What are safe exercise “dosing” steps to avoid post-exertional crashes?

The short version: there is no evidence-based "dosing" protocol for exercise in people prone to post-exertional crashes. The studies that exist test pacing, not dosing, and the one randomized trial of graded exercise in this population found no benefit and more deterioration.

The evidence you need does not exist. No study has tested a specific exercise "dose" — measured in minutes, heart rate, or intensity — against post-exertional crashes in people with ME/CFS or long COVID who already have post-exertional malaise (PEM). What exists is the opposite: a 2024 randomized trial of 241 children with ME/CFS found that graded exercise therapy produced no significant improvement in physical function compared to activity management, and the graded exercise group had more deterioration events (27 vs 17) [13]. A reanalysis of the adult PACE trial found that the graded exercise group had more serious adverse events (8% vs 4%) and no improvement in fitness, work days lost, or disability benefits [15].

What the evidence does support is pacing, not dosing. A 2011 consensus document defines pacing as staying within your individual energy envelope — recognizing early signs of PEM and switching activities or resting before you crash [19]. In non-randomized data, people who stayed within their energy envelope showed a 16% improvement in fatigue and 37.6% improvement in physical function, compared to 2.7% and 6% in those who pushed past it [19]. The one randomized trial of symptom-contingent pacing (12 weeks) showed a 30% drop in fatigue scores versus 15% in controls, but had no follow-up or objective activity monitoring [19].

The case report of a 27-year-old runner with post-COVID POTS who improved over a year used a graduated program starting from recumbent exercise, progressing to seated, then standing and walking [3]. That is a single patient, not a protocol.

My call: do not follow a fixed exercise "dose" — that is what the failed trials tested. Use symptom-contingent pacing: stop before you crash, switch muscle groups, rest at the first sign of PEM. The evidence for any other approach is absent or negative.

Confidence: moderate — the evidence against graded exercise is solid, but the evidence for pacing comes from small, short, non-randomized studies.