Please note: In 2003, the CTF updated its Grades of Recommendations to include an "I Recommendation" for situations where insufficient evidence exists to allow a recommendation to be made. (Formerly, these situations were captured under a "C Recommendation".) This change is not retroactive, and all "C Recommendations" made prior to 2003 have not been reevaluated in light of the new "I" recommendation grade. For a discussion of these recommendation grades, please link to the 2003 article in the Canadian Medical Association Journal here.
These recommendations were finalized by the Task Force in June 1994
Contents
OverviewIn the ensuing 10 years there was a dramatic increase in knowledge about and general interest in cholesterol. The Task Force has re-examined the evidence published subsequent to its 1993 update< 1> and again finds insufficient evidence to recommend routine cholesterol screening but endorses case-finding in men 30-59 years old, corresponding to the group where drug therapy has been shown to be effective.
Burden
of SufferingAn increased blood cholesterol level, specifically a high low-density lipoprotein (LDL) cholesterol, is closely linked to the severity of atherosclerosis in the coronary arteries and is a major risk predictor of clinical CHD.<2,3> In addition, there is a continuous, graded relation over most of the serum cholesterol range.<3> There is also increasing evidence from epidemiologic studies that other cholesterol subfractions (high-density lipoprotein cholesterol (HDL-C), very low-density lipoprotein cholesterol (VLDL-C)), the carrier apolipoproteins of cholesterol and a-lipoprotein) may help to predict the risk of myocardial infarction.<4> Several studies suggest an independent and inverse association between levels of HDL-C and risk for CHD<5> and HDL-C and triglyceride levels are independent predictors of CVD death in women.<4>
Recent data on risk factors for cardiovascular disease in Canada come from heart health surveys in each province between 1986 and 1990 involving over 20,000 men and women aged 18 to 74 years. Among men and women the prevalence rates of a total plasma cholesterol level of 6.20 mmol/L or more were 18% and 17% respectively, an LDL-C level of 4.10 mmol/L or more 16% and 14%, an HDL-C level below 0.90 mmol/L 13% and 4%, and a triglyceride level of 2.30 mmol/L or more 20% and 11%.
The prevalence rates of hypercholesterolemia and CHD are higher among older people;<6,7> menopause has also been associated with a major increase in serum total cholesterol concentration (19% over an 8-year period).<8> Among older people the gradient of CHD risk with increasing total cholesterol levels is flatter, but risk has been shown to be significant for both elderly women and men in large studies<9> although in some reports it disappears after 64 years of age. While the strength of the association between cholesterol and CHD may decline with age, absolute excess risk does not<9> given the almost ten-fold difference in the incidence rate of death from CHD in men 65 years or older and men under 65.
As people become very old, however, the relation between total cholesterol and longevity appears to change. Several studies, including very elderly women and men, reported higher death rates in those with lower cholesterol levels or among those with the greatest decline in cholesterol levels.<10> It may be that those who survive to a very advanced age are no longer susceptible to elevated total cholesterol levels or that a low cholesterol concentration in the very old is simply a marker of chronic noncardiac disease. The benefits of therapy for this group are questionable.
Although the risk of CHD is strongly related to the serum total cholesterol level, at any level of cholesterol the risk varies widely depending on the presence of other risk factors. For example, in the Framingham Heart Study it was found that a 45-year-old man with a serum cholesterol level of 7.40 mmol/L, an elevated systolic pressure, electrocardiographic evidence of left ventricular hypertrophy and glucose intolerance who smokes is 12 times more likely to have cardiovascular disease over an 8-year period than a man of the same age with the same cholesterol level who is free of the other risk factors. In contrast, over 8 years the risk for a 45-year-old man with a serum cholesterol level of 8.00 mmol/L but no other risk factors would be 2.4 times that of a man with the same risk profile whose cholesterol level is normal (5.40 mmol/L). Excess body weight in white American men has been associated with deleterious changes in the lipoprotein profile.<11> Such evidence was the basis for several trials of multifactorial interventions to prevent cardiovascular disease.<12,13>
In the context of routine clinical practice the use of the serum total cholesterol level to predict future CHD events in individual patients is not straightforward. In familial hypercholesterolemia (FH), for example, the clinical expression of CHD can occur at very different ages, even among siblings with closely matched cholesterol levels. Lower cutoffs have been proposed for screening close relatives of confirmed FH cases.<14> In the Framingham cohort the distribution of serum cholesterol levels in men under the age of 50 years who had CHD was very similar to that in subjects who remained disease-free; the average cholesterol level in the former group was only 6.30 mmol/L. In the Pooling Project more than 40% of asymptomatic men aged 40 to 59 years who experienced their initial CHD event during an 8.6-year follow-up period had a serum cholesterol level of less than 6.30 mmol/L, and about 88% of those with a higher level had no clinical signs of CHD. Thus, the serum cholesterol level is a poor discriminator between people destined to have symptomatic CHD and those who will remain symptom-free. It is this weak predictive power that has been a source of concern about recommending universal screening.
