The N of 1: Why You Are the Most Important Study You Will Ever Run

Sports science produces a lot of useful findings, and we cite them regularly. But there is a limitation built into almost every study in the exercise physiology literature that rarely gets acknowledged when the findings get distilled into training advice: the results describe what happened to the average participant across the sample, and you are not the average participant. You are one specific person with a specific training history, specific genetics, specific sleep patterns, specific stress load, specific gut behaviour at kilometre 60 of a mountain race, and a specific set of responses to the interventions being studied. The average finding may apply to you closely, loosely, or not at all.

This is not an argument against research. It is an argument for treating research as a starting point rather than a prescription, and for developing the habit of running deliberate experiments on yourself to find out what actually works for you specifically. In clinical research this is called an n=1 (or N of 1) study - a single-subject trial in which the subject and the experimenter are the same person. In the context of endurance training, it is one of the most underused tools available.

What Makes a Good N of 1 Study

The difference between a useful self-experiment and just trying something and seeing what happens is structure. Without structure, the signal gets lost in the noise of everything else that is changing simultaneously. With structure, even a small number of data points can be genuinely informative.

A good n=1 study changes one variable at a time. This is harder than it sounds in endurance training, where training load, sleep, nutrition, stress, terrain, and weather are all fluctuating simultaneously. It requires consciously holding as much as possible constant while the variable of interest changes - which is impractical in many real-world training contexts, but more achievable than most runners assume if they think about it in advance.

A good n=1 study has a clear outcome measure. Not "I want to see how I feel," but a specific, trackable signal: performance on a consistent test effort, RPE at a fixed pace on a known route, body weight at a consistent time of day, time to fatigue at a given intensity, or subjective recovery score against a consistent scale. The outcome measure does not need to be precise in the laboratory sense. It needs to be consistent enough that meaningful changes are distinguishable from day-to-day noise.

A good n=1 experiment runs long enough for the adaptation in question to express itself. Testing whether adding one more strength session per week improves performance over two weeks tells you almost nothing, because two weeks is not enough time for the relevant neuromuscular adaptations to develop. Testing the same thing over eight weeks, with consistent outcome measures at the beginning and end, gives you something you can actually learn from.

And a good n=1 experiment is honest about confounders. If the experiment period included a period of high work stress, disrupted sleep, or a minor illness, those need to be noted and accounted for when interpreting the results. The value of the self-experiment is in the learning, and learning requires honest interpretation, not just confirmation of what you hoped would happen.

Applied to Strength Training

The published research on strength training for endurance runners is consistent enough in its broad findings - heavy compound lifting improves running economy, posterior chain training reduces injury risk, plyometrics improve rate of force development - that the question of whether to do it is largely settled. The more useful questions for any individual runner are ones the research cannot answer on their behalf.

How many strength sessions per week can you absorb without compromising running quality? For some runners, two sessions per week sits easily alongside their training load with no degradation in running performance. For others, a second session bleeds into the following day's run quality in ways that undermine the whole enterprise. The research cannot tell you which category you are in. A structured eight-week experiment - two sessions per week for four weeks, one session per week for four weeks, with consistent run quality measures across both periods - can.

What is your personal response to heavy load versus moderate load? The literature generally supports heavier loading for neuromuscular adaptation, but some runners find that heavy squats and deadlifts create enough systemic fatigue that their running suffers in the days following, while others find no such interference. Testing this systematically over successive training blocks, with a consistent measure of running quality in the 48 hours following each strength session, gives you a useful individual data point that generalises from no published study.

Which specific exercises produce the most obvious transfer to your running? This is genuinely individual. Some runners notice a clear improvement in climb efficiency when Romanian deadlifts are in the programme. Others notice it from single-leg work. Some find that heavy calf isometrics produce the most obvious change in how their feet feel on technical terrain. Paying deliberate attention to the relationship between what you are lifting and what you are noticing on the trails, over a sustained period with one exercise change at a time, is a form of n=1 experimentation that costs nothing and teaches you a great deal.

Applied to Nutrition

Endurance nutrition is an area where population-level research is particularly poor at predicting individual responses, for reasons that are increasingly well understood. Gut microbiome composition, the rate at which individuals oxidise fat versus carbohydrate at various intensities, and the degree to which specific foods cause GI distress under the stress of sustained running effort all vary enormously between individuals. The average finding from a nutrition study is often less useful here than in almost any other area of training science.

The most practically significant questions a trail runner can investigate through n=1 experimentation in nutrition are typically these.

What is your personal fat oxidation capacity at race pace? Fat oxidation rates during exercise vary by roughly threefold between individuals at similar fitness levels, and the practical implication - how much exogenous carbohydrate you need per hour at a given race intensity - varies accordingly. A runner who oxidises fat efficiently at threshold pace needs substantially less gel intake per hour than one who is more carbohydrate-dependent at the same intensity. Finding out which end of that distribution you sit on requires testing: long efforts at controlled intensity with varying carbohydrate intake, with energy levels and GI comfort as the outcome measures.

