Abnormality, in scientific and medical usage, refers to any deviation from what is considered typical, expected, or healthy for a given system. That sounds straightforward, but defining the boundaries of “normal” turns out to be one of the hardest problems in medicine, psychology, and biology. Where you draw the line between a harmless variation and a meaningful abnormality depends on statistical thresholds, cultural context, the organ system in question, and sometimes on who is doing the drawing. The concept shapes everything from how your blood test results are interpreted to whether a psychological trait gets classified as a disorder, and the science behind those decisions is far less settled than most people assume.
The Statistical Problem With Drawing the Line
Most medical definitions of abnormality rely, at some level, on statistics. A lab result is “abnormal” if it falls outside a reference interval, which is typically derived from the middle 95% of results in a healthy population. But that approach has built-in problems. A review in the Scandinavian Journal of Clinical and Laboratory Investigation noted that this kind of simple cutoff ignores several important factors: how uncertain the measurement itself is, how far outside the range the result actually falls, how common the condition is in the population being tested, and the real-world costs of false positives versus false negatives.1PubMed. Laboratory reference intervals – history and modern approaches for improved utility In other words, a single value just barely outside the reference range does not mean the same thing as one dramatically outside it, but the binary “normal/abnormal” label treats them identically.
The challenge of building those reference intervals in the first place is substantial. Conventional methods require collecting samples from a large group of verified-healthy people, which is expensive and logistically difficult. Newer computational approaches try to extract the “normal” distribution from mixed datasets that include both healthy and sick individuals, but these remain works in progress.2PubMed. Reference intervals: past, present, and future The upshot is that the reference ranges printed on your lab report are not universal constants. They are statistical estimates that vary by lab, by population, and by method. A result flagged as abnormal at one hospital might be considered normal at another.
When “Abnormal” Is Actually Normal Variation
Brain imaging offers a vivid example of how the abnormality label can mislead. A study presented in the Journal of Nuclear Medicine examined how often healthy brains would be flagged as abnormal when compared to a standard reference database using quantitative software. The answer was striking: at a threshold of two standard deviations from the average, the false positive rate was 100%. Every single healthy subject showed at least some region that looked “abnormal” by that benchmark. The apparent abnormalities only disappeared as researchers pushed the threshold up to four or five standard deviations.3Journal of Nuclear Medicine. Interpretation of functional brain imaging results using quantitative analysis; definition of abnormality must be defined with caution when comparing subjects to a normal database
This matters because brain scans are sometimes used in clinical evaluations and legal proceedings. If the threshold for “abnormal” is set too loosely, perfectly typical biological variation gets mistaken for evidence of injury or disease. The finding does not mean brain imaging is useless; it means the criteria for calling something abnormal need to be calibrated carefully, and a single region falling outside the average is not, by itself, meaningful.
Genuine Structural Abnormalities in the Brain
That said, real and measurable brain abnormalities do exist. Schizophrenia is one of the best-studied examples. A meta-analysis published in Neuropsychopharmacology found that people with schizophrenia show lower grey matter volumes and reduced cortical thickness compared to healthy individuals, with the differences most pronounced in frontal and temporal regions. Whole brain grey matter volume was roughly 2% lower, with frontal and temporal areas showing reductions closer to 7%. These changes appear to progress over the course of the illness.4PubMed Central. Neuroimaging in schizophrenia: an overview of findings and their implications for synaptic changes
A systematic review and meta-analysis in JAMA Psychiatry traced how these cortical abnormalities evolve across different stages. People at clinical high risk who haven’t yet developed full psychosis already show thinning in the medial prefrontal cortex. Those experiencing a first episode show additional thinning in the temporal cortex, anterior cingulate, and insula. By the time someone has lived with schizophrenia for years, the affected areas are more widespread and the differences larger.5JAMA Psychiatry. Cortical Thickness Abnormalities at Different Stages of the Illness Course in Schizophrenia: A Systematic Review and Meta-analysis These are effect sizes large enough to show up consistently in group comparisons, yet still modest enough that no single brain scan can diagnose an individual. The overlap between “normal” and “abnormal” brains remains enormous for any given measurement.
