Tip · Stress & mind

Gratitude journaling

Q: How many things should I write down, and how long does it take?

The studies vary widely and have never systematically compared the number. What was tested is the duration of the intervention: in a randomized experiment with 119 participants, two weeks of daily practice were enough to measurably change sleep quality, optimizm and diastolic blood pressure. The widespread figure of three things comes from practice, not from a dose comparison. The claim that five minutes are enough is also plausible, but untested.

Q: Does it really work, or is it placebo?

It depends on what it is compared with. Against doing nothing and against merely filling in questionnaires, it works slightly but measurably. Against another, similarly plausible writing exercise, it no longer works at all for well-being; there the effect is d = −0.03. What reliably rises is the measured feeling of gratitude itself, with d = 0.46 even against active comparison conditions.

Q: Does gratitude journaling help you fall asleep?

The indications are there, but not strong. In the systematic review of physical endpoints, subjective sleep quality improved in 5 of 8 studies. The randomized experiment with an active control group also found better sleep quality after two weeks. The likely pathway runs through thoughts before falling asleep: anyone who looks back on the day with gratitude has less negative rumination on their mind.

Q: Does gratitude lower my cortisol?

In the only study in this literature that measured cortisol with an active control group, there was no association whatsoever, even though well-being, sleep quality and diastolic blood pressure improved. So there are no reliable data for an effect of gratitude journaling on cortisol. That is not an argument against the exercise, but against this one justification.

Q: Can I use it to treat depression?

No. The meta-analysis that tested exactly this found a small effect on symptoms of depression and anxiety across 27 studies with 3,675 participants and explicitly recommends choosing approaches with stronger evidence for complaints that require treatment. The exercise can accompany treatment; it does not replace it.

Q: Does it work the same in every culture?

The largest analysis to date, across 28 countries, found significant differences between countries but could not identify any factor that explains these differences. The authors note that more primary research is needed on this. The pooled effect across all countries remains small and positive, at g = 0.19.

Randomized trials show small effects on mood – mostly shortly after the exercise, and much smaller against active comparison groups.

Writing down three things you are grateful for: the exercise takes five minutes and by now has a very large evidence base. The effect is small, stable and depends heavily on what it is compared with. This page shows what the exercise reliably affects — and where the story about the brain and cortisol does not hold up.

In short

Gratitude journaling increases well-being measurably, but only slightly: g = 0.19 across 145 papers with 24,804 participants from 28 countries, robust to publication bias. The clearest increase is in the feeling of gratitude itself, with d = 0.46 even against an alternative activity. On symptoms of depression and anxiety the exercise works with g = −0.29 and still holds at g = −0.23 at follow-up. The catch lies in the comparison group: against a psychologically active comparison condition, the effect on well-being is d = −0.03.

What is behind it

No physiological mechanism for gratitude journaling has been described. A psychological one has: writing things down systematically directs attention to positive events and thereby shifts the balance of what is remembered at the end of the day. The most stable expression of this is the rise in positive affect — the authors of the classic series of experiments themselves call it the most robust finding, and in the pooled analysis the effect was larger when well-being was measured as positive affect.

For sleep, there is a tested intermediate step. In a cross-sectional study with 401 people, gratitude predicted better subjective sleep quality, longer sleep duration, shorter sleep onset latency and less daytime impairment, fully mediated by more positive and fewer negative thoughts before falling asleep, independently of the Big Five traits. A cross-sectional study does not prove causation, but it is the only mechanism that has been specifically tested here.

The staircase of comparison groups

The same literature, analyzed in stages by control condition, produces a clear staircase downwards: against a measurement-only control d = 0.31 for psychological well-being, against any alternative activity d = 0.17, against an activity matched for effort d = 0.14, against a psychologically active comparison condition d = −0.03.

What is notable is where the effect stays large: not in well-being, but in the measured feeling of gratitude itself, with d = 0.46 against an alternative activity. Above all, then, the exercise makes people more grateful. For anxiety, no significant effect was found in any constellation.

What the body notices

The systematic review of physical endpoints shows how thin this area is. Of 1,433 articles found, 19 met the criteria. Subjective sleep quality improved in 5 of 8 studies, self-reported physical symptoms in 2 of 8, inflammatory markers in 1 of 2. Blood pressure, blood glucose control, asthma control and eating behavior each improved in 1 of 1 study — each of these endpoints was therefore examined by exactly one paper, and the majority showed concerns about risk of bias.

