Article Text
PDF
Abstract
Objectives National dietary guidelines were introduced in 1977 and 1983, by the USA and UK governments to reduce coronary heart disease (CHD) mortality by reducing dietary fat intake. Our 2015 systematic review examined randomised controlled trial (RCT) evidence available to the dietary committees at the time; we found no support for the recommendations to restrict dietary fat. What epidemiological evidence was available to the dietary guideline committees in 1983?
Methods A systematic review of prospective cohort studies, published prior to 1983, which examined the relationship between dietary fat, serum cholesterol and the development of CHD.
Results Across 6 studies, involving 31 445 participants, there were 1521 deaths from all-causes and 360 deaths from CHD during the mean follow-up of 7.5±6.2 years. The death rates were 4.8% and 1.1% from all-causes and CHD respectively. One study included men with previous heart disease. The death rate from CHD for those with, and without previous myocardial infarction was 20.9% and 1.0% respectively. None of the six studies found a significant relationship between CHD deaths and total dietary fat intake. One of the six studies found a correlation between CHD deaths and saturated dietary fat intake across countries; none found a relationship between CHD deaths and saturated dietary fat in the same population.
Conclusions 1983 dietary recommendations for 220 million US and 56 million UK citizens lacked supporting evidence from RCT or prospective cohort studies. The extant research had been undertaken exclusively on males, so lacked generalisability for population-wide guidelines.
- Public health
- Heart disease
- Dietary
- Cholesterol
- Cohort study
Statistics from Altmetric.com
Introduction
US public health dietary advice was announced by the Select Committee on Nutrition and Human needs in 19771 and was followed by UK public health dietary advice issued by the National Advisory Committee on Nutritional Education in 1983.2 Dietary recommendations in both cases focused on reducing dietary fat intake; specifically to (1) reduce overall fat consumption to 30% of total energy intake and (2) reduce saturated fat consumption to 10% of total energy intake.
The recommendations were intended to address mortality from coronary heart disease (CHD). In 2015, we published a systematic review and meta-analysis,3 which reported that evidence from randomised controlled trials (RCTs), available to the dietary guideline committees, did not support the introduced dietary fat recommendations. This systematic review extends this work by re-examining the totality of epidemiological evidence available at the time the dietary fat guidelines were introduced.
The most comprehensive population study undertaken was the Seven Countries Study.4 This reported that CHD ‘tended to be related’ to serum cholesterol values and that these in turn ‘tended to be related’ to the proportion of calories provided by saturated fats in the diet.5 Keys acknowledged that epidemiological studies could reveal relationships, not causation.4 RCTs provide the best evidence.6
While the UK nutritional guidelines2 made reference to the Seven Countries Study, the US committee document1 did not. Neither publication made reference to any of the RCTs available at that time. However, the US Committee report reported data from the non-randomised, cross-over trial, the Finnish Mental Hospital Study.7 ,8
Although a number of reviews of prospective cohort studies have been undertaken,9–11 no review has examined the epidemiological evidence available at the time dietary fat guidelines were introduced. These dietary fat guidelines have prevailed until 2016 and thus the validity of their evidence base remains important to examine. UK dietary fat guidelines are unchanged. The Dietary Guidelines for Americans, issued in January 2016,12 were conspicuously silent on the subject of total fat, but reiterated that saturated fat should be restricted to no more than 10% of calorie intake.
This systematic review will assess if the published prospective cohort studies available to the dietary committees supported their recommendations on dietary fat. With this in mind, we hypothesised that prospective cohort study evidence available to the dietary committees at the time of issuing recommendations did not support the contention that reducing dietary fat intake would contribute to a reduction in CHD risk or related mortality.
Methods
This systematic review uses the Meta-analysis Of Observational Studies in Epidemiology methodology (MOOSE).13 MOOSE uses parts of the Cochrane methodology for systematic review and meta-analysis of RCTs, Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA), for example the figure for presenting search methodology.14
Cross-sectional studies cannot determine cause and effect. For this reason, and to adhere to best practice evidence available, this review has focused on prospective cohort studies.15
Search strategy
A search was undertaken to identify prospective cohort studies that examined the relationship between dietary fat intake, serum cholesterol and mortality from CHD. Exclusion criteria were: clinical trials; cross-sectional studies; case–control studies. Inclusion criteria were: prospective cohort studies; participants were human adults; primary study outcome was CHD mortality; data related to dietary fat consumption were available; data on CHD mortality and serum cholesterol measurements were available.
