Training Strain
Also known as: Foster Strain, Weekly Strain Index, Monotony × Load
A composite weekly load metric that multiplies weekly AU by training monotony, producing a number that rewards productive volume and penalises uniform distribution. Strain is the second half of Foster's monitoring pair (monotony × weekly load = strain). Where weekly AU alone says 'how much did you do', strain says 'how much did you do, weighted by how blurred together it was'.
Formula
strain = weeklyAU × monotony
Where weeklyAU = sum of dailyAU across 7 days, and monotony = mean(dailyAU) ÷ SD(dailyAU). Useful interpretive bands from Foster's team-sport work:
strain < 4,000 — sustainable training week
strain 4,000-6,000 — productive accumulation, monitor for drift
strain > 6,000 — Foster's original threshold for elevated overreaching / illness risk
A week of 2,450 AU at monotony 1.43 ≈ 3,500 strain. The same 2,450 AU at monotony 2.4 ≈ 5,900 strain — the same volume reads as nearly twice the physiological cost because the body never gets a clean recovery day inside the week.Example
Two athletes complete identical 2,800 AU weeks. Athlete A: 4 hard sessions (700 AU each) + 3 rest days. SD ≈ 360, monotony ≈ 1.11, strain ≈ 3,100. Sustainable. Athlete B: 7 days at 400 AU each, no rest day. SD ≈ 0 (in practice ~50), monotony ≈ 8, strain ≈ 22,000+. The math screams; the body screams a week later. Even if you allow that B's reality has some inter-day variance, monotony lands above 2.5 and strain pushes past 7,000 — well over Foster's threshold.
How Afitpilot Uses This
Afitpilot's primary weekly load readout is weekly AU split by modality (strength / endurance / mixed), with the chronic / acute load chart providing the trend view. Strain is on the same roadmap candidate as monotony — both are cheap to compute from existing sRPE inputs and would surface a useful distinction the weekly AU alone misses: same volume, different recovery cost. When surfaced, strain would not auto-modify the plan, but a sustained 2+ week drift above Foster's threshold band would be flagged in the coach drawer alongside Hooper drift and ACWR — the same multi-signal pattern that flags the early overreaching continuum.
Strain in practice
| Who / Context | Value | Note |
|---|---|---|
| Foster's elevated-risk threshold | strain > 6,000 | Derived from team-sport athletes; use as a starting reference |
| Recreational productive week | strain 2,000-4,000 | Typical for adult athletes with at least one rest day |
| Marathon build week | strain 5,000-7,500 | High weekly volume × moderate monotony — sustainable for 4-8 weeks then deload |
| Polarized-training week | strain 1,500-3,500 at the same volume | Strong hard/easy contrast keeps monotony low, strain low |
| Strain after returning from a planned rest week | Often spikes 50-100% | Expected transient — read the 2-3 week trend, not the spike |
| Sport with highest baseline strain in Foster cohorts | Endurance running and cross-country skiing | Multi-day uniform-load weeks are the modality, not a flaw |
Known Limitations
- •Strain inherits every limitation of monotony — modality-blind, threshold-dependent, undefined at zero variance — and stacks them on top of weekly AU's own limitations as a load metric for strength work.
- •The 6,000 threshold from Foster's team-sport data is widely cited and rarely re-validated. Endurance athletes in deliberate high-volume base blocks routinely exceed it without injury or illness; strength athletes rarely reach it. Treat the threshold as a heuristic, not a stop sign.
- •Strain spikes hard when athletes complete a planned rest week as actual full rest (low load, low monotony) followed by a return-to-training week (moderate load, high monotony from contrast with the rest week). The metric reads the transition as risky when it's structurally productive.
- •The multiplicative structure means small monotony increases produce large strain increases at high weekly volumes. An athlete at 5,000 weekly AU with monotony rising 1.5 → 1.8 sees strain go 7,500 → 9,000 — a 20% jump from a small distribution change.
- •Strain is a population-level monitoring tool, not an individual predictor. Foster himself emphasises trend over absolute value: rising strain over 2-3 weeks is more actionable than any single week's number.
Science Context
Strain emerged from Foster's 1998 framework as the actionable composite of monotony and weekly AU — the metric athletes and coaches actually look at when monotony is the underlying input. The original validation was on team-sport athletes (soccer, basketball, swimming) over multi-week observation windows, with strain above ~6,000 associated with elevated reports of upper-respiratory infection and overreaching symptoms. Subsequent work (Foster et al. 2001; later replications in cycling, rugby, and rowing) supports the qualitative pattern — rising strain tracks subjective complaints — without strongly validating any single numerical threshold. Modern training-monitoring reviews (Halson 2014; Bourdon et al. 2017 consensus statement) include strain alongside ACWR, monotony, and Hooper as one of the cheap-to-compute scalars worth tracking, with the standard caveat that no single number diagnoses overreaching and the multi-signal trend across weeks is the actual signal.