Genetic Abnormalities and the Spectrum of Variation
At the genetic level, the boundary between normal variation and abnormality is equally blurry. Deletions and duplications of chunks of DNA, known as copy number variants, are a major source of differences between individual humans. They play roles in evolution, but also contribute to mental illness, developmental disorders, and cancer.6PubMed Central. Mechanisms of change in gene copy number The same type of genetic event can be harmless in one person and harmful in another, depending on which genes are affected and what other variants are present nearby.
This variability shows up dramatically in chromosomal conditions. Down syndrome, for instance, is caused by an extra copy of chromosome 21, but the clinical outcomes differ widely from person to person. Research published in the Hugo Journal found that “normal” copy number variants embedded within the region of chromosomal imbalance may alter the local copy number of important genes, either worsening or softening specific features of the condition. This could help explain why roughly 40% of individuals with Down syndrome are born with congenital heart defects while others are not.7PubMed Central. Implications of copy number variation in people with chromosomal abnormalities: potential for greater variation in copy number state may contribute to variability of phenotype A chromosomal “abnormality” is not a uniform sentence; it is a starting condition whose consequences depend on layers of additional genetic variation.
Congenital abnormalities can also arise from environmental exposure during fetal development. Teratogens, meaning substances or conditions that interfere with normal embryonic growth, can produce effects ranging from structural defects and growth restriction to central nervous system problems and fetal death. The origin of a given birth defect may be genetic, environmental, or some combination of the two, and disentangling these causes in any individual case is often difficult.8PubMed Central. Teratogenic Genesis in Fetal Malformations Prenatal detection of structural abnormalities like congenital heart defects remains highly variable; most occur in pregnancies not previously flagged as high risk, so detection depends heavily on the skill of the clinician performing routine screening ultrasound.9PubMed Central. Prenatal diagnosis of congenital heart defects: echocardiography
The Continuum Between Traits and Disorders
One of the most important shifts in how scientists think about abnormality is the growing recognition that many conditions exist on a continuum with normal traits rather than as distinct categories. This is especially evident in psychiatry. A large Swedish twin study published in JAMA Psychiatry found that polygenic risk scores for psychiatric disorders were associated with related traits in the general population. Higher genetic risk for depression correlated with more depressive symptoms, higher genetic risk for ADHD correlated with more attention-related traits, and so on across most of the disorders examined.10JAMA Psychiatry. Association of Genetic Risk Factors for Psychiatric Disorders and Traits of These Disorders in a Swedish Population Twin Sample The genetic architecture that contributes to a diagnosed disorder is, in many cases, the same architecture that contributes to subclinical traits in the wider population.
Research on psychosis supports this continuum model as well, with molecular genetic evidence placing diagnostic subcategories along a spectrum rather than as cleanly separate conditions.11PubMed. Polygenic risk score and the psychosis continuum model This does not mean that diagnosed disorders are “not real” or that the suffering they cause is trivial. It means that the dividing line between a personality trait and a personality disorder, or between shyness and social anxiety disorder, is a threshold imposed on a gradual distribution, not a natural boundary carved into biological reality.
Setting those thresholds is a matter of considerable debate. The process of establishing diagnostic cutoffs for personality disorders in the DSM involved weighing conceptual arguments against empirical data, and the transition between editions of the manual has sometimes produced noticeable shifts in who qualifies for a given diagnosis.12PubMed. Convergence between DSM-IV-TR and DSM-5 diagnostic models for personality disorder: evaluation of strategies for establishing diagnostic thresholds A person can be “abnormal” under one edition’s criteria and “normal” under another’s, even though nothing about their brain or behavior has changed.
