Wilhelmsen 1975.

Methods	Study design: Single centre RCT Country: Sweden Dates patients recruited: 1968‐1970 Maximum follow up: 5 years Patients randomised on discharge.
Participants	Inclusion criteria: All patients born in 1913 or later who suffered a MI during the period 1968‐1970 and were discharged alive from the hospital. Exclusion criteria: none described. N Randomised: total: 315; intervention: 158; comparator: 157 Diagnosis (% of pts): MI: 100% Age (years): intervention: 50.6; comparator: 50.6 Percentage male: intervention: 87%; comparator: 90% Ethnicity: NR
Interventions	Intervention: The training programme started 3 months after the MI. The programme at the hospital consisted of three supervised half‐hour training sessions a week. It included dynamic work, such as calisthenics, cycling, and running in an interval programme with individualised intensity. If a patient found it difficult to attend the hospital for training then individualised programmes were developed for training at home or in the workplace. Components: exercise. Setting: supervised in a centre. Aerobic exercise: Modality: e.g. cycling, running. Length of session: 1/2 hour. Frequency: three times a week. Intensity: 144 ± 18 beats/min; 80% of their heart rate increasing capacity (if no sign of cardiac limitation); 136 ± 19 beats/min in mean highest training heart rate (if limited by angina pectoris). Resistance training included? Calisthenics. Total duration: NR ‐ see notes below. Co‐interventions: At discharge from hospital all patients were given general recommendations about gradually increasing physical activity during the convalescence period. Comparator: usual care. Co‐interventions: as above.
Outcomes	Mortality, re‐infarction.
Source of funding	NR
Conflicts of interest	NR
Notes	1 year post‐MI, only 39% of those who started training were training at the hospital. A further 21% trained at home or at work.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"By the use of a random number table the patients were allocated..."
Allocation concealment (selection bias)	Unclear risk	Not reported.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"The exercise test 1 yr after the MI followed the same protocol but was conducted by another physician, who did not know if the patients belonged to the experimental or the control group".
Incomplete outcome data (attrition bias) All outcomes	Low risk	No losses to follow up for clinical events.
Selective reporting (reporting bias)	Unclear risk	Outcomes to be collected were not clearly described in the methods.
Groups balanced at baseline	High risk	"Despite the fact that the randomization was done according to a random number table, there were slightly more patients with a history of hypertension, previous MI, treatment with digitalis prior to the MI, and also with high physical activity during leisure time in the control group than in the experimental group. Dyspnea during the acute phase was more common in the experimental group than in the control group. Body weight and systolic blood pressure were slightly lower in the control group than in the experimental group 3 months after the MI".
Intention‐to‐treat analysis conducted	High risk	No.
Groups received same treatment (apart from the intervention)	Low risk	“... the two patient groups were treated in a uniform manner by three physicians, and all possible precautions were taken to standardize the follow‐up and treatment with the exception of the training programme.”