ManeuverThe rate of false positive results can be lessened by repeated measurement of the total cholesterol level which will reduce biologic variability or by selective lipoprotein analysis which will identify individuals with high levels of protective HDL (usually in women). The rate of false negative results may be reduced by lowering the level of total cholesterol for retesting or by adding lipoprotein analysis to initial cholesterol testing. The latter approach, for example, would identify individuals with low HDL levels who are at increased risk for CHD, but would also increase costs, introduce greater analytic and biologic variability,<18> and reduce accessibility to high-quality analytic procedures.
Some authorities have recommended including a non-fasting measurement of HDL-cholesterol in screening for lipid disorders<15> because of the association of a low HDL level with CHD mortality and morbidity even amongst subjects without elevated total cholesterol. Apart from greater analytic and biologic variability of HDL determinations<17,18> and increased cost there are few clinical trial data linking changes in plasma concentration of this or other lipid fractions to reductions in the CHD death rate. The effects of labelling a person as having hypercholesterolemia have been considered by several groups and there appear to be few or no adverse consequences.<19,20>
Key elements in current dietary recommendations are to: 1) reduce the intake of total fat to 30% of the total energy intake and the saturated fatty acids to below 10% of the total energy intake; 2) increase the intake of complex carbohydrates and either cis-monounsaturated or polyunsaturated fatty acids; and 3) restrict the number of kilojoules for overweight people.<15,16,21> Lowering of the dietary cholesterol intake to less than 300 mg/d is also advocated, although the additive effect of this change on the plasma cholesterol and lipoprotein concentrations against the recommended background diet is controversial and may be small. When compared with the average North American diet, these dietary changes can reduce the plasma total cholesterol level by about 14%, the LDL-C level by 16% and the HDL-C level by 3% in healthy, noninstitutionalized people, although a recent trial in free-living individuals produced only a 5% reduction in total cholesterol.<22>
The evidence for a favourable effect of short-term dietary manipulations on blood lipids in controlled settings is compelling, but the long-term effectiveness in asymptomatic outpatients has generally been disappointing. Net reductions in total cholesterol level of 1-13% have been reported in various populations with the largest reductions occurring in patients with symptomatic CHD<23,24> or patients with diets very high in fat. The disappointing results have been attributed generally to poor dietary compliance and sustained contact appears to be a critical element in maintaining a good response. Several other factors appear to influence the magnitude of the cholesterol response to dietary changes including baseline cholesterol level, weight change, level of physical activity, presence or treatment of other risk factors, and alcohol intake.
Several classes of lipid-lowering drugs are available. The bile acid sequestrants (cholestyramine and colestipol) and nicotinic acids or niacin have been studied extensively in clinical trials, and their long-term safety has been established.<15> Typically, these agents lower the total cholesterol and LDL-C levels by 10% to 15% in usual practice settings. Their acceptability, however, has been limited by the hardships related to their use, their lack of palatability and the high incidence rate of side effects, which include gastrointestinal distress, pruritus and severe flushing.
The fibric acid derivatives (clofibrate and gemfibrosil) have also been extensively evaluated in long-term clinical trials and are particularly effective in lowering elevated triglyceride levels. They also lower the total cholesterol and LDL-C levels by 10% to 15%, and gemfibrozil, at least, increases the HDL-C level by about 10%.<25,26> Of the two drugs gemfibrozil is preferred because of its better safety record.<15>
Probucol lowers total cholesterol level in part by reducing HDL-C level, the cardioprotective lipid fraction. The drug, however, has an antioxidant action that may inhibit LDL oxidation and retard atherosclerosis. Estrogens may lower LDL-C by 15% and raise HDL-C by about 15%. Some authorities suggest women consider estrogen before other cholesterol-lowering drugs,<15> although their influence on the incidence of uterine and breast cancer is still uncertain (see Chapter 52).
