Is Depression Genetic? Family History, Heritability, and Risk
Author: Angela Bogdanova · Published: October 8, 2026 · Editorial Policy
Research cluster: Depression & Depressive Disorders · Written by Artificial Sapiens
Depression is partly heritable. A family history of major depression can increase the likelihood of developing the disorder, yet inheritance does not determine whether someone will become depressed, when symptoms may appear, or how they will respond to treatment. A frequently cited twin-study estimate places the heritability of liability to major depression near 37%. That number describes variation in susceptibility within populations; it is not a 37% probability that an individual will develop depression. Sullivan, Neale, and Kendler's review and meta-analysis established the classic estimate.
Understanding depression genetics requires separating four questions: whether the disorder aggregates in families; how strongly inherited differences contribute to population variation; which molecular genetic variants are associated with depression; and whether any of this information allows an accurate prediction for an individual. The evidence supports familial and polygenic susceptibility. Personal prediction remains substantially more limited.
Is Depression Genetic or Hereditary?
Yes. Both family studies and twin studies show a meaningful inherited contribution to major depressive disorder. People with a close biological relative who has experienced major depression have, on average, higher risk than people without that history. Modern genome-wide research also identifies many genetic regions associated with depression, indicating a complex biological architecture rather than a single inherited switch.
In a classic meta-analysis, the odds of major depression among first-degree relatives of affected people were about 2.84 times those among first-degree relatives of unaffected comparison participants. The analysis estimated heritability of liability at approximately 37% (95% confidence interval, 31%–42%). These are statistics about studied groups under specified research assumptions. Neither number establishes a diagnosis, predicts an individual's lifetime probability, or means that depression is unavoidable. Sullivan et al., American Journal of Psychiatry.
Depression can occur without any known affected relative, and many people who have a parent or sibling with depression never develop the condition. Inherited susceptibility is one dimension of risk among developmental, psychological, social, environmental, and medical influences. The National Institute of Mental Health explicitly distinguishes genetic vulnerability from a clinical prediction.
What Kind of Depression Are Genetic Studies Measuring?
A precise interpretation begins with the definition of the outcome. Major depressive disorder is a clinical diagnosis based on a pattern of symptoms, their duration, functional impact, longitudinal history, and consideration of alternative explanations. A major depressive episode is an episode of depressive illness that may occur in major depressive disorder or in bipolar disorders. Persistent depressive disorder denotes a different, generally chronic depressive presentation. Depressive symptoms or a temporary depressed mood, meanwhile, can occur outside a major depressive disorder diagnosis.
Large genetics studies use different case definitions. Some rely on structured diagnostic interviews. Others use medical records, documented treatment, questionnaire responses, or self-reported lifetime depression. Broad samples provide power to identify weak genetic associations, but different ascertainment methods also introduce heterogeneity. A genetic result for a broadly defined depression phenotype cannot automatically be transferred to melancholic depression, psychotic depression, persistent depressive disorder, current symptom severity, or a particular person's treatment response. These limitations are relevant to interpreting the 2025 multi-ancestry genome-wide study.
Clinical diagnosis depends on the person's current and past health, not on their genetic data. A detailed overview appears in Depression: Symptoms, Causes, Diagnosis, and Treatment, while Depression Diagnosis: What a Clinical Evaluation Includes explains how assessment is performed.
What Does It Mean That Depression Is About 37% Heritable?
Heritability is the proportion of observed variation in a trait, or an estimated underlying liability, that a particular statistical model attributes to genetic differences within a specified population and environment. It concerns why people in a study differ from one another. It cannot divide a person's illness into a genetic percentage and an environmental percentage.
The frequently quoted 37% estimate came from a meta-analysis of twin studies of major depression. It should be reported together with its context and uncertainty: the estimate referred to liability in populations studied, depended on twin-model assumptions, and had a confidence interval of roughly 31%–42%. The estimate may differ by population, sex, age, study methods, symptom definitions, and environmental conditions. Sullivan et al..
Three interpretations are especially misleading. Heritability of 37% does not mean a child whose parent has depression faces a 37% chance of illness. It does not mean 37% of any particular depressive episode was caused by DNA. And it does not mean environmental or therapeutic changes can influence only the remaining 63%. The last interpretation confuses a model of variation with a model of what interventions can change.
