Self-esteem research shows a consistent link between self-worth and mental health, relationships, and life outcomes, though most evidence is correlational rather than causal. Major longitudinal and meta-analytic studies find self-esteem tends to rise from adolescence into midlife and then declines slightly. High self-esteem differs from narcissism. This self-esteem research summary organizes the core findings and their limits.
- What self-esteem research actually measures
- Five takeaway findings from self-esteem research
- Self-esteem studies at a glance
- How self-esteem changes across the lifespan
- Self-esteem vs narcissism: what studies do and don't show
- What the intervention evidence says so far
- How to use this summary and its sources
- FAQ
- Sources
What self-esteem research actually measures
When teachers, parents, and students ask what does research say about self-esteem, they usually mean three questions: how it is defined, what it predicts, and whether it can change.
According to the American Psychological Association, self-esteem is a person’s overall evaluation of their own worth. Researchers often separate global self-esteem, a broad judgment about oneself, from domain-specific self-esteem, which applies to areas such as school, work, appearance, or relationships. They also distinguish trait self-esteem, a fairly stable pattern over time, from state self-esteem, which shifts with a situation or event.
Self-concept is the wider set of beliefs a person holds about themselves; self-efficacy is narrower, describing confidence in a specific task. A broad score does not capture how a person feels about their skills at work or their body image.
These distinctions matter because a finding about one type does not automatically apply to another. Studies usually measure these constructs through self-report scales and then look for patterns across groups, ages, and time periods. For a plain-language explanation of the terms, see what self-esteem is and how it is measured.
Five takeaway findings from self-esteem research
These five takeaway findings come from self-esteem research findings published in peer-reviewed journals and summarized by major health and psychology organizations. The points below come from meta-analyses, longitudinal studies, randomized trials, and reviews; each notes the kind of evidence behind it. Across these designs, a finding becomes more credible when it appears in several countries and at several ages.
- Self-esteem and mental health are consistently linked. Research indexed in PubMed Central includes studies that link lower self-esteem to higher anxiety and depression symptoms. Most of this evidence is correlational, so it does not prove that low self-esteem causes poor mental health. See the effects of low self-esteem for a reader-friendly overview.
- Social comparison and social media use are linked to self-esteem in student samples. A 2023 review by Yadav and Reddy reports that heavy social comparison and cyberbullying are associated with lower self-esteem among students. These studies are mostly cross-sectional; the finding is strongest for association, not direction. See social media and self-esteem for a closer look.
- Self-esteem has moderate, not uniform, links with life outcomes. Orth and Robins (2022) report that self-esteem correlates with several aspects of well-being, relationships, and work, but the effects are generally modest and may run in both directions. Some longitudinal studies suggest the relationship is reciprocal.
- A common lifespan pattern appears across studies. Bleidorn and colleagues (2018) describe a tendency for self-esteem to rise from adolescence into midlife before leveling off or declining in later life. Individual paths vary widely. Cultural norms appear to shape the timing and size of these changes.
- High self-esteem is not narcissism. Brummelman and colleagues find that interventions can raise self-esteem without producing narcissistic traits, although follow-up periods tend to be short. The studies vary, so this is a pattern, not a rule for every intervention.
Self-esteem studies at a glance
The scientific studies on self-esteem that anchor this summary differ in design. The four studies were chosen because they represent different methods rather than a single design. The table lists the evidence type and the main limitation of each study because correlation is common in this field. That design is helpful because a single cross-sectional sample may reflect one school or one year. For background on how these constructs developed, see the psychology behind self-esteem.
| Author (year) | Evidence type | Key finding | Main limitation |
|---|---|---|---|
| Orth & Robins (2022) | Meta-analysis of longitudinal studies | Self-esteem follows a long-term trajectory, generally rising into midlife before a slight late-life decline. | Most evidence is correlational; individual change is not uniform. |
| Bleidorn et al. (2018) | Cross-national age-trend analysis | A similar age pattern appears across several countries, with self-esteem often rising into midlife. | Cannot establish causality; cohort and cultural differences remain. |
| Brummelman et al. | Randomized controlled trials | Targeted interventions can improve self-esteem without necessarily increasing narcissism. | Short follow-up periods; long-term effects unclear. |
| Yadav & Reddy (2023) | Literature review, mostly cross-sectional | Social comparison and cyberbullying are associated with lower self-esteem among students. | Cross-sectional designs cannot confirm direction; samples often limited to students. |
A meta-analysis aggregates effect sizes from primary studies, while a cross-national analysis looks at age differences across samples. A randomized controlled trial randomizes participants to intervention or control, so each design carries a different kind of confidence. No study listed here should be read as a complete picture on its own; the table is a map, not an exhaustive list.