The most significant advance in the treatment of hypercholes-terolemia is the introduction of 3-hydroxy-3-methylglutaryl-coenzyme A (HMG CoA) reductase inhibitors such as lovastatin into clinical practice. Unlike other lipid lowering agents this class of agent is well-tolerated, very effective in lowering cholesterol levels, and widely accepted. Clinical studies have reported reductions in LDL-C levels by as much as 40%.<27> Adverse effects including a rise in liver or muscle enzymes appear to be uncommon. The possibility that drugs of this class produce cancer, lens opacity, sleep disturbance and other health problems has not been excluded, since the long-term safety of these agents has not yet been established.
Primary
Prevention TrialsThe main features of the primary prevention trials are summarized in Table 1. In brief, the trials enrolled predominantly middle-aged men without any clinical evidence of CHD. There were, however, exceptions. About half of the participants in the LAVDS were 65 years or older, and roughly an equal proportion of study subjects in the UCS, MCS, and EXCEL were women. In addition, a sizeable minority of the subjects in the LAVDS, UCS and EXCEL had clinical signs of CHD. In the WHO trial 2% of the enrollees reported that they had angina pectoris, and men with a positive exercise test result but no other manifestations of CHD were not excluded from the LRC trial. The mean baseline total cholesterol levels were elevated only in the drug trials as the presence of hyperlipidemia was not a criterion for entry into the diet studies. Each drug trial used a different cholesterol-lowering agent, whereas in the two dietary trials polyunsaturated fatty acids were substituted for saturated fatty acids as the primary therapeutic maneuver leaving total fat intake unchanged.
The quality of the UCS was the poorest of the intervention studies. It had no uniform criteria for CHD, no double-blinding and a high withdrawal rate. The other four drug studies were of comparable and high quality. The two dietary intervention studies had quality scores between those of the UCS and the other three drug studies. The results of data aggregation were not altered by excluding the UCS from the analysis.
The intermediate measure of effectiveness, alterations in the blood lipid levels, is presented in Table 2. The mean net reduction in the blood total cholesterol level ranged from 9 to 23% in the drug studies, and 14% in the diet trials. The direction of change in the triglyceride levels in the studies in which it was measured was consistent with the known effects of the interventions on this blood lipid.
The data in men from the primary prevention trials were combined and published in a Task Force report on cholesterol lowering.<1> Several points emerged. First, there was a consistent and significant reduction in the number of non-fatal cardiac events in the actively treated group (odds ratio (OR) 0.74, 95% confidence interval (CI): 0.64 to 0.85) and a slight, but insignificant, fall in the number of cardiac deaths (OR 0.90, 95% CI: 0.71 to 1.14). These benefits were offset by a significant increase in the rate of death from non-cardiac causes (OR 1.19, 95% CI: 1.03 to 1.39), which resulted in a slight but insignificant excess in total mortality. Cholesterol lowering was associated with consistent and significant increases in the number of violent deaths (OR 1.78, 95% CI: 1.17 to 2.72) and events of gallbladder disease (OR 1.69, 95% CI: 1.28 to 2.23), and a less consistent and borderline significant increase in the number of deaths from cancer (OR 1.41, 95% CI: 1.05 to 1.89). These small excesses of cancer and trauma deaths have been attributed to chance.<33>
Since this publication the results of a trial examining the short-term efficacy and safety of lovastatin have appeared which included data on 3,390 women with moderate primary hypercholes-terolemia.< 27> In light of increasing recognition of the importance of preventing CHD in women, the data of all primary prevention trials including EXCEL and the two older studies, UCS<30> and MCS<32>, that enrolled women were reanalyzed using standard meta-analytic techniques. Data on 38,940 study subjects were pooled (Tables 3 and 4). Surprisingly cardiac deaths were not reduced in the drug trials of cholesterol lowering (OR 1.00, 95% CI: 0.80 to 1.26) and only slightly so when the data of the drug and diet trials were combined (OR 0.99, 95% CI: 0.83 to 1.18). As noted before<1> cholesterol-lowering was associated with a significant increase in non-cardiac deaths (OR 1.17, 95% CI: 1.03 to 1.34) and the increase in total mortality (OR 1.10, 95% CI: 0.99 to 1.22) approached but did not reach conventional levels of significance (p=0.082). There was a significant increase in deaths from cancer (OR 1.30, 95% CI: 1.02 to 1.66) and a non-significant increase in deaths from violence. These increases were observed in both the drug and diet trials.