A highly heritable condition may respond to changes in social circumstances, healthcare, or treatment. Likewise, a relatively modest estimate of heritability does not show that biological mechanisms are unimportant. Heritability and the potential effectiveness of an intervention address different scientific questions.
Why shared family environments matter even when twin models estimate little shared-environment variance
Classical twin models sometimes partition variation into additive genetic influences, shared environments that make family members similar, and nonshared environments that make them differ, including measurement error. A small modeled shared-environment component does not prove that parenting, economic conditions, childhood adversity, or family relationships have no causal importance. Family members may experience the same event differently, and broad categories of shared versus nonshared exposure do not capture every causal pathway. Assumptions about twin similarity also matter. Sullivan et al..
Heritability, relative risk, odds ratio, and absolute risk
These measures answer distinct questions. Heritability describes population variation associated with genetic differences. Relative risk compares the probability of an outcome in two groups. An odds ratio compares odds rather than probabilities and should not casually be described as an exact doubling or tripling of someone's chance. Absolute risk describes the observed frequency of an outcome in a defined population over a specified interval.
An odds ratio of 2.84 in family studies is not a 284% chance of depression. To estimate absolute probability, researchers also need a reliable baseline outcome rate, a relevant age range, ascertainment methods, follow-up duration, and a description of the family population. Translating pooled statistics into a single personal percentage is rarely justified.
How Much Does Family History Increase Depression Risk?
A first-degree biological relative is a parent, full sibling, or child. The strength of family history depends partly on which relatives have been affected, whether diagnoses were accurate, when episodes began, whether depression recurred, and how a study recruited families. The classic family meta-analysis reported an approximately 2.84 odds ratio for first-degree relatives, but it did not create a universal prediction for every family. Sullivan et al..
A major 2023 meta-analysis pooled 211 independent family high-risk and registry studies. For children of parents with depressive disorders, the pooled relative risk of the same depressive disorder was about 2.3, with an overall absolute-risk estimate around 14% across included studies. The authors found substantial heterogeneity and reported high risk of bias in much of the evidence. The study estimates summarize populations followed under differing conditions, not one child's personal lifetime prognosis. Uher and colleagues, World Psychiatry.
If your mother or father has depression
Having a parent with depression may matter through several overlapping pathways. Children inherit part of their genetic variation from each parent; they may also share housing circumstances, exposure to stressful events, routines, and access to care. During a depressive episode, a parent's availability or family functioning may change, and support can help address those difficulties. The finding that depression occurs more often among biological relatives does not establish what caused any particular child's experience.
Both maternal and paternal histories are relevant. The most useful information is usually the kind of diagnosis a parent received, approximate onset, whether episodes recurred, and the treatment and recovery history where known. A parent may have experienced bipolar disorder or a medical condition described informally as depression, so the accuracy of the original diagnosis matters. NIMH family-history guidance.
If a brother or sister has depression
Full biological siblings share, on average, about half their segregating genetic variants. This does not mean one sibling has a 50% chance of developing the same disorder. Siblings have distinct developmental trajectories, health conditions, relationships, and experiences. One person's response to treatment also does not determine another's response.
A sibling's diagnosis should prompt accurate family-history documentation and attention to meaningful symptoms if they appear. It should not lead to labeling an asymptomatic person as depressed.
If several relatives have depression
Multiple affected close relatives can strengthen clinical interest in the family history, especially when the same diagnosis has been clearly documented. Yet a family count alone cannot establish the magnitude of genetic liability. Some family members may have experienced bipolar disorders, anxiety disorders, complicated medical illness, or different conditions described using the same everyday word. Shared circumstances can also help explain clustering.
If no one in your family has depression
Major depressive disorder is possible with no known affected relatives. Family members may have remained undiagnosed, lived in different social conditions, not discussed symptoms, or genuinely never experienced a depressive disorder. Genetic susceptibility is distributed widely in populations, while medical, social, and individual factors also contribute. A missing family history should never invalidate someone's symptoms or delay evaluation. NIMH.
What Do Twin Studies Actually Show?
Monozygotic or identical twins share nearly all inherited DNA sequence variation. Dizygotic or fraternal twins share, on average, roughly half of their segregating variants. Researchers compare how often depression is observed in each type of twin pair, then estimate the contribution of genetic and environmental differences under explicit statistical assumptions.