What is well supported

The core is a preregistered meta-analysis of 145 papers, 163 samples, 727 effect sizes and 24,804 participants from 28 countries. The pooled effect on well-being is g = 0.19 (95% CI 0.15 to 0.22), robust to publication bias. That is small, but with an evidence base of this size it can hardly be argued away. Added to this is the finding that is easily overlooked: on the feeling of gratitude itself, the exercise works with d = 0.46 even against an alternative activity. And the effect on symptoms of depression and anxiety holds: g = −0.29 after the intervention, g = −0.23 afterwards.

What the studies show

The largest analysis: small but stable

Preregistered meta-analysis of 145 papers, 163 samples, 727 effect sizes, 24,804 participants from 28 countries. Well-being g = 0.19 (0.15 to 0.22), robust to publication bias. The moderator analysis found significant differences between countries, but no moderator that explains them. The effect was larger when well-being was measured as positive affect and when several types of exercise were combined.

The effect depends on the comparison condition

Series of meta-analyses on the same literature, staged by control condition (k = 4–18, N = 395–1,755). Against measurement-only control, well-being d = 0.31 (0.04–0.58). Against an alternative activity: gratitude d = 0.46 (0.27–0.64), well-being d = 0.17, anxiety d = 0.11 and thus not significant. Against an effort-matched activity d = 0.14, against a psychologically active comparison condition d = −0.03.

As self-help for depression and anxiety

Meta-analysis of 27 studies with 3,675 participants. After the intervention g = −0.29 (SE 0.06; p < 0.01), at follow-up g = −0.23 (SE 0.06; p < 0.01), at both time points with larger effects against waitlist than against active control conditions. The authors explicitly call the effects moderate and, for symptoms requiring treatment, recommend approaches with a better evidence base.

Two weeks, measured with blood pressure and sleep

Randomized experiment with 119 young women and three conditions: two weeks of gratitude intervention, an active control writing down everyday events, and no treatment. Hedonic well-being, optimizm and sleep quality rose, and diastolic blood pressure fell. The improvements correlated with better sleep quality and lower blood pressure, but explicitly not with cortisol.

Where the data stop

Against a psychologically active comparison condition, that is, another exercise that seems similarly plausible, the effect on well-being is zero: d = −0.03 (−0.13 to 0.07; k = 9). No significant effect on anxiety was found in any constellation. The overall effect remains small at g = 0.19, and no moderator could explain the differences between countries. Cortisol did not move in the only paper that measured it with an active control group.

The statement that gratitude changes neural patterns within four to eight weeks does not hold up against its source. There, people who were starting psychotherapy anyway wrote gratitude letters; measurement took place three months later in a group comparison, and what was found was stronger neural modulation in the medial prefrontal cortex. That is a functional difference in people in therapy, not a structural change in healthy people.

How to do it

The classic exercise is listing things you are grateful for, on a weekly or daily basis. In a randomized experiment with 119 young women, two weeks were enough for measurable changes in sleep quality, optimizm and diastolic blood pressure, and the effect lasted through follow-up.

The meta-analysis identified variety as an effective feature: the effect was larger when different types of gratitude exercises were combined, that is, not only listing but also, for example, gratitude letters. Two claims found everywhere, by contrast, are untested. The number of entries has never been systematically compared; the widespread figure of three things comes from practice. And no paper has systematically varied whether people write in the morning or in the evening.

Safety

Gratitude journaling is among the lowest-risk measures of all: no substance, no physical strain, no intervention in the body, and the literature reports no physical adverse effects. One limitation is nonetheless important: in depression or an anxiety disorder that requires treatment, the exercise is no substitute for treatment. The meta-analysis that asked exactly this question arrives at g = −0.29 and explicitly recommends that those affected choose approaches with stronger empirical support. Anyone suffering from persistent low mood, lack of drive or anxiety belongs in the hands of a doctor or psychotherapist; the exercise can stand alongside that, not in front of it. The second point concerns expectations: anyone who does it because it is pleasant is right. Anyone who expects a transformation will be disappointed.

BK-Score Thin human evidence

Human evidence		5
Mechanism		4
Safety data		10
Hype gap		5
Track record of use		9

Randomized trials show small effects on mood and well-being, often only shortly after the intervention. More recent reviews with active control groups find considerably less than the early papers.

The score rates the state of knowledge, not the effect. “Safety data 9” means well studied – not harmless.
Subjective assessment by Biohacking Kompakt based on published scoring rules – not a scientific rating and not a medical recommendation. Rules and all ratings (German)

Frequently asked questions about gratitude journaling

How many things should I write down, and how long does it take?