The fall half-marathon taper — what 2 weeks of reduced volume should look like

The Bosquet 2007 taper meta-analysis

The single most-cited piece of research on tapering is the Bosquet, Montpetit, Arvisais and Mujika meta-analysis published in Medicine and Science in Sports and Exercise (Bosquet et al. 2007, PMID 17762369). The team pooled 27 studies that met inclusion criteria for randomised, controlled, or repeated-measures taper designs in endurance athletes. The headline finding: performance improvements were maximised when volume was reduced by 41 to 60 percent over a taper lasting 8 to 14 days, with training frequency and intensity held constant.

The effect size on race performance was not trivial. Bosquet's group calculated a mean improvement of roughly 1.96 percent on time-trial outcomes — which for a 1:45 half marathon translates to about two minutes shaved off. That is the difference between a personal best and a near-miss. For amateurs who have spent four months building mileage, two minutes is not a rounding error. It is the entire return on training.

What the meta-analysis ruled out is just as useful. Tapers shorter than 7 days did not produce the same fitness retention. Tapers longer than 21 days began to erode aerobic capacity. Reducing intensity alongside volume — the instinctive amateur move — blunted the performance benefit. The Mujika group has reproduced these findings in subsequent work (Mujika 2010), and the practical guidance has barely changed.

Volume vs intensity vs frequency

The three levers of training load are volume (total kilometres), intensity (pace zones), and frequency (number of sessions per week). Tapering is not about doing less of everything. It is about doing less of one thing — volume — while protecting the other two.

Volume should drop sharply. A runner who has been logging 60 km per week should be running 25 to 35 km during the second-last week and slightly less during race week. The drop is front-loaded; the bulk of the cut happens in week one of the taper, not week two.

Intensity stays. The threshold runs, tempo intervals, and race-pace strides that built fitness should not disappear. They get shorter, but the pace and effort stay sharp. A 10-km tempo at half-marathon pace might become a 5-km tempo. A 6 × 1 km interval session might become 4 × 800 m. The body needs to be reminded what race pace feels like — going long and slow during taper week leaves the legs sluggish on race morning.

Frequency holds. If you have been running five or six days a week, keep running five or six days a week. Cutting sessions creates a detraining signal the body interprets as injury or illness. A 5-km easy shake-out the day before a race is more useful than a rest day for most runners, even if it feels counterintuitive.

The 2-week structure (week 1 then race week)

Week one of the taper, with a Sunday goal race in view: Monday — easy 6 km. Tuesday — 4 × 1 km at 5K pace with full recovery, total 8 km. Wednesday — easy 6 km or cross-train. Thursday — 8 km with the middle 4 km at half-marathon pace. Friday — rest or 30 minutes easy. Saturday — 6 km easy with 4 × 100 m strides. Sunday — 10 km easy long run. Total: roughly 40 to 45 km, down from a peak of 60 to 65.

Race week: Monday — easy 5 km with strides. Tuesday — 5 km with 3 × 800 m at race pace. Wednesday — 4 km easy. Thursday — rest or 25 minutes easy. Friday — 4 km easy with 4 × 100 m strides. Saturday — 3 km shake-out or rest. Sunday — race. Total non-race volume: 20 to 25 km. The frequency holds. The intensity points stay. Only the mileage drops.

Carbohydrate loading 36-48 hours out

The classical Sherman/Costill carbohydrate-loading protocols of the 1980s have been refined. Burke et al. (2011) published a comprehensive review in the Journal of Sports Sciences concluding that 36 to 48 hours of high carbohydrate intake (10 to 12 g per kg of body weight per day) was sufficient to maximise muscle glycogen for endurance events lasting longer than 90 minutes. Longer loading periods produced no additional benefit and frequently caused gastrointestinal distress.

For a 70-kg runner, that is roughly 700 to 840 g of carbohydrate per day for two days. This is a lot of food. Practically, that looks like oatmeal at breakfast, a large pasta lunch, rice and chicken at dinner, plus bread, fruit, and sports drinks throughout the day. Fibre intake should drop sharply during this window to reduce gut volume on race morning.

The night-before pasta dinner is a tradition with mixed evidence. The Burke review and the Pfeiffer (2012) work on race-day GI distress both suggest that the meal eaten 12 to 14 hours before the race matters far less than the cumulative intake over the prior 36 hours. A moderate dinner of familiar food is better than a heavy load of unfamiliar pasta.

Race-pace strides during taper

Strides — short, fast accelerations of 80 to 120 metres — are the most underused tool in the amateur taper. The Bosquet finding that intensity must be maintained does not mean every session needs to be hard. It means the neuromuscular system needs occasional reminders of what fast feels like.

Two to four sets of 100-metre strides at slightly faster than 5K pace, with full recovery between each, take about 8 minutes to complete. They produce no measurable fatigue. They preserve running economy, the metric most likely to deteriorate during a sedentary taper. Daniels (2014) recommended strides three to four times per week during taper for exactly this reason.