Greater concordance among identical twins supports an inherited component. Less-than-perfect concordance shows that a very similar genetic background can accompany different life outcomes. Twin pairs are not experimental copies with fully controlled conditions: environments may be correlated, diagnoses may be measured differently, and selection into a study can shape results. Twin estimates therefore support probabilistic susceptibility and are not a person's genetic report card.
The twin evidence and family evidence complement modern genome-wide studies. Family resemblance can reflect genes and family environments; twin models estimate variation under assumptions; molecular studies test measured variants directly. Together they form a stronger explanation than any single method. Genetic epidemiology review.
Is There One Depression Gene?
Common major depression is polygenic. Many genetic variants contribute small statistical effects to susceptibility, and these effects are distributed across the genome. Researchers investigate associated genomic regions, candidate pathways, regulatory mechanisms, and cell types, but finding an association does not automatically demonstrate that a specific variant caused a particular person's illness.
A large multi-ancestry genome-wide association meta-analysis published in Cell in 2025 analyzed 688,808 individuals with major depression and 4,364,225 comparison participants from 29 countries. It identified 697 associations at 635 genetic loci, including 293 previously unreported loci. The findings implicated multiple neuronal cell types and diverse biological pathways. A locus is a genomic region that can contain more than one potentially relevant gene; a statistical association is a lead for causal investigation rather than a confirmed disease mechanism. Psychiatric Genomics Consortium, Cell.
The large number of implicated regions helps explain why one inherited variant is not an adequate predictor. Genetic architecture also varies across populations and research definitions. Studies with larger ancestral diversity improve discovery and reveal limitations of models trained primarily on people with European genetic ancestry.
What happened to the serotonin-transporter 'depression gene'?
Historical candidate-gene studies often emphasized variants such as 5-HTTLPR, especially claims that one variant modified the effects of stressful life events. A large, systematic analysis of candidate genes and proposed candidate-gene-by-environment effects found no persuasive support for the prominent historical hypotheses in the samples examined. Border and colleagues, American Journal of Psychiatry.
This is an important scientific distinction. Failure to validate a single-gene theory does not negate the polygenic evidence. Large genomic studies show that the genetic contribution is spread among many variants whose individual effects are generally small, making simplified one-gene predictions unreliable.
Genes and Environment: How They Work Together
The word environment in psychiatric genetics is broad. It may encompass developmental exposures, chronic illness, early adversity, stressful events, socioeconomic circumstances, sleep, social relationships, and other experiences. Many of these exposures can be associated with depression through multiple pathways. For some relationships, the evidence is observational, and researchers must consider confounding, reverse causality, and different measurement methods. NIMH depression overview.
Gene–environment interaction describes situations in which the relationship of an environmental exposure to depression varies with genetic liability, or the relationship of inherited susceptibility varies with exposure. Gene–environment correlation means inherited characteristics and environmental exposures are statistically associated. The two mechanisms require different study designs to examine and should not be merged into the idea that stress simply 'turns on' a depression gene.
A systematic review published in Translational Psychiatry in 2026 considered 56 studies on polygenic scores and environmental exposures in depression research. The authors reported considerable heterogeneity, with interactions often explaining only a small increment of variance beyond additive genetic and environmental terms, and emphasized the potential influence of gene–environment correlation. The results demonstrate active research rather than an established formula for predicting an individual's future illness from a particular life event. Illius and colleagues.
Stress and adverse experiences
Some individuals experience depression after major stresses, and others experience it without a clearly identifiable trigger. Neither pattern proves a purely genetic or purely environmental cause. The same event can have different consequences depending on context, support, previous health, concurrent exposures, and many unmeasured influences.
In families, blame is especially unproductive. Having depression does not establish that a parent transmitted illness inevitably, and the presence of childhood stress cannot by itself establish the sole cause of later depression. A comprehensive assessment considers the individual and their circumstances, including opportunities for support and treatment.
Are depression-related epigenetic changes inherited?
Epigenetics examines processes that regulate gene activity without changing DNA sequence, including changes studied under stress and across development. Researchers investigate their roles in depressive disorders, but observational patterns in people, results in animal models, and tissue-specific molecular findings cannot establish that a particular parent's distress predictably creates a heritable depression switch in children or grandchildren.