The period searched was up to 5 September 1983, as this was the date of the UK dietary guideline committee publication.2 In practical terms, the UK publication would have been in preparation, review and print for some weeks or months before this date. However, the search found that no relevant observational studies were published in 1983, giving confidence that available studies could have been considered. The US dietary guidelines were first issued in February 1977 and so publications before and after this date were separated to clarify which evidence was available to each committee.1
Searches of the literature were performed using MEDLINE, EMBASE and the Cochrane Library. AMED and SIGLE (grey literature sources) were not relied on, as their periods covered were not compatible: from 1985 to 1992 respectively (figure 1).16 ,17
Summary of systematic review profile.
Selection of studies
Of 285 identified articles, 253 were rejected on review of the title and abstract. Of these, 52 were rejected for being review articles. Fifty-four were commentaries, editorials or letters. Twenty-nine were clinical trials, 11 were not observational studies. Twenty-eight related to conditions other than heart disease, primarily cancer, diabetes, hypertension and pregnancy. There were 14 studies where animals or children/adolescents were the primary focus. Twenty-six were articles about pharmacology/blood analysis. Twenty-two were rejected for being related to a particular food or supplement, rather than dietary fat. A further 12 papers were educational material and 5 reviewed dietary compliance in nutritional studies. Thirty-two papers remained. Eighteen were rejected on closer inspection of the full paper: 9 were reviews, commentaries or expert opinion pieces; 4 were cross-sectional/cluster observation studies; 3 were papers on the same study which had no end point data; 1 was educational material and the final 1 was a study of blood clotting. Fourteen met the inclusion criteria. No case–control studies or retrospective cohort studies were found, so no data were lost with the inclusion criteria of prospective cohort studies. Where an abstract was unavailable, the publication type, journal name and meta-tags were reviewed to assess if the article should be rejected (eg, ‘letter’, ‘clinical trial’, ‘paediatric’). Copies of the remaining articles were obtained from university libraries or the British Library.
Once duplication was removed, the remaining 14 papers produced 4 articles documenting 6 studies.4 ,18–20 The four articles were manually searched for references to the earliest available publications. The six studies were: the Western Electric Study;18 ,21 The Puerto Rico Heart Health Program;19 ,22 ,23 The Seven Countries Study;4 The Framingham Heart Study;19 ,24 The Honolulu Heart Program;19 ,25 ,26 and a study conducted in London and the South East.20
To ascertain the validity of eligible observational studies, a pair of reviewers (ZH and BD) worked independently to determine which studies met the inclusion criteria. The same six were agreed on. Risk of bias was further assessed using the Cochrane Collaboration assessment tool for component parts relevant to observational studies, defined as follows: selection bias (cohort appropriately reflected wider population characteristics); detection bias (blinding of outcome assessment); attrition bias (incomplete data outcome); and reporting bias (selective reporting) (figure 2).27
Methodological quality summary: review authors’ judgements about each methodological quality item for each included study.
Data extraction
Table 1 details data extraction of: study name; participant characteristics; whether free from CHD at study entry; years of follow-up; outcomes related to serum cholesterol, total dietary fat and saturated dietary fat for those who developed/died from CHD and those who did not. Other significant associations with CHD, reported by the studies, have been extracted.
Outcome data from included prospective cohort studies for: study name; participant number and age range; years of follow-up; serum cholesterol, total fat and saturated fat for CHD-free versus CHD deaths22 ,24 ,25 or CHD-free versus development of CHD;4 ,20 ,21 and other significant associations found
The process of data extraction revealed some gaps. The Western Electric Study was the only study with complete data for serum cholesterol, total fat and saturated fat in the form in which the development of CHD could be contrasted with those who remained CHD-free.21 These data were not available for CHD mortality. The Seven Countries Study reported correlation coefficients for the relationship between serum cholesterol, total fat, saturated fat and CHD for different countries.4 It did not report dietary fat intake for participants who developed CHD versus those who did not and it did not report dietary fat intake in comparable measurements, for example, grams, or proportion of calorie intake. The London Bus and Bank Study did not report saturated fat data and reported serum cholesterol and total fat in tertiles, which impaired interpretation.20 The three studies of Framingham, Honolulu and Puerto Rico did not report serum cholesterol data in relation to CHD.19 ,22 ,24 ,25 These three studies reported total fat and saturated fat in an optimal way for analysis: the dietary fat intake in grams for those who died from CHD versus those who did not.