Other meta-analyses of primary prevention randomized trials of cholesterol-lowering produced conclusions that were similar to those conducted by the Task Force.<34,35> Law and colleagues, however, in an analysis that included subjects with and without known ischemic heart disease, reached different conclusions.<36> In their report which combined the results of both primary and secondary prevention trials there was a significant reduction in CHD, a non-significant reduction in all-cause mortality, and a non-significant rise in non-cardiac deaths. The difference in findings and conclusions may well be explained by the inclusion of secondary prevention trials in their pooled analysis. Unlike primary prevention trials where less than one-half of deaths were cardiac in origin, the cause of death in these trials were largely from CHD (on average greater than 85%). Thus there is little "opportunity" to die from a non-cardiac cause.
Observational
StudiesCohort studies conducted within communities have also demonstrated an association between CHD and dietary con-stituents.< 37-42> The risk of death from CHD is positively related to the baseline level of dietary scores, inversely related to the intake of polyunsaturated fatty acids and fibre, and positively related to the dietary intake of cholesterol and saturated fatty acids. There are several caveats to these observations. First, these relations often remained statistically significant even after adjustment for the serum total cholesterol level and other variables. Second, the subjects enrolled in the cohort diet studies were largely adult men aged 30 to 69 years, thus, limiting, at least potentially, the generalizability of the results to adult women, the elderly, children, and young adults. Finally, the dose-response relation of dietary scores and CHD risk is not linear, as might be predicted.<37,38>
Recommendations
of OthersThe Canadian Cholesterol Consensus Conference<16> recommended case-finding of adults 18 years of age and older with particular attention being given to those with cardiovascular disease risk factors, family history of CHD, or clinically apparent cardiovascular disease. Testing should be repeated every 5 years if blood lipids are normal. If total cholesterol is between 5.2 and 6.2 mmol/L, lipoprotein analysis should be done.
In 1993 the U.S. National Cholesterol Education Program recommended that HDL-C determinations (and in some cases TG level measurement) accompany measurements of total cholesterol when healthy individuals are being assessed for CHD risk.<15> Emphasis was placed on screening those at higher risk of CHD based on total cholesterol measurement or other known risk factors, in locations where accuracy of measurement, appropriate counselling and follow-up can be assured – at least 2 and preferably 3 samples taken in the fasting state at least 1 week apart.
More selective screening strategies have been proposed that entail the use of prescreen information to focus cholesterol screening on people at increased risk for CHD. Information on risk factors for CHD would be obtained during the course of clinical evaluation, as proposed by the Toronto Working Group on Cholesterol Policy,<43> or by questionnaire. Although the ability of a risk assessment to identify people with a high total cholesterol level has not been formally evaluated, experience with the questionnaire method indicates that almost 25% of those with hypercholesterolemia were not identified. This case-finding strategy is built on the premise that people at increased risk because of other CHD risk factors have more to gain from treatment. The results of multifactorial trials of primary prevention of cardiovascular disease in high-risk subjects, however, generally have not supported this assumption, despite favourable responses to interventions on risk factors.
The recommendations of the U.S. Preventive Services Task Force are currently under review.
Conclusions
and RecommendationsCase-finding should be directed to all men aged 30 to 59 years who present to their physician’s office for any reason, individual clinical judgement being exercised in all other circumstances. This selective form of case-finding stresses the importance of the link between the detection of hypercholesterolemia and the favourable effect of lowering the cholesterol level on the incidence rate of CHD in this group. Cholesterol testing should be considered when other CHD risk factors are present such as smoking, hypertension, or diabetes mellitus, or when there is a strong family history of hypercholes-terolemia or premature CHD. People with an initial total cholesterol level above 6.2 mmol/L should undergo another nonfasting test in l to 8 weeks.
The optimum frequency of repeat testing for people with a total cholesterol level of 6.2 mmol/L or less is unknown, but a prudent approach might be to have another test done within 5 years. Because the effectiveness of cholesterol screening has not been evaluated, the value of measuring the blood total cholesterol level is based on expert opinion (C Recommendation).
