by interventions aimed at free-living populations (2–5) may be attributable to one of 2 factors, or both: people are not complying with the dietary advice given in community settings, or advice givers could be more effective in their role.
By contrast, substantial reductions in blood cholesterol may be achieved in "metabolic ward" settings over short time periods with strict dietary control (6). Dietitians are specifically trained and motivated to provide high-quality dietary advice. Dietitians have a variety of different approaches and select the most appropriate for each patient. Because of the limited number of dietitians and the large proportion of the population who are at risk for, or have, ischemic heart disease, much of the dietary advice is given by physicians and nurses rather than by dietitians with extensive nutrition training (7). The effectiveness of dietary advice given by dietitians compared with that given by other health professionals or self-help resources is unknown; this area of health care has not been extensively researched. Knowledge of the relative effectiveness would inform policy decisions on the best way to manage raised blood cholesterol in the general population. If dietitians prove to be much better than other health professionals at reducing blood cholesterol, it may be sensible to increase the number of dietitians; alternatively, if there is not a large difference, some dietetic work could be channeled into training nurses and doctors to give appropriate advice. It is important that patients get the best treatment but that this also fits within current resources and financial constraints.
The Cochrane Collaboration (8) is an international organization that aims to help people make well-informed decisions about health care by preparing, maintaining, and ensuring the accessibility of systematic reviews of the effects of health care interventions. A Cochrane Review addresses a clinically relevant question and is carried out using a strict methodology according to a previously published protocol. The methods of the review are evidence based where possible and aim to minimize bias. Cochrane Reviews are regularly updated to take account of accumulating evidence.
We carried out a systematic review according to the Cochrane methodology to answer the following question: In adults, what is the relative efficacy of dietary advice given by dietitians compared with advice from other health professionals, or self-help resources, in reducing blood cholesterol?
METHODS
A full and regularly updated version of this review—including methods, a detailed description of the included studies, and data tables—is available in the Cochrane Library (9). This article concentrates on the main outcome of interest; further information on secondary outcomes is available in the full version. The Cochrane Library is a database of systematic reviews and other evidence on the effects of health care available from Update Software (http://www.cochrane.org). Briefly, we developed a search strategy to identify randomized controlled trials of dietary advice to reduce blood cholesterol. Accepted interventions included dietary advice given by dietitians or nutritionists compared with advice from other health professionals (eg, doctors, nurses) or self-help resources. Any study in which the term "random" was used to describe the method of study group allocation was included. The participants were noninstitutionalized adults with or without existing heart disease or previous myocardial infarction. Studies were at least 6 wk in duration from baseline. Interventions of dietary advice using dietary supplements were excluded. Studies in which the intervention included the provision of meals or food items, or where lipid-lowering drugs were given to the intervention group only, were excluded. The outcome of interest was difference in blood cholesterol for the dietitian group compared with other intervention groups.
We searched MEDLINE (1966 to January 1999), EMBASE (1980 to December 1998), Human Nutrition (1991–98), CINAHL (1982 to December 1998), and the Cochrane Library (to Issue 2 1999). The search was not limited to English language publications. The search terms included food- and diet-related terms and ischemic heart disease (we did not search specifically on the health professional). Hand-searching of the proceedings of nutrition and cardiology conferences was undertaken. We also used the reference lists of included studies and published review articles on diet and heart disease.
Experts in the field were contacted for references to studies not yet identified by the search process. A forward search on included studies was conducted using the Science Citation Index (1981–98) and the Social Sciences Citation Index (1981–98).
Each potentially relevant study was assessed for inclusion in the review independently by at least 2 reviewers (RLT, LH, PSL). Differences between the reviewers were resolved by discussion. It was planned that if after discussion agreement between the reviewers could not be established a third reviewer would be used. However, this was unnecessary. The statistic for the agreement between the reviewers was 0.68 (a good agreement).
The following information on outcome variables from each trial was extracted: number of subjects, baseline and final values, mean change (final value minus baseline value), and standard deviation of change for both intervention and control groups. If standard deviations were not reported, they were estimated using the methods described by Follmann et al (10).
The review was carried out under the auspices of the Effective Practice and Organization of Care (EPOC) Cochrane Group. The standard EPOC criteria (11) were used to assess the methodological quality of randomized trials. In brief, the criteria were concealment of allocation (protection against selection bias), follow-up with participants at end of study (at least 80%), blinded assessment of blood cholesterol, similar baseline blood cholesterol measures for each intervention group, reliable blood cholesterol measure, and protection against contamination.
