If-then plans can help turn diet and exercise intentions into action. Research also shows limits: old habits can remain, and weight loss is not assured.
In the morning, you see your exercise clothes and decide to go for a walk. By evening, you are home and the clothes are still where you left them. The intention made it through the working day; the walk did not. On days like this, it is tempting to assume that a stronger resolution is the missing ingredient.
A meta-analysis of physical activity intentions and behavior found that 47.6% of people who intended to be active did not translate that intention into action. The analysis included 25 independent samples and 29,600 participants. This does not mean that half of everyone in the studies failed to exercise. The figure describes the gap within the group who intended to act, a distinction that matters when reading an apparently dramatic statistic.[14]
Implementation intentions address that gap by connecting a situation with a response: “If this happens, then I will do that.” Much of the research asks whether people carry out eating or physical activity behaviors. Writing a plan and losing weight are different outcomes. The evidence gives reasons to try this planning method, but also reasons to keep its promise specific.[3][5][6]
Implementation intentions link goal-directed actions to cues
An implementation intention specifies when, where, and how you will carry out an action that serves a goal. “I want to exercise more” states a goal. “When I finish lunch and get up from my seat, I will go outside for a walk” connects an identifiable moment with something you can do. A review of fat-intake reduction describes implementation intentions as action plans subordinate to a goal: the goal supplies the direction, while the plan gives an occasion for acting.[10]
The everyday examples in this article are suggestions for applying that format. Their individual timings and actions have not been tested as separate interventions.
Consider “I will exercise when I get home.” There are still unanswered questions. Will you start at the door, after dinner, or after finishing household tasks? What activity do you mean? “Before I take off my shoes at the door, I will put out my walking clothes” describes a more recognizable scene. It lets you check what you expect to happen, rather than simply whether the sentence sounds determined.
A broad meta-analysis combining 642 independent tests found effects on cognitive, emotional, and behavioral outcomes, with effect sizes ranging from d=0.27 to 0.66. Effects were larger when plans used a contingent if-then format, participants were highly motivated to pursue the goal, and the plans were rehearsed. These tests extended beyond food and exercise. The range describes results across many tasks, rather than an expected amount of weight loss from planning.[6]
Rehearsal need not become an elaborate ritual. With the lunchtime example, you could mentally follow the sequence of finishing your meal, standing up, and heading toward the exit. The part you are rehearsing is the transition from noticing the cue to beginning the action.[6]
Before polishing the wording, ask whether the action is something you actually want to do. Someone who likes walking but does not want to run cannot acquire a desire to run merely by borrowing an impressive running plan. The stronger effects among highly motivated participants are a reason to keep the plan connected to the person’s own goal. A clear sentence cannot supply every condition needed to carry it out.[6]
This article focuses on the entrance to an action: which ordinary situation will prompt you to begin? The companion article on mindful eating examines attention during eating. The articles on social support and financial incentives consider help from other people and rewards. Those are separate questions. Here, the useful starting point is a condition you can recognize in your own day and a response you are willing to make.
初日の汗は、予約した人だけが持ち帰れる。
Adding healthy foods often works better than banning foods
A 2011 systematic review and meta-analysis included 23 studies of eating behavior. Plans aimed at adding healthy foods, such as fruit, had an effect size of d=0.51. Plans aimed at reducing unhealthy eating, such as unhealthy snacks, had a smaller effect size of d=0.29. The same planning approach produced different average results depending on what people were trying to change.[7]
The review also identified a weakness in some of the favorable evidence. In studies promoting healthy eating, less adequate control conditions may have inflated effect sizes. If the planning group receives more careful support than the comparison group, the entire difference cannot necessarily be attributed to the if-then sentence. A positive average and a concern about the comparison can both be true.[7]
A 2019 meta-analysis found a similar direction of difference. Across 70 interventions involving 9,689 participants, the effect size was d=0.33 for increasing healthy eating and d=0.18 for reducing unhealthy eating. These values differ from the earlier review, but both analyses found larger effects for adding healthy behaviors than for reducing unhealthy ones. They describe eating outcomes, not equivalent changes in body weight.[5][7]
The 2019 analysis also examined why results varied. For healthy eating, effects tended to be larger in younger participants. Yet student samples had smaller effects than nonstudent samples. Initial training and interventions delivered offline were positively associated with effect size. Younger age and student status did not point in the same direction, so “this works better for young people” would lose part of the picture. Both the participants and the way the intervention was delivered mattered.[5]