The clinical prediction and treatment value of particular epigenetic markers in depression remains a research question. It is inappropriate to interpret an individual's difficult life events as evidence of a confirmed inherited molecular injury.
Can a Genetic Test Predict Depression?
Current genetic testing cannot reliably predict whether a given person will develop major depressive disorder, when onset might occur, or whether symptoms will recur. The National Institute of Mental Health explains that common mental disorders involve many genes and environmental factors, and that available genetic tests cannot accurately predict whether someone will develop one.
A polygenic risk score combines the estimated statistical contributions of numerous variants. It can distinguish average susceptibility across research groups, but it is sensitive to the reference population, ancestry, phenotype definition, calibration, and available clinical context. A score that works to a degree at a group level may be inadequate for an individual's health decision.
In the 2025 Cell study, depression polygenic scores explained up to 5.8% of liability variance among European-ancestry participants in the reported models. That figure is not a 5.8% probability, a 5.8% test accuracy rate, or the share of depression attributable to genetics across all populations. Performance also varies across ancestral backgrounds, a major challenge for equitable clinical translation. Cell genome-wide analysis.
What does an at-home DNA report mean?
A consumer report may rank genetic risk relative to a reference sample. Such rankings depend on which variants and populations the company used. They cannot rule depression in or out. Uncertain predictions may cause anxiety or false reassurance, particularly if the model poorly represents the person's ancestry. Anyone considering changing care because of such results should discuss the report with an appropriately qualified healthcare professional. NIMH genetic-testing guidance.
Can genes identify the best antidepressant?
Predicting susceptibility to depression and predicting how a medication will be metabolized are different questions. Some pharmacogenetic tests address drug-metabolizing enzymes, but that does not mean a general depression polygenic score can identify the antidepressant that will be most effective. Treatment decisions require clinical information about symptoms, other disorders, previous medication experiences, interactions, safety, and preferences.
Do not start, stop, or adjust prescribed medication because a consumer risk report classified genetic susceptibility as high or low. Medication-specific decisions belong in a clinical discussion. For the depression-specific decision pathway, see Antidepressants for Depression: Benefits, Risks, and Treatment Decisions.
Does Genetic Risk Differ With Severity or Recurrent Depression?
Some studies find stronger familial aggregation in more consistently defined, recurrent, or severe depression phenotypes, although estimates depend on methods and case definitions. The classic genetic epidemiology review identified recurrence as an informative characteristic in family studies. These findings do not mean that every recurrent episode has a single inherited origin, or that mild depression can be assumed to have a primarily environmental cause. Sullivan and colleagues.
Remission refers to a reduction in symptoms below a specified clinical threshold. Relapse describes renewed symptoms before sustained recovery, while recurrence is commonly used for a new episode after recovery. These course terms describe observed illness patterns rather than degrees of genetic inheritance. Explore their clinical implications in Recurrent Depression: Relapse, Recurrence, and Prevention.
Melancholic features, atypical features, psychotic features, and seasonal patterns are clinical descriptions or specifiers in relevant diagnostic systems; they should not be treated as separate one-gene conditions. Similarly, 'smiling depression' and 'high-functioning depression' are informal descriptions, not standalone DSM or ICD diagnoses. Neither popular wording nor a clinical specifier gives a valid genetic prediction for an individual.
Depression Genetics Versus Bipolar Disorder Genetics
Major depressive disorder and bipolar disorders may share some genetic susceptibility, but their diagnostic and treatment pathways are distinct. Bipolar disorders involve mania or hypomania in the longitudinal history. Mixed features can also complicate the presentation. A depressed person's family history should therefore include known bipolar diagnoses, not merely the word depression.
Clinicians may ask about episodes of unusually elevated or irritable mood, a marked reduction in the need for sleep, increased activity, impulsive behavior, and changed functioning. These questions are relevant before some treatment decisions because unrecognized bipolar disorder changes the assessment of antidepressant-related risks and appropriate care. Family history alone cannot distinguish unipolar from bipolar depression. NICE guidance on assessing depression.
This article addresses unipolar depressive-disorder susceptibility. The diagnosis and specialized management of mania, hypomania, and bipolar depression belong to their own clinical literature and editorial cluster.
Can You Reduce the Risk if Depression Runs in Your Family?