A stepped dietary intervention program, as recommended by several expert groups,<15,16,21> should be the initial therapeutic intervention for all men aged 30 to 59 years with a serum total cholesterol level above 6.85 mmol/L or an LDL-C level above 4.90 mmol/L. Evidence for this approach comes largely from randomized controlled trials of drug treatment of hypercholes-terolemia in which dietary therapy was used to identify men whose cholesterol levels were highly sensitive to diet.<25,28> People with a total cholesterol level of 6.20 to 6.85 mmol/L or an LDL-C level of 4.15 to 4.90 mmol/L may also benefit from a therapeutic diet, although supporting clinical data for this recommendation are weaker. Drug therapy should be considered only in those who have failed to respond to an adequate dietary trial. The exceptions to this approach are patients with a severely elevated serum total cholesterol level. To classify people as having an abnormal cholesterol level necessitating drug treatment, the average of three determinations of the total cholesterol level, or the average of two determinations of the total cholesterol level and at least one measurement of the LDL-C level must be calculated.
Currently the efficacy and short-term safety of drug treatment in the primary prevention of CHD have only been adequately determined in middle-aged men with hypercholesterolemia.<25,26,28-30> The Task Force therefore recommends this form of treatment in asymptomatic men aged 30 to 59 years with a mean serum total cholesterol level persistently above 6.85 mmol/L or an LDL-C level above 4.50 mmol/L after an adequate trial of intensive dietary therapy for at least 6 months. There is fair evidence to support this approach (B Recommendation) since the long-term safety of lipid-lowering drugs is unknown and a favourable impact on the overall death rate has not been established.
The decision to extend drug treatment to other people with a serum total cholesterol level above 6.85 mmol/L or an LDL-C level above 4.50 mmol/L must take the following issues into consideration. There are no efficacy trials involving elderly people, children or young adults, and the limited data available on adult women do not support therapeutic intervention.<30,32> The significant excess in the rate of death from noncardiac causes and the increase in total mortality noted in meta-analyses of primary prevention trials of cholesterol-lowering cannot be ignored in people at lower risk. Therefore, a grade C Recommendation (indicating insufficient evidence to include or exclude this approach) was considered appropriate until more compelling data become available.
The decision to offer drug treatment to people with a blood total cholesterol level of 6.85 mmol/L or less or an LDL-C level of 4.50 mmol/L or less is even more complex.
If it is assumed that the positive relation of blood total cholesterol to the risk of CHD is curvilinear in all segments of the population, the potential benefit of cholesterol lowering becomes progressively less with as the baseline cholesterol level gets lower. For example, among MRFIT participants a reduction of the total cholesterol level by 1.29 mmol/L from a baseline level of 7.76 mmol/L would hypothetically lower the absolute risk of CHD by 50%; however, the same reduction would lower the risk by 25% if the initial level was 6.47 mmol/L and by only 8% if the initial level was 5.17 mmol/L. For women and elderly people the benefits would be significantly less, since the gradient of CHD risk is much less steep.
A related issue is the number of people who must be treated (NNT) in order to prevent one critical event. From the LRC trial the number of patients with hypercholesterolemia needed to be treated for 5 years to prevent one CHD event (fatal or nonfatal) was 89. Because the rate of baseline events is lower among people at low and moderate risk the NNT would be substantially higher. If it is estimated that the average rate is 25% of that in the high-risk group and that the same relative risk reduction were achieved with therapy, a rough estimate of the NNT for 5 years to prevent one CHD event would be 356 in this group. Thus, the benefits of therapy would be marginal.
Given the uncertainties about the long-term safety of the available lipid-lowering drugs, the primary mode of treatment will likely be dietary. Apart from the aforementioned lack of data on the efficacy and safety of this intervention, credence to use data from drug studies to make dietary recommendations rests on the assumption that the mode of cholesterol lowering is less important than the actual blood total cholesterol level achieved. This premise, although reasonable, does not negate the possibility that the beneficial effects of drug treatment on the rate of CHD events may, at least in part, be independent of changes in the cholesterol level.
On the basis of these considerations the Task Force concluded that there was insufficient evidence to include or exclude a stepped fat-modified therapeutic diet to which a cholesterol-lowering drug would be added if the dietary response was deemed inadequate (grade C Recommendation).