No one study satisfied all inclusion criteria. The decision was taken to include all six prospective cohort studies available to the dietary committees, with deficiencies noted. This decision was supported by the fact that the UK dietary guidelines publication referenced The Seven Countries Study and the studies of Framingham, Hawaii/Honolulu, London and Puerto Rico.19 ,20 ,26 ,28
Statistical analysis
The data available are not conducive to meta-analysis. Three of the studies are presented in a comparable format, but meta-analysis was not possible on these alone, as the information required for dichotomous or continuous analysis was not available.19 ,22 ,24 ,25 The data presented in the other three studies was in different formats and also lacked the information that would enable meta-analysis to be undertaken.4 ,20 ,21 Table 1 presents the available data in the absence of forest plots being possible. Table 1 illustrates the lack of significant findings.
Results
Participants and study design
The 6 identified prospective cohort studies included 31 445 male participants. All but 337 participants20 were over the age of 40 years at baseline and were followed for a minimum of 4 years and a maximum of 20 years.
The mean duration of the six cohorts was 7.5±6.2 years. The weighted mean duration (person years by participants) was 5.6±0.8 years.
Five cohort studies excluded men with previous heart disease. The one study that included men with previous heart disease enabled examination of men healthy at baseline in some circumstances.4
All studies had complete outcome data, showing no evidence of attrition bias (figure 2). Two studies were unclear in the assessment of blinding of outcome.4 ,21 The three comparator population studies showed no evidence of selection bias.19 ,22 ,24 ,25 The London Bus and Bank Study20 and The Western Electric Study21 were assessed as having some selection bias for drawing from limited occupations and/or corporations. The selection bias in The Seven Countries Study was stronger, having initially been informed by Keys' analysis of country data29 and having subsequently relied greatly on friendship for cohort selection.28 Five studies were judged low risk for reporting bias, as there was no evidence of any data being withheld. The Seven Countries Study was judged as unclear for reporting bias, as the presentation of data according to thresholds introduced subjectivity (figure 2).27
In the absence of meta-analysis, statistical review of study heterogeneity was not possible. The studies were homogenous in their focus on men; largely free from previous heart disease and in their study aims to review the impact of dietary fat on CHD mortality for a follow-up period of at least 4 years. The age of participants was an area of homogeneity: three of the studies included men aged 45–64 years (52% of participants);19 ,22 ,24 ,25 two of the studies included men between 40 and 55 or 59 years (47% of participants);4 ,21 only one study, accounting for 1% of participants, included a wider age range of 30–67 years.20
Areas of heterogeneity were related to the populations chosen. London bankers, Japanese fishermen, American railroad workers and Puerto Rican farmers, like all the cohorts studied, differed by location, climate, politics, income, ethnicity, genetics, traditional diet, soil quality and other confounding factors.
Examination of the dietary guidelines
None of the six cohorts examined either of the introduced dietary guidelines: a total fat consumption of 30%, or a saturated fat consumption of 10%, of energy intake. Three studies examined the total fat intake and the saturated fat intake, as a percentage of calorie intake, for men who died from CHD compared to men who remained alive.19 ,22 ,24 ,25 One study examined the total fat intake and the saturated fat intake, as a percentage of calorie intake, for men who developed CHD compared to those who did not (the comparison was not available for CHD deaths).21 One study reviewed total dietary fat in tertiles and noted no difference in the saturated fat intake of men in different calorie intake tertiles.20 One study compared total and saturated fat intake by 7 countries and 16 cohorts, as opposed to by men who developed CHD compared with those who did not in each region.4
Outcomes: all-cause mortality
Across 6 studies, involving 31 445 participants, there were 1521 deaths from all-causes during the period of follow-up. The death rate from all-causes was 4.84%, during the mean follow-up of 7.5±6.2 years.
CHD mortality
Across six studies, involving 31 445 participants, there were 360 deaths from CHD during the period of follow-up. The death rate from CHD was 1.14%, during the mean follow-up of 7.5±6.2 years.