How self-esteem changes across the lifespan
Longitudinal research shows a common age pattern. Self-esteem often starts lower in adolescence, rises through young adulthood and midlife, and then may decline in later life. This pattern appears in the longitudinal meta-analysis by Orth and Robins (2022) and in the cross-national age-trend data reported by Bleidorn and colleagues (2018). The rise may reflect growing independence, work roles, and stable relationships; the late-life decline may reflect retirement, illness, or loss of social roles, but these explanations are not fully settled.
Longitudinal studies are stronger than cross-sectional surveys for describing change because they follow the same people over time. Because the studies track patterns rather than assign conditions, they describe trajectories, not causes. Individual paths vary widely: some people decline early, others remain stable.
Adolescence tends to show more fluctuation; midlife often brings stability; late life may bring gradual change. Cultural norms also shape which ages show gains or losses. See self-esteem across the lifespan for a fuller discussion.
Self-esteem vs narcissism: what studies do and don't show
Narcissism is not a synonym for high self-esteem. Self-esteem is about feeling worthy; narcissism includes entitlement, grandiosity, and a need for admiration. Research by Brummelman and colleagues uses intervention studies to test whether raising self-esteem increases narcissism.
The evidence so far suggests the two are separate: interventions can improve self-worth without increasing narcissistic traits. However, these trials often measure outcomes over short periods, so long-term effects are less clear. In everyday terms, this means a student can learn to value themselves without becoming self-centered.
The distinction also appears in everyday behavior: a person can feel good about their skills without claiming they are more deserving than others. Narcissism is a multidimensional construct; measuring it requires scales that capture entitlement and admiration, not just self-worth. The finding does not mean every praise intervention is harmless; results vary by study and method. The distinction matters when parents or teachers worry that praising children will create narcissism.
What the intervention evidence says so far
Intervention research is promising but mixed. Studies report that realistic goal-setting, self-compassion exercises, and cognitive-behavioral techniques can help people build self-esteem. CBT-style practices usually involve identifying unhelpful thoughts and testing them against evidence. Self-compassion exercises often focus on treating oneself with the same kindness one would offer a friend. Realistic goal-setting breaks larger aims into smaller, achievable steps.
Some studies use group formats, while others use self-guided materials, which may affect results. Most studies measure outcomes by asking participants to rate self-esteem before and after a program, using validated self-report scales. Randomized trials are rare, and many interventions have short follow-up periods or small samples. For that reason, research is mixed on whether improvements last and which ingredients matter most. These practices do not promise results, and they are not treatments for depression or anxiety.
How to use this summary and its sources
This summary is a starting point, not medical or psychological advice. Most findings describe correlations; they do not prove that changing self-esteem causes a change in another condition. When a study reports an association, that association may reflect other factors, such as stress, history, or social context.
Because many findings come from student or adult samples, they may not apply to every age group. For plain-language health information, MedlinePlus offers summaries of mental health topics. To explore primary studies, search the research indexed in PubMed Central.
The 988 Lifeline is a free, confidential service for mental health crises. If low mood lasts for weeks or interferes with daily life, a licensed mental health professional can help; in the U.S., the 988 Lifeline is available by call or text.
FAQ
What does research say about self-esteem and mental health?
Repeated findings link low self-esteem to higher risk of anxiety and depression, but most studies are correlational, not causal.
Does self-esteem decline with age?
Longitudinal studies suggest it often rises from adolescence to midlife and may decline in late life; individual paths vary.
Is high self-esteem the same as narcissism?
No: self-esteem reflects self-worth, while narcissism involves entitlement and grandiosity; interventions can raise one without the other.
Does social media lower self-esteem?
Studies of students link heavy social comparison and cyberbullying to lower self-esteem, though they are mostly cross-sectional.
What improves self-esteem, according to research?
Realistic goal-setting, self-compassion and CBT-style practices show promise, but the evidence is mixed and no practice promises specific results.
Sources
- American Psychological Association, “Self-Esteem” topic page.
- MedlinePlus, National Library of Medicine.
- PubMed Central, National Library of Medicine.
- Primary studies named in the text: Orth & Robins (2022); Bleidorn et al. (2018); Brummelman et al.; Yadav & Reddy (2023).

Keep reading