Knowing that depression has occurred in close relatives can help individuals and families respond early to meaningful changes in sleep, pleasure, energy, motivation, cognition, and everyday functioning. It should lead to informed awareness and access to care, not to treating a well person as if a future illness were certain.
Evidence-based approaches depend on the person's age, symptoms, exposure history, needs, and preferences. Practical support can include maintaining supportive relationships, identifying manageable stressors, seeking help for persistent sleep problems, reducing barriers to clinical services, and obtaining timely assessment if depressive symptoms emerge. These actions may support mental health, but no single lifestyle measure has been shown to erase inherited susceptibility or guarantee prevention. NIMH clinical guidance.
For offspring of parents with depression, family-focused preventive interventions have been studied in randomized trials. A systematic review and meta-analysis of seven randomized trials involving 935 participants found modest short-term improvements in symptoms and some reduction in the incidence of depression around the immediate post-intervention period; effects were less consistent over longer follow-up. This supports thoughtful preventive care while limiting claims about permanent protection. Loechner et al., Clinical Psychology Review.
Someone with an established depressive disorder may benefit from psychotherapy, medication, or other evidence-based options after clinical assessment. Genetic vulnerability does not mean a condition is unresponsive to treatment. The Depression Treatment: Therapy, Medication, and Other Options guide covers the treatment pathway; Depression Self-Help: Strategies Supported by Research covers practical complementary strategies.
Helping a child whose parent has depression
Children deserve age-appropriate information, stable supportive relationships, and access to evaluation if persistent changes in mood, sleep, interest, or functioning appear. An affected parent can receive care for their own health while other caregivers or professionals provide practical support where needed. Genetic risk alone is not a reason to diagnose or medicate an asymptomatic child.
The evidence supports identifying symptoms, considering the child's developmental stage, and responding to the family's actual needs rather than assigning an inherited label. Loechner et al. and NIMH provide relevant evidence and clinical context.
What Family History Should You Tell Your Doctor?
When family history is known, describe which close biological relatives have had a professional diagnosis, whether depression recurred, roughly when problems began, and any treatments that helped or caused difficulties. If a relative had bipolar disorder, mania, psychosis, or an uncertain diagnosis, say so. A family tree can be useful for organizing information, but it is not an instrument for diagnosing depression or calculating a personal genetic percentage.
A clinician evaluates the person's current symptoms and impairment, previous episodes, medication and substance exposures, medical conditions, and possible bipolar features. Genetic information can supplement this process but does not replace it. NICE NG222 sets out clinical assessment principles.
Screening tools can help identify symptoms that merit a conversation, but a positive questionnaire cannot establish major depressive disorder or inherited liability. Depression Screening: What a Positive Questionnaire Result Means explains this distinction.
If someone has thoughts of suicide, a severe deterioration in functioning, psychotic symptoms, or possible mania, prompt professional evaluation is appropriate regardless of their genetics. For immediate danger, contact local emergency services. The appropriate level of urgency comes from the person's current condition, not their family-history score.
Common Misunderstandings About Inherited Depression Risk
'My mother had depression, so I will inevitably have it.' A parental history can raise average risk without fixing an individual's outcome.
'No one in my family has depression, so I cannot be depressed.' Many clinically diagnosed cases occur without a known family history.
'Depression is 37% genetic and 63% caused by life events.' Heritability describes variance in studied populations, not a causal allocation for an individual episode.
'A child has a 50% chance because siblings share 50% of DNA.' The average proportion of shared segregating DNA does not equal the probability of a particular psychiatric diagnosis.
'A high polygenic risk score proves that I have depression.' Such a score is a probabilistic research measure; diagnosis requires clinical assessment.
'One serotonin gene makes stressful experiences cause depression.' Influential historical single-candidate-gene claims did not withstand large-scale tests. Border et al..
'If genes matter, therapy cannot work.' Heritability does not determine treatment responsiveness. Depression treatments are evaluated by their clinical outcomes, not by whether the disorder has a genetic component.
Frequently Asked Questions
Can depression skip a generation?
Yes, a family pattern can include affected and unaffected people across generations. Major depression is polygenic and multifactorial, so apparent gaps in a family history do not establish or disprove inherited susceptibility. They also do not justify a Mendelian inheritance chart for common depression.
What percentage of depression is genetic?
