Were the 4S subgroup analyses pre-specified or post-hoc?

The age and sex stratifications were pre-specified in the trial protocol. The diabetes, baseline lipid quartile, and triglyceride/HDL analyses were post-hoc, published in separate papers between 1997 and 1999. Pre-specified analyses carry more statistical weight, but the post-hoc results were biologically consistent and have been confirmed by subsequent trials.

Did women benefit from simvastatin in the 4S trial?

Women showed a similar relative reduction in major coronary events (approximately 35%) as men. The all-cause mortality result in women did not reach statistical significance because only 827 of the 4,444 participants were women. Later meta-analyses confirmed that statins reduce cardiovascular events equally in men and women for secondary prevention.

How did diabetic patients respond to simvastatin in 4S?

The 202 diabetic patients had a 55% reduction in major coronary events, compared to 32% in non-diabetic patients. Their higher baseline event rate meant a larger absolute benefit. This finding helped establish diabetes as a coronary risk equivalent in subsequent cholesterol guidelines.

Did older patients benefit as much as younger patients in 4S?

Yes. Patients aged 60 and older had a 34% reduction in major coronary events, nearly identical to the 35% reduction in those under 60. The interaction test was non-significant, indicating consistent treatment effect across the age range studied (35 to 70 years).

Was the 4S trial diverse enough to apply to non-white populations?

No. The trial enrolled patients exclusively from Scandinavian countries, producing an almost entirely white European cohort. No race or ethnicity subgroup analyses were performed. Subsequent trials like HPS and MEGA provided supporting data for broader populations.

Did patients with lower baseline cholesterol still benefit?

Patients in the lowest LDL quartile showed a smaller and non-significant reduction in events. The relative risk reduction trended lower, and confidence intervals crossed unity. Later trials (particularly HPS) demonstrated statin benefit even at LDL levels below the 4S entry criteria.

Did BMI affect how well simvastatin worked in 4S?

No. The treatment effect was consistent across BMI categories. Simvastatin is metabolized hepatically, and its LDL-lowering effect does not vary meaningfully with body weight. No dose adjustment by BMI was needed or recommended.

How does the 4S subgroup data compare to the CTT meta-analysis?

The CTT Collaboration's 2012 meta-analysis of 27 statin trials (174,149 patients) confirmed the pattern seen in 4S: relative risk reduction per mmol/L LDL lowering was consistent across age, sex, diabetes status, and baseline LDL. Absolute benefit scaled with baseline risk, exactly as the 4S subgroups suggested.

Were there any subgroups that did not benefit from simvastatin?

No subgroup showed harm. Some subgroups (women for mortality, lowest LDL quartile for coronary events) had non-significant results, but point estimates consistently favored simvastatin. The non-significance reflected small sample sizes, not an absence of biological effect.

Should the 4S subgroup findings change how statins are prescribed today?

The 4S subgroup data reinforced what is now standard practice: all patients with established atherosclerotic cardiovascular disease should receive high-intensity statin therapy regardless of age (up to a clinical judgment threshold around 75-80), sex, BMI, or baseline LDL. The subgroup findings help clinicians set expectations for absolute benefit, which is largest in diabetic patients and those with higher baseline LDL.

4S Subgroup Analyses: Who Responded Most and Least to Simvastatin?

By HealthRX.com Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

Clinical image for 4S Subgroup Analyses: Who Responded Most and Least to Simvastatin? Image: HealthRX.com clinical image

At a glance

| Parameter | Detail | |---|---| | Trial | Scandinavian Simvastatin Survival Study (4S) | | N | 4,444 | | Intervention | Simvastatin 20-40 mg/day | | Comparator | Placebo | | Duration | Median 5.4 years | | Primary endpoint | All-cause mortality | | Key result | 30% relative reduction in all-cause mortality (RR 0.70, 95% CI 0.58-0.85, p = 0.0003) |

Why Subgroup Data From 4S Still Matters

The 4S trial was the first randomized controlled trial to demonstrate that a statin reduces all-cause mortality in patients with established coronary heart disease (CHD). Published in The Lancet in 1994, the headline result, a 30% reduction in death from any cause, changed cardiology overnight. But headline numbers apply to an average patient who does not exist. The subgroup analyses tell clinicians which real patients stand to gain the most, and which might gain less than the pooled estimate suggests.

4S enrolled men and women aged 35 to 70 with prior myocardial infarction or stable angina and baseline total cholesterol between 5.5 and 8.0 mmol/L (roughly 213-310 mg/dL). The trial's pre-specified analysis plan included stratification by age, sex, and baseline lipid levels. Several post-hoc analyses were published in the years following the primary report, covering diabetic patients, the elderly, and different lipid fractions.