Unanswered
Questions (Research Agenda)Future investigations toward the ideal diet should focus on ways to raise the plasma HDL-C concentration while reducing the LDL-C level, leaving the plasma triglyceride level unchanged and ensuring palatability. Furthermore, the adverse effects of labelling a person as having hypercholesterolemia, need to be systematically studied.
EvidenceThe Mantel-Haenszel method was used to obtain an overall estimate of benefits and risks of lowering the cholesterol level in primary prevention trials; the iterative method of Gart was used to compute 95% confidence intervals (CIs) for these estimates. The significance of the summary measure of treatment effect was judged by means of the uncorrected Mantel-Haenszel c 2 test with 1 degree of freedom; the consistency of effect was assessed by means of the Breslow-Day c 2 test. Only data for men were considered in the analysis, because data for women were limited and generally reported in summary form.
This review was initiated in June 1993
to update the Task Force report<1>
published in 1993
and recommendations were finalized by the Task Force in June 1994;
only new references or evidence that was crucial to the recommendations
are cited below and readers are referred to the 1993
publication for a full citation list.
Table
1: Main characteristics of trials of primary prevention of coronary heart
disease
| Study1 | LRC | HHS | WHO | UCS | EXCEL | LAVDS | MCS |
| No. of patients enrolled | 3,806 | 4,081 | 10,627 | 2,278 | 8,245 | 846 | 9,057 |
| % of male patients | 100 | 100 | 100 | 48 | 59 | 100 | 49 |
| Mean duration of trial, yr | 7.4 | 5.3 | 5.3 | 1.9 | 0.92 | 3.5 | 1.1 |
| Mean age of patients (and range), yr | 47.8
(35-59) |
47.3
(40-55) |
45.9
(30-59) |
53.9
(>17) |
55.8
(18-70) |
65.5
(50-89) |
--
(>17) |
| Baseline level, mmol/L
Total cholesterol LDL-C HDL-C Non-HDL-C Triglycerides |
7.50 5.60 1.20 -- 1.80 |
7.00 4.90 1.20 5.80 2.00 |
6.40 -- -- -- -- |
8.13 -- -- -- 2.38 |
6.67
4.65 1.16 -- 1.75 |
6.00 -- -- -- -- |
5.40 -- -- -- 1.30 |
| Intervention+ | Choles | Gemfib | Clofib | Colest | Lovast | D P:S | D P:S |
| Mean blood pressure, mmHg | 121/80 | 141/90 | 135/87 | 133/83 | (39.6)§ | 136/78 | -- |
| % of patients who smoked | 37 | 36 | 56 | -- | 18 | 70 | -- |
| People with diabetes included? | No | Yes | Yes | Yes | Yes | -- | -- |
| % of patients with prior MI | 0 | 0 | 0 | 6 | 18 | 7 | -- |
| Atherosclerotic heart disease or angina | 0 | 0 | 2 | 25 | 29 | 18 | -- |
1LRC = Lipid
Research Clinics Coronary Primary Prevention Trial (28), HHS = Helsinki
Heart Study (25), WHO = World Health Organization Clofibrate Trial (26,
29), UCS = Upjohn’s Colestipol Study (30), EXCEL = Expanded Clinical Evaluation
of Lovastatin (27), LAVDS = Los Angeles Veterans Diet Study (31), MCS =
Minnesota Coronary Survey (32); LDL-C = low-density lipoprotein cholesterol,
HDL-C = high density lipoprotein cholesterol, MI = myocardial infarction;
+ Choles = Cholestyramine, Gemfib = Gemfibrozil, Clofib = Clofibrate, Colest
= Colestipol, Lovast = Lovastatin, D P:S
= change in polyunsaturated : saturated fat ratio in diet; § % of
patients with hypertension.
