HOW TO STAY MOTIVATED TO EXERCISE AFTER 50: WHAT THE RESEARCH MEASURED

How do you stay motivated to exercise after 50? The research has more to say about this than most people expect. Behavior scientists have measured how long a new routine takes to feel automatic, whether a written plan changes what people do, whether tracking helps, and whether training with a group beats training alone. Some of those findings are steady. Others disagree with each other, and a few point the opposite way from the usual advice. This guide sets out what the studies measured, next to the daily system I use from my own books.

Where I'm coming from: I'm a Marine veteran and an endurance athlete with more than 300 races behind me, and I wrote The Roman Protocol, The Spartan Protocol and The Discipline Protocol. I'm not a psychologist, a physician or a researcher. What follows is what published studies report, and what my books do. When something is a study finding and when it's a choice from my own system, I'll say which.

Motivation Isn't One Thing

"Motivation" covers several different reasons for showing up, and the research separates them.

• Why you train matters. A 2012 systematic review of 66 studies (Teixeira and colleagues) found consistent support for a link between more self-chosen, or "autonomous," motivation and exercise. Training because you value what it does for you was more strongly tied to starting, while training because you enjoy it was more strongly tied to sticking with it long term. Most of the studies were descriptive rather than experiments, which the authors note.
• How often older adults stick with programs. A 2014 systematic review of nine studies in older people (Picorelli and colleagues) found that 65 to 86% of participants completed their exercise programs, and attendance ran from 58 to 77% of available sessions. Adherence was generally higher in supervised programs. Better adherence went with fewer health conditions, better self-rated health, better physical ability, better cognitive ability and fewer depressive symptoms.

Put plainly: in these studies, many people who started a program skipped sessions, and some stopped. That is common. The rest of this guide covers the tools researchers have tested for closing that gap.

How Long It Takes a Habit to Form

The idea that a habit forms in 21 days is everywhere. The best-known study of real-world habit formation found something slower and more variable.

• The Lally study. In 2010, Phillippa Lally and colleagues at University College London had 96 volunteers pick one eating, drinking or activity behavior and do it daily in the same context, such as "after breakfast," for 12 weeks. They rated how automatic it felt each day. 82 gave enough data to analyze, and the researchers' curve fit well for 39 of them.
• The timeline. The time it took people to reach 95% of their own peak automaticity ranged from 18 to 254 days. UCL's summary of the study reported an average of 66 days among the people whose data the model fit well.
• Missing a day. Missing one opportunity to do the behavior did not materially affect the habit formation process. Doing the behavior more consistently went with a better fit to the habit curve.
• Exercise specifically. In a 12-week study of 111 new gym members (Kaushal and Rhodes, 2015), exercising at least four times a week for about six weeks was the minimum the authors found for an exercise habit to form. Consistency, keeping the routine simple, the environment and how enjoyable people found exercise all predicted stronger habits.

The Lally study was small, and a walk before dinner is not the same as a structured training plan. What it does show is a long and wide range. To see what 66 days, 18 days and 254 days look like from your own start date, use the free Habit Timeline Calculator, which also walks you through writing an if-then plan. More in The 21-Day Habit Myth.

If-Then Plans: What the Trials Found

Researchers call a specific if-then plan an implementation intention: "If it's 5:45 a.m. on Monday, then I walk out the back door for my run." A goal says what you want. An implementation intention says when, where and how.

• Peng and colleagues (2022): 41 randomized trials with 5,439 participants. In the 35 trials pooled, planning interventions increased physical activity compared with control groups (standardized mean difference 0.35). In their subgroup analyses, planning worked better in patients, in men, with action planning, when delivered face to face, and when it was reinforced during follow-up.
• Silva and colleagues (2018): 13 randomized trials in adults aged 18 to 76, many with conditions such as heart disease, obesity or diabetes. In trials where the if-then planning was reinforced, the pooled effect was 0.25 in favor of planning.
• Kompf (2020): a systematic review of 24 studies asking who if-then plans work for. The answer was people who already intend to be active and have strong self-efficacy, meaning confidence they can do it. A plan helped turn an existing intention into action. It did not create the intention.
• Bélanger-Gravel and colleagues (2013): a randomized trial in 101 inactive older adults with obesity. Adding if-then plans to a standard program produced no main effect on any outcome. Steps per day did change differently over time between groups, and at six months the improvement in physical activity was greater in the planning group (effect size 0.59). It had little effect on weight beyond the standard program.

These are small-to-moderate effects, and they depended on context. I could not find a trial testing if-then plans in masters athletes. The studies are mostly in less active adults or people with health conditions.

Writing It Down: Self-Monitoring

Tracking what you do is one of the most studied behavior-change techniques. In older adults, the findings disagree, and the disagreement is worth seeing.

• Michie and colleagues (2009): 122 evaluations with 44,747 participants of programs to increase physical activity and healthy eating. Self-monitoring explained more of the difference between studies than any other technique. Programs that combined self-monitoring with at least one other technique from the same theory of self-regulation (control theory) had larger effects than the rest (0.42 versus 0.26).
• Harkin and colleagues (2016): 138 randomized studies with 19,951 participants, across many kinds of goals. Interventions that got people to monitor their progress improved goal attainment (effect size 0.40). The effect was larger when progress was reported to others or made public, and when it was physically recorded.
• French and colleagues (2014), adults 60 and over: 24 studies in community-dwelling older adults. Interventions overall raised physical activity a little (effect size 0.14). But self-regulation techniques, including setting behavioral goals, prompting self-monitoring, planning for relapses and giving feedback on performance, were associated with lower self-efficacy and lower physical activity. The authors concluded that techniques that work for younger adults may not work for older adults.

So the general evidence favors writing things down, while the one review limited to adults 60 and over found the opposite pattern for several self-regulation techniques. That review compared program features across studies, which shows association rather than cause. It is still the closest evidence to this audience, and it's the reason I keep my own tracking short.

Training With Others

"Find a training partner" is standard advice. The largest recent review measured it directly.

• Kritz and colleagues (2026): 71 studies comparing individual and group-based physical activity programs. Group programs showed a small advantage for physical activity that was not statistically significant (effect size 0.086, confidence interval crossing zero). Psychosocial and health outcomes were also not significantly different. Functional outcomes such as strength and flexibility favored groups after outlier removal (0.164). The authors concluded that both individual and group approaches, in person or online, may be effective.
• Supervision: in the Picorelli review above, adherence was generally higher in supervised programs.

In The Spartan Protocol, the chapter "The Shield of the Man Beside You" builds a mutual accountability arrangement between two people who hold the same standard for each other. That's the book's approach, not a measured result. This post sets it against the Kritz findings.