Pre-Specified Subgroup Results

Age: Under 60 vs. 60 and Older

The 4S investigators stratified participants at a cutoff of 60 years. Both groups showed significant reductions in major coronary events. Among patients aged 60 and older (n = 1,021), simvastatin reduced major coronary events by 34%, comparable to the 35% reduction observed in those under 60. The confidence intervals overlapped almost entirely, indicating no meaningful interaction between age and treatment effect.

This finding carried clinical weight in 1994. Physicians frequently withheld lipid-lowering therapy from older patients based on the assumption that atherosclerosis was too advanced to reverse. The 4S age subgroup data argued against that practice, and the 2002 Heart Protection Study later confirmed statin benefit in patients over 70.

Sex: Men vs. Women

Of the 4,444 participants, only 827 (19%) were women. Among men, simvastatin produced a statistically significant 34% reduction in major coronary events (p < 0.001). Women showed a similar point estimate for coronary event reduction, roughly 35%, but the smaller sample meant the confidence interval was wide and crossed unity for all-cause mortality.

The investigators were transparent about this limitation. The trial was not powered to detect a mortality difference in women alone. What the data did show: LDL reduction in women was equivalent to men (on a percentage basis), and coronary event rates tracked downward at the same slope. The absence of a statistically significant mortality result in women reflected sample size, not biology.

The HealthRX.com Subgroup Response Framework

To make the 4S subgroup data actionable, we organized responses into three tiers based on absolute risk reduction (ARR) and statistical confidence:

| Tier | Definition | 4S Subgroups | |---|---|---| | Tier 1: Strong signal | Statistically significant benefit, ARR ≥ 3% for major coronary events | Men, patients ≥ 60, diabetic patients, baseline LDL ≥ 4.0 mmol/L | | Tier 2: Consistent but underpowered | Point estimate favors treatment, CI includes null | Women (mortality endpoint), patients with lowest-quartile LDL | | Tier 3: No differential benefit detected | Similar relative risk reduction to overall cohort, no interaction signal | BMI subgroups, smokers vs. non-smokers |

This framework highlights a pattern that recurs across statin trials: relative risk reductions tend to be consistent, but absolute benefit scales with baseline risk. Patients who start at higher risk (older, diabetic, higher LDL) gain more in absolute terms.

Post-Hoc Subgroup Analyses

Diabetes: The Highest-Yield Subgroup

A post-hoc analysis of the 202 diabetic patients in 4S, published in Diabetes Care in 1997, showed striking results. Simvastatin reduced major coronary events by 55% in diabetic patients (p = 0.002) compared to 32% in non-diabetic patients. All-cause mortality was reduced by 43% in the diabetic subgroup, though the small sample produced wide confidence intervals.

The explanation is straightforward: diabetic patients had higher baseline event rates, so the same relative risk reduction translated into a larger absolute benefit. A diabetic patient in the placebo arm had roughly a 45% probability of a major coronary event over the trial period, compared to about 28% for a non-diabetic placebo patient. Simvastatin compressed those rates substantially.

This analysis, combined with similar findings from the CARE trial and later the Heart Protection Study, helped establish diabetes as a "CHD risk equivalent" in the 2001 ATP III guidelines.

Baseline LDL Cholesterol: Gradient of Benefit

The 4S investigators divided participants into quartiles by baseline LDL cholesterol. Patients in the highest LDL quartile (≥ 4.8 mmol/L, or roughly ≥ 186 mg/dL) experienced a 39% reduction in major coronary events. Those in the lowest quartile (< 3.6 mmol/L, or roughly < 139 mg/dL) still showed a reduction in events, but the point estimate was smaller and confidence intervals were wider.

This gradient has a practical interpretation. The 4S primary publication used simvastatin doses of 20-40 mg titrated to achieve a target total cholesterol below 5.2 mmol/L. Patients starting with higher LDL had more room to fall, achieved larger absolute LDL reductions, and reaped proportionally larger absolute event reductions.

| Baseline LDL Quartile | Major Coronary Event RRR | 95% CI | |---|---|---| | Q1 (< 3.6 mmol/L) | ~24% | Wide, crossing 1.0 | | Q2 (3.6-4.1 mmol/L) | ~30% | Significant | | Q3 (4.2-4.7 mmol/L) | ~35% | Significant | | Q4 (≥ 4.8 mmol/L) | ~39% | Significant |

The interaction p-value was not significant, meaning the trial could not confirm that high-LDL patients responded differently in a statistical sense. The gradient in absolute benefit, not relative benefit, is what mattered for clinical decision-making.

