To assess whether low serum cholesterol concentration increases mortality from any cause.

The authors do not specify which sources were searched, or how the search was performed.

Study designs of evaluations included in the review

1. Cohort studies of serum cholesterol concentration and IHD, which were limited to studies with more than 350 IHD events.

2. International prospective community studies in which serum cholesterol concentrations, and mortality from IHD and other causes, were measured. Only 2 of the original 3 international studies had data on causes of death other than IHD.

3. RCTs of reduction in cholesterol concentration (by drugs, diet, or ileal bypass surgery) and IHD events (deaths and nonfatal infarcts). Trials had to record at least 1 death and document a reduction in serum cholesterol concentration of at least 1%.

Information in this section has been taken from a companion article (see Other Publications of Related Interest).

Specific interventions included in the review

Reduction of serum cholesterol concentration through use of drugs, diet or ileal bypass.

Participants included in the review

Participants varied according to the study design, and included those with and without previous history of ischaemic heart disease (IHD). The populations in the randomised controlled trials (RCTs) were mostly male.

Outcomes assessed in the review

Excess-cause specific mortality associated with low or lowered serum cholesterol concentration. Deaths from causes other than IHD, i.e. circulatory diseases other than IHD, cancer, accidents, or suicide and other causes.

How were decisions on the relevance of primary studies made?

The authors do not state how the papers were selected for the review, or how many of the authors performed the selection.

RCTs: diagnoses were made without knowledge of treatment allocation,; data were analysed on the basis of intention-to-treat, and follow-up at the end of the trials was 99.5%. The authors do not state how the papers were assessed for validity, or how many of the authors performed the validity assessment.

The authors do not state how the data were extracted for the review, or how many of the authors performed the data extraction.

How were the studies combined?

1. Cohort studies: these were divided into 2 groups, 4 studies recruiting employed males and 6 recruiting from the general population (community cohorts).

2. International studies: data were combined, and the ratio of the mortality in communities with a mean serum cholesterol concentration below and above 5 mmol/l were calculated.

3. RCTs: original data on causes of death other than IHD were combined and analysed on an intention to treat basis. Logistic regression analysis was used to combine the odds ratios (ORs) from each trial to obtain a summary OR for the effect of cholesterol reduction on each of 4 specified categories of deaths (circulatory disease other than IHD, cancer, accidents and suicides, and other causes). The OR from each trial was weighted by its reduction in cholesterol concentration.

How were differences between studies investigated?

Logistic regression analysis, weighted by reduction in cholesterol concentration, was used to investigate study differences.

Ten cohort studies: 494,804 men and 18,811 events recorded; 54,832 women and 4,097 events recorded.

Two international studies: number of participants not given.

Twenty-eight published RCTs that recruited 52,350 men and recorded 4,241 events.

1. Cohort Studies.

Relative risk in subgroup with lowest cholesterol concentration, compared with the rest of the cohort:

Circulatory disease (not IHD): employed men 0.92 (95% confidence interval, CI: 0.80, 1.05), cohort from community setting 1.12 (95% CI: 0.98, 1.28).

Cancer: employed men 1.00 (95% CI: 0.91, 1.10), cohort from community setting 1.23 (95% CI: 1.17, 1.30).

Accidents and suicides: employed men 0.95 (95% CI: 0.70, 1.30), cohort from community setting 1.29 (95% CI: 1.13, 1.47).

Other diseases: employed men 1.08 (95% CI: 0.97, 1.21), cohort from community setting 1.62 (95% CI: 1.47, 1.78).

All causes except IHD: employed men 1.00 (95% CI: 0.94, 1.06), cohort from community setting 1.20 (95% CI: 1.15, 1.24).

2. International Studies.

Relative risk in communities with mean cholesterol concentration below 5 mmol/l, compared with mean cholesterol concentration greater than 5 mmol/l:

Circulatory disease (not IHD): 7 countries 1.85 (p<0.001), international comparison 1.37 (p>0.05).

Cancer: 7 countries 0.98 (p>0.05), international comparison 0.88 (p>0.05).

Accident and suicide: 7 countries 0.87 (p>0.05), international comparison 0.63 (p>0.05).

Other: 7 countries (not applicable), international comparison 0.92 (p>0.05).

All causes except IHD: 7 countries 1.16 (p<0.05), international comparison 0.98 (p>0.05).

3. RCTs.

OR for reduction in cholesterol concentration of 0.6 mmol/l:

Circulatory disease (not IHD): all trials 1.00 (95% CI: 0.82, 1.21).

Cancer: all trials1.07 (95% CI: 0.90, 1.26).

Accidents and suicide: all trials 1.17 (95% CI: 0.90, 1.52).

Other: all trials 1.07 (95% CI: 0.92, 1.26).

All causes except IHD: all trials 1.07 (95% CI: 0.97, 1.18).

Relative risk of death in treated or control groups per 0.6 mmol/l (10%) reduction in cholesterol:

Drug trials: all causes except IHD 1.20 (95% CI: 1.02, 1.40).

Dietary trials: all causes except IHD 1.01 (95% CI: 0.88, 1.15).

Participants without IHD: all causes 1.06 (95% CI: 0.97, 1.17).

Participants with IHD: all causes 0.90 (95% CI: 0.84, 0.97).

There is no evidence that low or reduced serum cholesterol concentration increases mortality from any cause other than haemorrhagic stroke. This risk affects only those people with a very low serum cholesterol concentration, and even in these people the risk will be outweighed by the benefits associated with the low risk of IHD.

This is a methodologically rigorous review. No search strategy is listed and therefore the completeness of the search cannot be ascertained. The authors' conclusions are not entirely consistent with the data presented. Readers are encouraged to refer to the letters to the editor regarding this paper (see Other Publications of Related Interest), which highlight several key areas of concern relating to the use of cholesterol-lowering agents and the assessment of initial risk for patients in such assessments.

Law MR, Wald NH, Thompson SG. By how much and how quickly does reduction in serum cholesterol concentration lower risk of ischaemic heart disease? BMJ 1994;308:367-72.

These additional published commentaries may also be of interest. Ischaemic heart disease and cholesterol [letters]. BMJ 1994;308:1038-41.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.