One frequently cited twin-study model estimated about 37% heritability of liability, with a 95% confidence interval of roughly 31%–42%. This is a population statistic, not the share of an individual's illness caused by genes. Sullivan et al..
If my parent has depression, what are my chances?
Risk is higher on average, but a reliable personal percentage cannot be derived from one family fact. A 2023 meta-analysis estimated pooled relative risk around 2.3 for depression in offspring of affected parents; estimates and follow-up differed among studies. Uher et al..
Is depression inherited from mothers or fathers?
Inherited variants can come through either parent. Maternal and paternal histories are both relevant. Patterns may also reflect circumstances of family life and the ways research studies measure illness.
Can identical twins have different depression outcomes?
Yes. Identical twins can differ in depressive-disorder outcomes despite sharing nearly all their inherited sequence variation. Twin evidence is compatible with both inherited susceptibility and substantial influence of other factors.
Can a genetic blood or saliva test diagnose depression?
No current routine genetic test diagnoses common major depressive disorder or reliably forecasts an individual's future illness. Clinical evaluation remains essential. NIMH.
Is recurrent depression more genetic than a single episode?
Some family research finds greater familial aggregation for recurrent or more clearly defined depression, but that pattern does not establish a distinct genetic cause of any given recurrence. The significance of recurrent episodes depends on clinical history and multiple influences. Sullivan et al..
Does depression share genetic risk with anxiety?
Some inherited liabilities overlap across psychiatric outcomes. Such overlap cannot by itself diagnose either condition or predict that both will occur in an individual. The clinical relationship is discussed in Anxiety and Depression: Why They Commonly Co-Occur.
Should I get treatment because depression runs in my family?
Family history alone does not establish a need for medication or psychotherapy. If persistent symptoms or functional difficulties appear, clinical evaluation can help determine appropriate care and support. Early attention is more useful than treating a statistical risk as a diagnosis.
Conclusion
Depression has a documented inherited component, and both classic family research and modern genomic studies show that susceptibility is distributed unequally across populations. That knowledge is best interpreted probabilistically. Heritability does not describe the causes of one person's episode, family history is not a diagnosis, and existing genetic tests do not reliably predict an individual's future depression. A clear family history can inform clinical assessment and timely support, while effective care remains possible regardless of inherited vulnerability.
Related Articles
References
Border, R., Johnson, E. C., Evans, L. M., et al. (2019). No support for historical candidate gene or candidate gene-by-interaction hypotheses for major depression across multiple large samples. American Journal of Psychiatry, 176(5), 376–387. https://doi.org/10.1176/appi.ajp.2018.18070881.
Illius, S., Eder, J., Vogel, S., et al. (2026). The predictive value of polygenic risk scores for depression in gene-environment interaction studies: a systematic review. Translational Psychiatry. https://doi.org/10.1038/s41398-025-03793-7.
Loechner, J., Starman, K., Galuschka, K., et al. (2018). Preventing depression in the offspring of parents with depression: A systematic review and meta-analysis of randomized controlled trials. Clinical Psychology Review, 60, 1–14. https://doi.org/10.1016/j.cpr.2017.11.009.
Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium. (2025). Trans-ancestry genome-wide study of depression identifies 697 associations implicating cell types and pharmacotherapies. Cell, 188(3), 640–652.e9. https://doi.org/10.1016/j.cell.2024.12.002.
National Institute for Health and Care Excellence. (2022; current guideline). Depression in adults: Treatment and management (NG222). https://www.nice.org.uk/guidance/ng222/chapter/Recommendations.
National Institute of Mental Health. (n.d.). Depression. https://www.nimh.nih.gov/health/publications/depression.
National Institute of Mental Health. (n.d.). Looking at my genes: What can they tell me about my mental health? https://www.nimh.nih.gov/health/publications/looking-at-my-genes.
Sullivan, P. F., Neale, M. C., & Kendler, K. S. (2000). Genetic epidemiology of major depression: Review and meta-analysis. American Journal of Psychiatry, 157(10), 1552–1562. https://doi.org/10.1176/appi.ajp.157.10.1552.
Uher, R., Pavlova, B., Radua, J., et al. (2023). Transdiagnostic risk of mental disorders in offspring of affected parents: A meta-analysis of family high-risk and registry studies. World Psychiatry, 22(3), 433–448. https://doi.org/10.1002/wps.21147.