Sleep + nervousness during the taper

Taper insomnia is real. The combination of reduced training load, race anxiety, and pre-race carbohydrate loading frequently disrupts sleep in the final 72 hours. Fullagar et al. (2015) in Sports Medicine reviewed the evidence on sleep and athletic performance and concluded that the cumulative sleep of the prior week matters more than the night immediately before competition. One bad night of sleep before a race rarely impairs performance if the prior six nights have been adequate.

The implication is practical. Bank sleep in the two weeks before the race. Aim for 8 hours per night during taper. If race-eve sleep is broken, do not catastrophise — the night before the night before matters more.

Fall weather variables (wind, temp, hydration)

Southern Ontario fall races run in conditions that swing wildly. A September race in Toronto can hit 24°C with humidity. A late-October race in Collingwood can start at 4°C with wind off Georgian Bay. The Tucker and Dugas heat work (2012) and the more recent Périard reviews suggest that race-pace adjustments of 2 to 4 percent are warranted for every 5°C above 15°C.

Hydration during taper does not require electrolyte loading. Sodium intake of roughly 3 to 5 g per day during the loading window is sufficient for most runners. Drinking to thirst, urine pale yellow on race morning — these are the markers. The Noakes et al. (2005) work on exercise-associated hyponatraemia made the case strongly against overdrinking, and that case has held up.

Race-morning checklist

Wake 3 hours before the start. Eat a familiar breakfast of 100 to 150 g of carbohydrate (oatmeal with banana and honey, or bagel with jam). Coffee if you normally drink coffee. Sip water but do not chug. 90 minutes out, light walk for 10 minutes. 30 minutes out, 5 minutes of jogging with 3 × 100 m strides. Race clothing tested in training, never new. Vaseline or anti-chafe balm on every contact point. Gels and water plan rehearsed in long runs.

Practical takeaways

• Cut volume by 41 to 60 percent over 8 to 14 days; hold intensity and frequency constant.
• Strides of 80 to 120 metres preserve running economy during the volume drop.
• Carbohydrate load over 36 to 48 hours, not a week; 10 to 12 g per kg per day.
• Bank sleep across the two-week window; one bad night before the race is not a problem if the prior week was solid.
• Race-morning routine should be entirely rehearsed in training; nothing new on race day.

Extended takeaways

The amateur instinct to do less and rest more during the final two weeks is the wrong instinct. The body interprets sudden, sweeping volume cuts as detraining. The Bosquet meta-analysis is unambiguous on this: the runners who held their training frequency and intensity, while cutting only the duration of their easy runs and the volume of their hard sessions, outperformed the runners who tapered comprehensively. The two-week structure outlined here is conservative — many elite programs run a 10-day taper with even sharper volume cuts — but it is robust to the small mistakes amateurs make.

The carbohydrate loading window has tightened over the past two decades. The week-long depletion-then-load protocols of the 1980s are no longer recommended; the gut distress they cause outweighs the marginal glycogen benefit. Thirty-six to 48 hours of high-carbohydrate intake, with fibre held low, achieves the same muscle-glycogen result without the GI risk. This is one of the few areas of endurance science where the recommendation has become simpler over time.

Race-day execution is where taper science either pays off or doesn't. The runners who arrive at the start line with rehearsed nutrition, tested kit, banked sleep, and a clear pace plan tend to run within 1 to 2 percent of their physiological ceiling. The runners who improvise — new gels, new shoes, new caffeine doses — frequently underperform their training by 3 to 5 percent. The taper itself is necessary but not sufficient. The discipline of rehearsing race-day variables in the long runs of weeks 8 through 12 of the build is what makes the taper investment pay out.

Frequently asked questions

Can I run a tune-up race two weeks before the half?

Yes, and the evidence supports it. A 10K race two weeks out functions as a high-intensity session and provides race-condition rehearsal. Treat it as a session, not a peak performance. Recovery from a 10K takes 3 to 4 days, which fits the taper window.

What if I get sick during the taper?

Reduce training to easy aerobic work and prioritise sleep. The Walsh et al. (2011) immune-function review found that upper-respiratory infections during taper are common and usually self-limiting. If symptoms are above the neck only, light running is fine. Below the neck (chest, lungs, gut) means rest.

Should I cut caffeine before race day?

Habitual caffeine users do not need to cut and reload. The Pickering and Kiely (2018) review found no consistent performance benefit to caffeine withdrawal before competition. Maintain your normal intake.

Do I need to gain weight during taper?

A small gain of 0.5 to 1.5 kg is expected and largely water bound to glycogen. Three grams of water bind to each gram of stored glycogen. This is loaded mass, not fat gain. Scale weight on race morning is meaningless.

What about the day after the race?

Walk for 30 minutes. Do not run. Refuel with mixed protein and carbohydrate within 2 hours of finishing. Sleep is the recovery tool that matters most in the 72 hours post-race.

References