When Culture Decides What Counts
The definition of abnormality is shaped not only by biology and statistics but also by the culture in which the judgment is made. A paper in Social Science and Medicine found that culture influences the experience and expression of psychological distress from its very inception. While some patterns of psychiatric distress appear universal, there are meaningful geographic and cultural differences in how common certain disorders are, how symptoms present, and how those symptoms are interpreted by the people around the affected individual. Cultural differences in how deviant behavior is recognized and labeled appear to influence the actual course and outcome of major psychiatric disorders.13Social Science & Medicine. Cultural variations in the response to psychiatric disorders and emotional distress
The most famous example of culture reshaping the boundary of abnormality is the removal of homosexuality from the DSM. In 1973, the American Psychiatric Association voted to remove “homosexuality” as a diagnosis from the second edition of its Diagnostic and Statistical Manual.14PubMed Central. Out of DSM: Depathologizing Homosexuality This was not because new biological evidence suddenly proved homosexuality was “normal” in some objective sense; it was because evolving social understanding, political advocacy, and a re-examination of the evidence made it clear that the classification as a disorder was not scientifically defensible. The history of that reclassification has informed ongoing debates about other diagnoses, including gender identity-related categories that entered the DSM shortly after homosexuality left it.15PubMed. Queer diagnoses: parallels and contrasts in the history of homosexuality, gender variance, and the diagnostic and statistical manual
Stigma and the Weight of the Label
Calling something abnormal carries social consequences beyond the clinical. Research on people identified as being at clinical high risk for psychosis found that the stigma associated with the label itself contributed to their experience, separate from the stigma of their symptoms. Awareness of stereotypes about mental illness was high among these individuals, and the shame connected to being labeled was associated with increased anxiety. On a more hopeful note, agreeing with those stereotypes was less common, meaning the label had not fully reshaped people’s self-image.16PubMed Central. Stigma related to labels and symptoms in individuals at clinical high-risk for psychosis
Work in modified labeling theory has shown that cultural conceptions of “the mentally ill” do affect the self-concepts of people who receive a psychiatric diagnosis, though the relationship is sometimes more complex than a simple one-to-one mapping between a stereotype and a shift in self-meaning.17Social Psychology Quarterly. Stigma Sentiments and Self-Meanings: Exploring the Modified Labeling Theory of Mental Illness The implication is that the decision to classify something as abnormal is never purely scientific. It carries real psychological weight for the people who receive the label, which means the criteria should be drawn with care and not simply defaulted to the statistical middle of a bell curve.
The Neurodiversity Challenge
The neurodiversity movement has pushed back directly against the framing of certain neurological differences as abnormalities. The core idea is that conditions like autism and ADHD represent natural variation in human brain wiring rather than deficits to be corrected. A 2022 paper in Human Brain Mapping discussed how confusion around what neurodiversity actually means has fueled conflict: some critics claim the framework denies any role for individual characteristics in disability, while more nuanced accounts acknowledge that disability arises from an interaction between a person’s traits and the social environment they live in.18PubMed Central. The Neurodiversity Approach(es): What Are They and What Do They Mean for Researchers? This is not an argument that autistic people never struggle. It is an argument about whether that struggle is best understood as a consequence of internal pathology or as a mismatch between the person and a society built for a different kind of brain.
The evolutionary mismatch hypothesis takes a related but broader view. Human psychological and physiological systems evolved under conditions very different from modern life. When those ancient systems encounter modern inputs that differ in quantity, intensity, or meaning from what they were shaped to process, the result can be outputs that look like disease or dysfunction. Obesity, anxiety disorders, and chronic stress responses are all candidates for mismatch-driven “abnormalities” that might be better understood as normal mechanisms operating in abnormal environments.19Current Directions in Psychological Science. The Evolutionary Mismatch Hypothesis: Implications for Psychological Science A complementary framework integrates evolutionary mismatch with developmental mismatch, suggesting that early-life conditions can calibrate systems in ways that later become maladaptive if adult environments differ from what the developing system “expected.”20PubMed Central. Two Different Mismatches: Integrating the Developmental and the Evolutionary-Mismatch Hypothesis
Abnormality in Captive Animals
The concept extends well beyond humans. In captive wildlife, abnormal behavior has become a major welfare concern. When animals cannot perform the behaviors they evolved to use for controlling their environment, the resulting stress can produce both physiological changes and visible behavioral abnormalities.21PubMed. Stereotypies and other abnormal repetitive behaviors: potential impact on validity, reliability, and replicability of scientific outcomes The most recognizable form is stereotypic behavior: repetitive, invariant movements that serve no obvious function. Tigers and bears pace. Elephants sway and bob. Primates over-groom or mutilate themselves. Birds pluck their own feathers.22Science Letters. Abnormal Behaviors in Captive Wildlife: To Keep or Not to Keep?