The death rates contrast with the 30% death rate for the six RCTs available to the dietary guideline committees3 and reinforce the high death rate in secondary studies. The one study that included men with previous heart disease found that the death rate from CHD for those with pre-existing myocardial infarction (MI) was 20.9%. The death rate from CHD for those without previous MI was 1.0%.4
Significance reported by the studies
None of the six studies found any significant relationship between CHD deaths and total dietary fat intake. One of the six studies found a statistically significant relationship between CHD deaths and saturated dietary fat intake.4
Discussion
The main findings of this systematic review are that the epidemiological evidence available to the dietary committees did not support the introduction of dietary fat guidelines. There were two prospective cohort studies available to the US committee,4 ,21 neither were referenced. An additional four were available to the UK committee,19 ,20 ,22 ,24 ,25 one of which was referenced.4 All available data was not taken into account by the dietary guideline committees and it would not have supported the introduced guidelines had it been considered.
Four of the studies found a significant relationship between higher calorie intake and lower incidence of CHD.19 ,20 ,22 ,24 ,25 The Framingham and Honolulu studies found a significant relationship between higher alcohol intake and lower incidence of CHD.19 ,24 ,25 The London and Honolulu studies found a significant relationship between higher starch/cereal intake and lower incidence of CHD.20 ,25 Affluence and/or activity levels could have been confounding variables in these findings.
Design limitations
As was found with the review of RCT evidence in Harcombe et al,3 the fundamental design limitation of epidemiological evidence available to inform the dietary committees of 1977 and 1983 was that all studies included men only. The epidemiological evidence available at the time of dietary guidelines being introduced had largely evaluated primary prevention men, while the RCTs available at the time had largely studied secondary prevention males.
One study alone provided support for the diet-heart hypothesis.4 This study suffered the most serious limitations: first of selection bias and second of not comparing the development of CHD against non-development of CHD in each cohort. Rather it was an inter country comparison, comparing the development of CHD in one cohort/country with the development of CHD in another cohort/country, which therefore introduced many other confounders.
The summary volume of The Seven Countries Study asserted that smoking, activity levels/exercise and weight played no part in CHD; blood pressure had some observed pattern and CHD tended to be related to total cholesterol and the average proportion of calories provided by saturated fats in the diet.5
By 1960, The Framingham Heart Study had found that smoking increased the risk of heart disease.30 By 1967 the same study had found that exercise had a positive impact on heart disease and weight had an adverse impact.31 In 1953, Morris et al32 demonstrated the benefit of vigorous physical activity to cardiovascular health. Published in 1970, the conclusions of The Seven Countries Study were contrary to evidence of the time. Importance was assigned to a study for its cholesterol and saturated fat findings, which contradicted contemporary evidence about smoking, activity and weight.
Study conclusions
Five of the studies made no mention of dietary fat in their conclusions. The Western Electric Study concluded: ‘No relation was encountered between body weight, mean blood sugar levels, lipoprotein lipase levels, or diet (other than coffee), and the development of coronary heart disease’ (p30).21 Morris et al20 identified healthy and unhealthy patterns of behaviour: ‘Meanwhile, a pattern of healthy living may have been identified: high energy intake and expenditure, high intake of cereal fibre, no cigarettes, with relatively little proneness to heart attack; and another behaviour pattern, of low energy intake and physical inactivity, low intake of cereal fibre, smoking cigarettes-carrying a relatively high risk’ (p.1313). The 1981 publication combining the Framingham, Honolulu and Puerto Rico cohorts summarised the findings of the three studies together: ‘In conclusion, men who developed MI or died of CHD consumed significantly fewer calories (but weighed more) and consumed less alcohol than average’ (p.514).19
The verbatim findings of The Seven Countries Study were: ‘The incidence rate of CHD tends to be directly related to the distributions of serum cholesterol values’; ‘The average serum cholesterol values of the cohorts tended to be directly related to the average proportion of calories provided by saturated fats in the diet’; and ‘The CHD incidence rates of the cohorts are just as closely related to the dietary saturated fatty acids as to the serum cholesterol level’(p.I-194).5
The Seven Countries Study was available to both dietary committees. While the UK nutritional guidelines2 made reference to The Seven Countries Study, the US committee document1 did not. However, the US committee document reported that: ‘The basic research is strongly corroborated by epidemiological studies of populations throughout the world who live quite well on a diet containing as little as 10 per cent calories from fat’ (p.XL)1 and the only study of populations throughout the world at that time was The Seven Countries Study.