What are your gut tolerances under race conditions? GI distress is the most common cause of DNF at the ultramarathon distance, and it is almost entirely individual. What works for one runner - gel type, solid food timing, caffeine dose, electrolyte formulation - is demonstrably different from what works for another. The only way to know what your gut will tolerate after six hours of running on heat-stressed mountain terrain is to test it in conditions that approximate race intensity and duration as closely as possible. Race day should never be the first time a nutrition strategy is tested. It should be the fifth or sixth.

How does your body respond to training fasted versus fed? The research on fasted training is genuinely equivocal at the population level, with some studies supporting improved fat adaptation and others finding no meaningful benefit and some cost to training quality. At the individual level, some runners train well fasted and others find their session quality degrades substantially without pre-session fuel. Testing this over a consistent block - same session type, same time of day, alternating fed and fasted across weeks, with RPE and session quality as outcome measures - gives you a personal answer that the literature cannot.

Applied to Ultrarunning

Ultrarunning is the domain where n=1 thinking is most critical and most obviously underused, because ultrarunning contains so many individual variables that population-level research is often nearly irrelevant.

Pacing strategy over mountainous terrain is one. Research on optimal pacing strategy in ultramarathons exists, but the findings are so dependent on terrain, elevation profile, course conditions, individual aerobic efficiency at different gradients, and fatigue patterns that applying a general pacing recommendation to a specific runner on a specific course is a crude approximation at best. What every experienced ultrarunner develops over time is an n=1 dataset of their own pacing responses - how fast they can climb a given gradient before crossing into unsustainable effort, how their pace holds across the back half of a 50K, what their late-race degradation curve looks like - that is far more useful for race planning than anything in the literature.

Sleep deprivation management is another. Events lasting more than 24 hours require runners to manage cognitive and physical function under significant sleep deprivation, and individual responses to sleep deprivation vary enormously. Some runners function reasonably well through a night section and recover quickly at dawn. Others experience severe cognitive degradation that makes navigation unreliable and pace judgement inconsistent. The only way to know your response is to train in conditions that replicate it - overnight training runs, late start times, or deliberately shortened sleep before long efforts - and to observe how your specific physiology handles it.

Altitude response is a third. Runners training and racing above 2,500 metres show highly individual patterns of acclimatisation, performance drop, and recovery. The average acclimatisation curve from the research is a reasonable starting point, but the range of individual responses is wide enough that some runners perform surprisingly well at altitude from day one while others are significantly impaired for longer than the population data would predict. If your racing involves altitude, your own acclimatisation history is your most important data source.

The Discipline of Keeping Records

N of 1 study only accumulates into useful knowledge if it is recorded. The human memory is a poor experimental instrument. It is biased toward confirming existing beliefs, it compresses and distorts timelines, and it consistently underweights data points that contradict the preferred narrative. Writing things down changes this.

A training log that goes beyond activity data - that captures RPE, sleep quality, nutrition, mood, and notable observations about how the body responded to specific sessions - is the raw material of useful self-experimentation. Over months and years, patterns emerge from this data that no coach and no published study could identify, because the data is specific to you. You begin to see that your running quality degrades reliably after two consecutive days of poor sleep but not after one. You notice that your leg turnover on technical terrain is noticeably better in the weeks when you are consistent with single-leg strength work. You observe that your GI system tolerates one brand of gels well at race pace and another causes problems consistently from hour three onwards.

None of this knowledge is generalisable to other runners. All of it is directly actionable for you.

The Relationship Between Research and Self-Experimentation

N of 1 thinking does not replace sports science. It operates alongside it. The research tells you which variables are worth experimenting with, what the plausible mechanisms are, and what range of responses other people have had. The self-experiment tells you where you sit within that range and how the mechanisms express themselves in your specific physiology and context.

The most effective runner-athletes tend to be the ones who read the research, take its broad findings seriously, and then test its specific implications against their own data rather than accepting the average finding as their personal prescription. They treat their training as a long-running experiment in which they are simultaneously the subject and the investigator, and they update their practice when the evidence - internal and external - changes.

That orientation is available to every runner at any level of experience or performance. It requires curiosity, consistency, and honesty more than any specialist knowledge. And the return on those investments compounds over time in a way that following someone else's programme never quite does.

For more on the science underpinning the training variables most worth experimenting with, read our posts on VO2max, RER, 1RM - Three Tests Every Endurance Athlete Should Get and CGM in Running and Endurance Sports: Promising Tool or Overhyped Fuel Sensor?.