Data analysis
Net differences (ie, dietitian minus other group differences) at the longest duration of follow-up available were computed. Unstandardized differences in means were examined using the random effects model. Standardized differences in means were also performed to allow for the difference in length of follow-up between the studies. Where there were no material differences between results using standardized and unstandardized differences in means, the unstandardized results are presented. Differences between the results of the trials were checked for heterogeneity by visual inspection of the graphs and by statistical test (chi-square).
RESULTS
The literature search yielded 14053 references. Figure 1 gives a flow chart of studies assessed and excluded at various stages of the review. Of these, 150 papers were obtained. After examination by 2 assessors, 11 studies met the inclusion criteria. The main reason for exclusion was that "intervention was not dietitian versus other health professional or self-help resources" (91%). Other criteria that were often not met were "blood cholesterol not measured" (32%) and "study was not a randomized trial" (18%). Details of the studies included in the review are shown in Tables 1 and 2. There were 4 studies comparing advice from dietitians with advice from doctors (13, 16, 18, 20), 7 comparing advice from dietitians with advice from self-help resources (12, 14–15, 17, 19, 21–22), and only one comparing dietitians with nurses (19). Most studies were carried out in the United Kingdom, the United States, and Australia. Studies were carried out in a range of settings, including general practice (13, 16, 19, 21), workplace (12, 18, 20), and a variety of clinic settings (14–15, 17, 22). The health status of participants varied between studies at baseline: participants had slightly raised blood cholesterol values (12, 15), raised lipid values requiring treatment (13, 16, 18–21), raised fasting plasma glucose (14), risk factors for diabetes (22), or previous heart disease (17).
FIGURE 1. . Summary of systematic review profile.
View this table:
TABLE 1 . Characteristics of participants, settings, and study length
View this table:
TABLE 2 . Characteristics of intervention groups
The duration of the studies varied from 6 wk to 104 wk. Most interventions were diet alone, but one study included exercise in both intervention and control arms (14). Table 3 describes the dietary interventions used in the trials. Those participants seen by a doctor tended to have less frequent appointments or less time at appointments than those seen by a dietitian (13, 16, 18, 20). The self-help resources were generally just simple leaflets. In 3 of the 4 (13, 18, 20) trials of dietitians versus doctors, participants randomized to the dietitian were also seen by a physician (although no dietary advice was received). In some studies, the advice was given in individual consultations (13–14, 16–17, 19), whereas in others it was given in group sessions (12, 15, 22); a few studies used both approaches (18, 20–21).
View this table:
TABLE 3 . Details of the dietary intervention methods
The methodological quality of the studies is described in Table 4. All 11 of the reported studies had methodological weaknesses according to the EPOC criteria (11). Judged by the 6 quality criteria, 2 studies met 5 of the criteria (18, 20), 4 studies met 4 criteria (14–16, 21), 2 studies met 3 criteria (13, 19), and 3 studies met 2 criteria (12, 17, 22). In general, studies comparing dietitians with doctors were of a higher quality than studies comparing dietitians with self-help resources.
View this table:
TABLE 4 . Quality assessment with the use of Cochrane criteria1
Table 5 and Figure 2 show the change in blood cholesterol for each intervention. The effects on blood cholesterol are presented separately for each comparison (dietitians vs doctors, dietitians vs self-help resources, dietitians vs nurses) in Table 5. The meta-analysis plot is shown in Figure 2. There was no heterogeneity detected. The random effects analyses demonstrated a reduction in blood cholesterol in participants receiving advice from dietitians compared with doctors (change -0.25 mmol/L, 95% CI -0.37, -0.12 mmol/L).
View this table:
TABLE 5 . Change in blood cholesterol for dietitian compared with other health professional or self-help resources
FIGURE 2. . Change in blood cholesterol for dietitians compared with other health professionals or self-help resources. Study-specific differences in means, with areas proportional to the weight assigned to that study in the meta-analysis, ; error bars indicate 95% CIs. The pooled differences between means (meta-analysis) with 95% CIs, .
The random effects analyse s for dietitians versus self-help resources showed a much smaller reduction in blood cholesterol (change -0.10 mmol/L, 95% CI -0.22, 0.03 mmol/L). The 2 studies demonstrating large effects for dietitians (17, 22) were both rated as being of lower quality than the other studies. Overall, comparing dietitians with either self-help resources or doctors, dietitians performed better than the other methods (change -0.14 mmol/L, 95% CI -0.24, -0.05). We did not include the data for nurses because the only study on nurses also reported on self-help resources, and therefore both these interventions were compared with the same group of participants receiving advice from dietitians.
Funnel plots were prepared combining all studies. These showed that the smaller studies tended to favor dietitians over the other methods. This may represent publication bias.