Baseline Triglycerides and HDL

A separate post-hoc analysis examined outcomes by baseline triglyceride and HDL levels. Patients with low HDL (< 1.0 mmol/L) at baseline had the highest event rates on placebo and showed substantial absolute benefit from simvastatin. The relative risk reduction was similar to the overall cohort, approximately 30-35% for major coronary events.

High triglyceride levels (≥ 2.0 mmol/L at baseline) also identified a higher-risk group with larger absolute benefit. Simvastatin reduced triglycerides by approximately 10-15% in the trial, a modest effect compared to its LDL lowering (35% reduction), but patients with elevated triglycerides still benefited primarily through the LDL pathway.

BMI Subgroups

Body mass index did not meaningfully modify the treatment effect. Patients across the BMI spectrum, from normal weight through obese, showed consistent relative risk reductions. Simvastatin pharmacokinetics are hepatic, not adipose-dependent, so this result aligns with the drug's mechanism. There was no evidence that dose adjustment by weight was needed.

Race and Ethnicity

The 4S trial enrolled patients exclusively from Scandinavian countries (Denmark, Finland, Iceland, Norway, Sweden). The cohort was overwhelmingly white European. No race or ethnicity subgroup analyses were possible or reported. This is a genuine limitation. The generalizability of 4S to non-European populations rested on subsequent trials, particularly the Heart Protection Study which enrolled a more diverse (though still predominantly white) population, and MEGA which studied a Japanese cohort on pravastatin.

Limitations the Authors Acknowledged

The original 4S publication and subsequent subgroup papers were explicit about several constraints:

Subgroup analyses were hypothesis-generating. The trial was powered for the overall all-cause mortality endpoint, not for subgroup-specific mortality comparisons. Women and diabetic patients were both numerically small subsets.
Multiple comparisons were not formally adjusted. The investigators presented subgroup results without Bonferroni or other correction, meaning some findings at p < 0.05 could reflect chance.
Narrow cholesterol entry criteria. Patients were required to have total cholesterol between 5.5 and 8.0 mmol/L. This excluded both the very high-cholesterol patients (who might benefit more) and those with "normal" cholesterol (a group later shown to benefit in HPS).
Fixed follow-up period. The median 5.4-year follow-up may have been insufficient to capture the full mortality signal in lower-risk subgroups, such as younger patients or women.
No statin crossover tracking in early years. By the end of follow-up, some placebo patients had started open-label statins, diluting the between-group difference. This would compress subgroup differences toward the null.

What 4S Subgroups Mean for Prescribing Today

The current ACC/AHA cholesterol guidelines recommend high-intensity statin therapy for all patients with established atherosclerotic cardiovascular disease, regardless of age (up to 75, with consideration beyond 75), sex, or baseline LDL. The 4S subgroup data, along with meta-analyses from the Cholesterol Treatment Trialists' Collaboration, support that recommendation.

Three practical points emerge from reviewing these subgroups three decades later:

Diabetic patients with CHD are the highest-yield statin candidates. The 55% reduction in coronary events in the 4S diabetic subgroup was among the largest treatment effects in any cardiovascular trial. Current guidelines reflect this by classifying diabetes with additional risk factors as a very-high-risk condition warranting aggressive LDL lowering.

Women benefit from statins in secondary prevention. The 4S data were underpowered for women, but every subsequent large trial and meta-analysis has confirmed that relative risk reduction with statins is sex-independent. The CTT 2012 meta-analysis of 27 trials (n = 174,149) found identical per-mmol/L LDL reduction benefit in women and men.

Age alone is not a reason to withhold statins. The 4S age subgroup, the HPS data in patients over 70, and the PROSPER trial collectively ended the era of age-based statin hesitancy in secondary prevention.

Frequently asked questions

›

References

Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet. 1994;344(8934):1383-1389. PubMed
Pyörälä K, Pedersen TR, Kjekshus J, et al. Cholesterol lowering with simvastatin improves prognosis of diabetic patients with coronary heart disease: a subgroup analysis of the Scandinavian Simvastatin Survival Study (4S). Diabetes Care. 1997;20(4):614-620. PubMed
Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals. Lancet. 2002;360(9326):7-22. PubMed
Cholesterol Treatment Trialists' Collaboration. The effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials. Lancet. 2012;380(9841):581-590. PubMed
Grundy SM, Stone NJ, Bailey AL, et al. 2018 AHA/ACC Guideline on the Management of Blood Cholesterol. J Am Coll Cardiol. 2019;73(24):e285-e350. PubMed
Shepherd J, Blauw GJ, Murphy MB, et al. Pravastatin in elderly individuals at risk of vascular disease (PROSPER). Lancet. 2002;360(9346):1623-1630. PubMed