Other associations
Keys reported correlation coefficients as follows: median serum cholesterol and saturated fat as a percentage of calories r=0.89 (p.I-170); median serum cholesterol and CHD deaths and infarctions (data for CHD deaths alone were not presented) per 100 people r=0.76 (p.I-172); CHD deaths and infarctions and saturated fat as a percentage of calories: r=0.84 (p.I-174); and CHD deaths and infarctions and total fat as a percentage of calories r=0.40 (p.I-173).33
These correlation coefficients established strong relationships between the component parts of the diet-heart hypothesis, using saturated, not total, dietary fat and using CHD deaths and infarctions. However, strong correlations were found with CHD and other factors, as different as animal protein and television sets, with Gross Domestic Product and living standards suggested as the confounding and possibly causal variables.34 ,35
The 25-year follow-up to The Seven Countries Study36 calculated Pearson correlation coefficients for mean serum cholesterol levels at baseline and CHD deaths at 5, 10, 15, 20 and 25 years of follow-up. The Pearson correlation coefficient was calculated as 0.72 for baseline cholesterol and CHD deaths at 25 years. The data in the 1993 Menotti article has been examined to repeat the correlations found with CHD death rates and mean serum cholesterol to understand the data and methodology used. The same methodology was then used to explore alternative correlations. The strongest relationship found was for CHD death rates and the latitude of the country or cohort in The Seven Countries Study.37 The correlation coefficient for CHD deaths and latitude of the cohort was 0.93. The correlation coefficient for CHD deaths and latitude of the country was 0.96. While Keys did find a strong association with median serum cholesterol and CHD deaths, there were stronger associations that were discoverable.
The latitude finding offers an alternative explanation for the observed relationship with cholesterol and CHD. Vitamin D is made when sunshine synthesises cholesterol in skin membranes.38 In cohorts further away from the equator, cholesterol is less able to be turned into vitamin D. Population mean serum cholesterol levels are higher and concomitantly population mean vitamin D levels are lower. Higher CHD could be associated with lower vitamin D, with cholesterol a marker, not a maker, of heart disease.39
Harcombe et al3 reported that the dietary fat guidelines were not supported by RCT evidence available at the time of their introduction. This systematic review finds that the prospective cohort study evidence available at the time did not support the introduced dietary guidelines. Both reviews reported serious limitations with the availability of primary prevention, both-sex, studies, which are the ones most likely to have generalisability for whole populations.
What are the findings?
Dietary recommendations were introduced in the USA (1977) and in the UK (1983) to reduce overall fat consumption to 30% of total energy intake and reduce saturated fat consumption to 10% of total energy intake. Randomised controlled trial (RCT) evidence available at the time did not support the introduction of these dietary fat guidelines. This study finds that epidemiological evidence at the time did not support the introduced dietary fat guidelines.
No prospective cohort study available to dietary guideline committees found any association between total fat intake and deaths from heart disease.
No prospective cohort study available to dietary guideline committees found any association between saturated fat intake and deaths from heart disease in the same population.
All RCT and epidemiological evidence available for the consideration of the dietary guideline committees had been undertaken on men alone. Evidence available at the time could not be generalised to women.
How might it impact on clinical practice in the future?
Public health advice on dietary fat has prevailed since 1977/1983 in the absence of supporting evidence. The US 2015 dietary guidelines excluded recommendations for total fat for the first time, but maintained the advice to restrict saturated fat. The pool of evidence does not support this recommendation. The UK advice has not changed since 1983. Dietary advice in the UK and the US at least, need re-examination.
Protein is present in all foods, except pure fats and sucrose, and thus tends to form ∼15% of total calorie intake. Restricting total fat intake to 30% concomitantly sets a carbohydrate intake of 55%. Diabetes and obesity have increased since guidelines to restrict fat intake. This association needs examination.
References
Footnotes
Twitter Follow Zoe Harcombe at @zoeharcombe
Contributors ZH conceived the study and was the major contributor to data extraction, writing of the manuscript and the meta-analysis. BD was involved in data extraction. All authors were involved in critical evaluation of content.
Competing interests ZH receives income from writing and from two small self-employment businesses: The Harcombe Diet Co and Columbus Publishing.
Provenance and peer review Not commissioned; externally peer reviewed.