DISCUSSION
Dietary advice from a dietitian appears to be more effective in reducing blood cholesterol than advice from a doctor. The comparisons of blood cholesterol reductions observed with self-help resources and with nurse advice were consistent with the effect observed in comparisons with doctors, although equally consistent with a lack of relative efficacy.
The effect on blood total cholesterol of advice from a dietitian compared with advice from a doctor was small (-0.25 mmol/L, about 4%) with a 95% confidence interval between 2% and 6%. A 6% lowering is significant for borderline lipid disorders before drugs are started. The quality of the 4 studies was superior to that of those comparing dietitian and self-help resources. It is worth noting that in 3 trials (13, 18, 20) the participants who were randomized to receive advice from the dietitian also saw a doctor, although not for dietary advice. It is possible that the presence of the doctor improves participant compliance. Participants randomized to the dietitian generally received more time with a health professional than those randomized to the doctor.
The effect of advice from dietitians or self-help resources on blood cholesterol is more uncertain. The overall effect was small (-0.10 mmol/L, <2%, with 95% CIs ranging from about -3.7% to 0.5%). The poorer performance of the dietitians in comparisons with self-help resources may reflect better adherence to diet among well-motivated patients, but study characteristics and the play of chance may explain this rather unexpected finding. Clearly, dietary advice is a weak intervention compared with use of cholesterol lowering statins, where reductions in blood cholesterol of around 20–25% are achieved (23).
The comparison between a dietitian and a nurse is important because many nurses give dietary advice as part of health promotion for ischemic heart disease. Nurses are considered an essential component of the primary care team and consequently have more opportunities to provide dietary advice than dietitians because they have more patient contact. One trial (19) had 3 arms: dietitian, nurse, and self-help resources. The number and length of appointments was similar for the dietitian and nurse. At the end of the trial there were no differences between the intervention groups in total blood cholesterol. The differences were all small, between 0.10 and 0.18 mmol/L. Given the rather disappointing effects of dietitians and nurses in this single trial, it would be worth attempting to conduct a further trial examining the effects of dietitians and nurses versus controls and to examine, using qualitative methods, the nature of the interaction with patients.
The number of studies included in the review is relatively small, and the studies span a large time frame, but this is the best evidence available. None of the studies fulfilled all the quality criteria. The smaller studies showed bigger effect sizes favoring the dietitian. This may relate to publication bias in which studies favoring self-help resources were unpublished. There was imbalance in the randomization process for some of the studies, most notably the smaller studies where participants with higher baseline blood cholesterol appeared to be randomized to the dietitian group. In these studies the size of the difference in means appeared greater. This may be partly due to regression to the mean, or it may be easier to motivate patients to reduce blood cholesterol when they know they have higher baseline levels. It is possible that in small trials dietitians were themselves more motivated and able to provide higher-quality intervention than in larger trials with more patients requiring treatment.
There was no apparent relationship between duration of trial and difference in outcome between intervention groups.
Cost-effectiveness of the interventions is an important factor that needs to be addressed. One study did assess treatment costs of dietitians versus self-help resources (21). The 1993 costs from randomization to the end of follow-up after 1 y were equivalent to $80 for the self-help resources and $370 for the dietitians—nearly a 5-fold difference. With a view to the self-help method being less expensive than using dietitians, nurses, or doctors and not substantially different in the effectiveness for reducing blood cholesterol, the self-help method is likely therefore to be more cost-effective; however, more research is required in this area. There is also a resource issue: risk of ischemic heart disease is an important public health problem and the limited availability of dietitians and other health workers to provide dietary advice hinders prevention efforts.
Only one study (13) reported patient satisfaction with the interventions. Participants receiving advice from a nutritionist indicated that they used the provided literature more, found nutritionists more helpful and less difficult, and were more positive in recommending the program than participants receiving advice from doctors. However, this may reflect the greater amount of time spent with the participants randomized to see the nutritionist.
In summary, the evidence of these studies indicates only a small effect on blood cholesterol of dietary advice given by a dietitian compared with other health professionals or self-help resources.
ACKNOWLEDGMENTS
We are grateful to the following investigators, who provided us with further information: Russell V Luepker (University of Minnesota), H Andrew W Neil (University of Oxford), and Alexandra Barratt (University of Sydney). We are also grateful to Stuart Logan, Betsy Anagnostelis, and Leanne Jones from the Systematic Reviews Training Unit, Institute of Child Health, University of London.
There were some potential conflicts of interest. Rachel Thompson, Carolyn Summerbell, Lee Hooper, and Diane Talbot are dietitians. This review was funded in part by the British Dietetic Association.
REFERENCES