Table
2: Change from baseline in blood lipid and lipoprotein levels after intervention
|
|
||||
|
|
||||
| Trial | Total Cholesterol | LDL-C | HDL-C | Triglycerides |
| LRC1 | -9 | -13 | +3 | +5 |
| HHS | -10 | -10 | +10 | -35 |
| WHO | -9 | - | - | - |
| UCS | -10 | - | - | - |
| EXCEL | -23 | -32 | +8 | -15 |
| LAVDS | -13 | - | - | - |
| MCS | -14 | - | - | -10 |
1 Abbreviations
the same as in Table 1
Table
3: Number of deaths by treatment status in asymptomatic adult men and women
|
|
||||||
|
|
||||||
| Trial | No. of patients | Cardiac | Noncardiac | Total | Cancer1 | Violence2 |
| Drug Therapy
LRC1 Treated Control |
1,900 |
44 |
27 |
71 |
15 |
4 |
| HHS
Treated Control |
2,030 |
19 |
23 |
42 |
11 |
4 |
| WHO
Treated Control |
5,296 |
77 |
104 |
181 |
55 |
25 |
| UCS
Treated Control |
1,129 |
16† |
17 |
48 |
2‡ |
0‡ |
| EXCEL
Treated Control |
1,663 |
3 |
0 |
3 |
0 |
0 |
| TOTAL
Treated Control |
12,018 |
159 |
171 |
345 |
83 |
33 |
| Dietary Therapy
LAVDS Treated Control |
422 |
50 |
127 |
177 |
20§ |
-- |
| MCS
Treated Control |
4,516 |
54 |
194 |
248 |
20 |
28 |
|
TOTAL
Treated Control |
4,938 |
104 |
321 |
425 |
40 |
28 |
|
|
||||||
| Overall Total
Treated Control |
16,956 |
263 |
492 |
770¶ |
123 |
61 |
1 Abbreviations
the same as in Table 1
2 Included in deaths from
noncardiac causes
† Denominator
includes only men without history of coronary heart disease at entry (372
in treatment group and 385 in control group)
‡ Only
data on men reported
§ A
total of 31 deaths from cancer in the treatment group and 17 in the control
group occurred during the diet phase; however, a follow-up report included
5 deaths from other types of cancer (1)
¶ Number
of deaths from cardiac and noncardiac causes does not add up to total because
in the UCS trial cardiac-related deaths in men with a history of coronary
heart disease were not included in the category of death from cardiac causes
Table
4: Effects of lowering the cholesterol level on cause-specific and overall
death rates in asymptomatic adult men and women Cause;
Odds Ratio (and 95% Confidence Interval)
| Trial | Cardiac | Noncardiac | Total | Cancer | Violence |
| Drug
Therapy |
|||||
| LRC | 0.72
(0.44-1.17) |
1.43
(0.79-2.27) |
0.95
(0.67-1.36) |
1.06
(0.50-2.27) |
2.75
(0.81-10.2) |
| HHS | 0.73
(0.35-1.53) |
1.34
(0.75-2.38) |
1.06
(0.68-1.66) |
0.99
(0.40-2.45) |
2.48
(0.72-9.39) |
| WHO | 1.18
(0.86-1.62) |
1.40
(1.07-1.82) |
1.31
(1.07-1.60) |
1.36
(0.94-1.96) |
0.95
(0.53-1.73) |
| UCS | 0.87
(0.40-1.89) |
1.04
(0.51-2.13) |
0.75
(0.47-1.18) |
0.981
(0.10-9.73) |
5.001
(.26->99) |
| EXCEL | 2.36
(0.69-9.76) |
2.78
(0.20->99) |
2.79
(0.82-11.4) |
2 | 2 |
| TOTAL | 1.00
(0.80-1.26) |
1.35
(1.10-1.65) |
1.15
(0.99-1.65) |
1.25
(0.92-1.68) |
1.41
(0.89-2.25) |
| Dietary
Therapy |
|||||
| LAVDS | 0.80
(0.50-1.26) |
1.06
(0.78-1.44) |
0.96
(0.73-1.28) |
1.70
(0.93-3.13) |
2 |
| MCS | 1.13
(0.77-1.65) |
1.07
(0.87-1.31) |
1.08
(0.90-1.30) |
1.14
(0.60-2.17) |
1.17
(0.69-2.00) |
| TOTAL | 0.97
(0.73-1.30) |
1.07
(0.90-1.26) |
1.05
(0.90-1.22) |
1.41
(0.92-2.18) |
1.17
(0.69-2.00) |
| Overall
Total |
0.99
(0.83-1.18) |
1.17
(1.03-1.34) |
1.10
(0.99-1.22) |
1.30
(1.02-1.66) |
1.30
(0.92-1.84) |
Link to Structured Abstract of this review
Link to Summary Tableof this review
Link to Selected References list of this review
Reprinted in modified format by the Canadian
Task Force on Preventive Health Care
with permission.
For any technical issues please contact: webmaster@ctfphc.org
Original Copyright
© 1994 Minister of Supply and Services Canada.
Last modified March 27, 1998.