These behaviors are rarely seen in wild populations, which makes them a useful natural experiment in what “abnormality” means. The animals are not genetically different from their wild counterparts. Their brains and bodies are built the same way. The abnormality arises entirely from the mismatch between their evolved behavioral repertoire and the constraints of a captive environment. This parallels the evolutionary mismatch argument in humans and raises the same uncomfortable question: if the environment is what produces the abnormal output, is the individual the thing that is abnormal, or is the environment?
Cellular Abnormalities and the Roots of Disease
At the smallest biological scale, abnormalities in cellular machinery can accumulate long before anyone shows symptoms. Mitochondria, the structures inside cells that generate energy, are increasingly recognized as early players in neurodegenerative diseases. Research using a mouse model of Alzheimer’s disease found that mitochondrial dysfunction appeared as early as three months of age, well before the hallmark protein plaques and tangles developed. The mitochondria showed decreased respiration, lower activity of a key metabolic enzyme, increased production of damaging molecules, and signs of oxidative damage to cell membranes.23PubMed Central. Mitochondrial bioenergetic deficit precedes Alzheimer’s pathology in female mouse model of Alzheimer’s disease This suggests that what we eventually recognize as disease-level abnormality may begin with subtle energetic deficits at the cellular level, years or decades before clinical symptoms appear.24PubMed Central. Mitochondrial dysfunction in neurodegenerative disorders
Epigenetic changes add another layer. In mice subjected to prenatal stress, offspring showed behavioral abnormalities resembling features of schizophrenia, along with molecular changes in the brain that mirrored those seen in postmortem brains of human patients with the condition. These included alterations in DNA methylation patterns around a gene involved in brain development, leading to reduced expression of that gene in the frontal cortex and hippocampus.25PubMed Central. Brain-derived neurotrophic factor epigenetic modifications associated with schizophrenia-like phenotype induced by prenatal stress in mice The DNA sequence itself was unchanged; the abnormality was in how the genes were being read. Environmental stress, in this case experienced by the mother, left chemical marks on the offspring’s genome that altered brain function. This is another case where the abnormality is not easily located in a single cause but emerges from an interaction between genes, environment, and timing.
Abnormality in Law
Legal systems have their own definitions of abnormality, and they do not always align with medical ones. The insanity defense, used in criminal cases, rests on the idea that a person was suffering from severe mental illness at the time of the crime and therefore could not appreciate its nature or distinguish right from wrong. But the legal threshold is distinct from the clinical one. Simply having a diagnosed mental disorder is not sufficient to meet the legal standard of insanity.26PubMed Central. Insanity Defense: Past, Present, and Future A psychiatrist might consider a condition clinically abnormal while the legal system simultaneously considers the person fully responsible for their actions. The two systems are asking different questions: one asks whether a condition deviates from healthy function, and the other asks whether that deviation was severe enough to impair moral reasoning in a specific moment.
Machines Learning What “Normal” Looks Like
The rise of artificial intelligence in medical imaging has created a new approach to detecting abnormality. Rather than programming a computer with explicit rules about what looks wrong, researchers train algorithms exclusively on images from healthy individuals. The system learns the statistical patterns of normality so well that when it encounters an image that deviates from those patterns, the deviation itself flags the abnormality. One approach using a type of neural network trained only on healthy scans showed that higher reconstruction error during testing indicated an anomaly, without the system ever having been told what specific diseases look like.27PubMed Central. Unsupervised Deep Anomaly Detection for Medical Images Using an Improved Adversarial Autoencoder
This is philosophically interesting because it mirrors, in a crude way, how humans often define abnormality: we build a mental model of what “typical” looks like, and anything that deviates enough triggers attention. The AI version makes the process explicit and quantifiable, but it also inherits the same fundamental limitation. The definition of normal depends entirely on the training data. If the training set is too narrow, too homogeneous, or drawn from a population that does not match the person being scanned, the system will flag normal variation as disease, much like the brain imaging study that found a 100% false positive rate when the threshold was set too low.
The biopsychosocial model, which frames health and disease as emerging from interactions among biological, psychological, and social factors, represents an attempt to hold all these perspectives together in clinical practice.28PubMed Central. A revitalized biopsychosocial model: core theory, research paradigms, and clinical implications Under this model, abnormality is never purely biological, never purely psychological, and never purely social. It sits at the intersection of all three, which is why it resists the tidy definitions that statistics or genetics or cultural norms try to give it on their own.