For daily life, that pattern suggests trying an action that adds something, rather than relying only on a prohibition. Instead of writing “I will not eat sweets,” you might write “When I choose lunch, I will also look for a dish containing vegetables.” The proposed action links a choice point to a food you want to include. It is an application of the planning format, not proof that this particular wording changes the total amount you eat or your weight.[5][7]
Reducing a food component is not automatically ineffective. A 2017 meta-analysis of fat-intake reduction combined 12 studies and 3,323 participants, reporting a moderate effect of d=0.488. Effects were larger among men and when there was no monitoring during the intervention. Changing the target behavior and the participants changes what the evidence says about an avoidance goal. The results do not justify dismissing every plan framed around reducing intake.[10]
The symbol d denotes a standardized effect size: a way of expressing a difference in the measured outcome on a common scale. It is not a unit of food intake or body weight. A value of d=0.33 does not tell you how much fruit someone ate, how many snacks they avoided, or how much weight they lost. To understand the number, return to the outcome measured in the paper. A standardized comparison helps summarize a result, but it does not turn unlike outcomes into the same event.[5][7][10]
You can preserve that distinction in your own assessment. “Did I choose the food when the planned situation occurred?” and “What happened to my weight?” are separate questions. This lets you acknowledge a change in choice while leaving the weight question open. It also makes feedback more concrete: a lunch selection can be checked against the plan, whereas a promise about the scale goes beyond the eating-behavior result.[3][5]
Exercise effects are modest and depend on the support
A 2013 meta-analysis reviewed 26 independent studies of implementation intentions and physical activity. The effect size immediately after the intervention was 0.31, with a 95% confidence interval of 0.11–0.51. At follow-up it was 0.24, with a 95% confidence interval of 0.13–0.35. The authors described the effects as small to medium. These are average differences supporting activity, with a more restrained size than a claim that planning transforms everyone’s routine.[2]
For these physical activity estimates, the 95% confidence intervals show the uncertainty around the immediate and follow-up effects; both size and precision matter when judging the planning benefit. In this review, effects were larger when plans included management of barriers. The planning question therefore extended beyond the ideal occasion to situations in which something interfered. A plan can specify the opening for action and also address a likely obstacle.[2]
Suppose you planned to walk on the way home, but overtime made that impossible. One possible response is: “If I work late, I will review today’s plan and decide the next occasion for the walk.” This is a suggested way to handle a barrier by choosing another opportunity. On a day when you need rest, rest can also be part of the schedule rather than something a rigid sentence forbids.
A 2018 systematic review of randomized trials in adults included 13 trials, with participants aged 18–76. The studies included people with conditions such as obesity, diabetes, and physical inactivity. In the subgroup of interventions that reinforced the implementation-intention strategy, the effect size was 0.25, with a 95% confidence interval of 0.05–0.45, favoring the intervention. The qualification is central: the reported subgroup result concerned reinforced planning, rather than every instance of writing a plan.[4]
A 2022 review narrowed its focus to people with chronic conditions living in the community. Across 54 studies, it found a small physical activity effect: a standardized mean difference of 0.24, with a 95% confidence interval of 0.10–0.39. Dietary behavior also improved by a small amount, with a standardized mean difference of −0.25 and a 95% confidence interval of −0.34 to −0.15. The negative sign in the dietary result indicated improvement in the measured outcome; it was not an amount of weight lost.[1]
A standardized mean difference expresses results measured on different scales using a shared metric. Within that review, physical activity effects were larger among men, older adults, overweight or obese patients with complications, and when a healthcare provider delivered the intervention.[1]
The 2013 review also reported larger effects among student and clinical samples. That student finding points in a different direction from the healthy-eating analysis. It is another reason to read the behavior and the population together.[1][2][5]
Taken together, these reviews make planning easier to understand as an aid. There are research clues for revisiting occasions that worked and barriers that interfered, rather than assuming the first written version finishes the job. The practical question is what kind of support fits the particular action.[1][2][4]
Numbers: implementation intentions and the action gap
47.6%: the share of people with physical activity intentions who did not act on them, across 25 samples and 29,600 participants.[14]
d=0.33 and 0.18: effects for increasing healthy eating and reducing unhealthy eating, respectively, across 70 interventions.[5]
0.31, 95% confidence interval 0.11–0.51: the immediate physical activity effect in a review of 26 studies.[2]
709 adults: a trial found no overall effect of forming plans on physical activity or BMI.[3]
0.31, 95% confidence interval 0.14–0.48: the difference in physical activity habit strength after habit-formation interventions, rather than a measure of exercise volume or weight loss.[15]
初日の汗は、予約した人だけが持ち帰れる。
1
2
