Anshika Jain
For decades, modern healthcare has relied on population averages to determine whether a person’s health measurements fall within an expected range. Blood pressure, blood glucose, cholesterol, heart rate, kidney function and numerous other clinical indicators are routinely compared with reference values derived from groups of people. This approach has helped physicians identify potential abnormalities, establish diagnostic standards and make medical decisions using consistent scientific evidence.
However, the concept of normal is more complicated than a number printed on a laboratory report. Two individuals of the same age may have different physiological baselines, genetic characteristics, lifestyles, medical histories and responses to environmental conditions. A measurement considered normal for one person may represent a meaningful change for another, even when both results remain within the same accepted reference range.
This has encouraged a broader shift in medical thinking, from interpreting health through isolated measurements to understanding individual health trajectories. Rather than asking only whether a patient differs from the average person, healthcare can also examine how that patient’s condition changes over time. This perspective does not eliminate the importance of population research. Instead, it combines established clinical standards with personal history to develop a more complete understanding of health, disease progression and preventive care.
How Healthcare Developed the Concept of Normal
Population-based reference ranges emerged from the need to interpret clinical measurements consistently. Laboratories require standardised ways to distinguish results that commonly occur among healthy people from those that may indicate disease. Reference intervals provide a practical framework for making these comparisons across large numbers of patients.
In laboratory medicine, a conventional reference interval is often established using the central 95% of results from an appropriately selected reference population. Consequently, some healthy individuals will naturally fall outside the interval, while a result inside it does not automatically establish that a person is healthy. Reference intervals are not universal definitions of health; their interpretation depends on the specific test, the population studied and the clinical circumstances.
Population-based standards remain essential because they are supported by research involving many people and help clinicians recognise established patterns of disease. They are particularly valuable when a patient has no previous test results available for comparison. Without such benchmarks, interpreting an unfamiliar measurement would be substantially more difficult.
The limitation emerges when a population reference range becomes the only framework for interpretation. A single result offers a snapshot of physiological status, but it may not reveal whether the patient’s biology is stable, improving or deteriorating. Healthcare therefore needs a complementary approach that considers both the broader population and the individual person.
Why an Average Does Not Always Describe an Individual
An average summarises a group, but it cannot fully describe the biological diversity within that group. People differ in their inherited characteristics, body composition, activity levels, sleep patterns, nutritional intake, medication use and underlying health conditions. These differences influence how their bodies regulate physiological processes.
Consider two individuals whose blood test results fall within the same laboratory reference interval. One person has maintained a relatively stable measurement for several years. The other has experienced a gradual change from a previously consistent baseline. Although neither result may trigger an automatic laboratory alert, the second person’s pattern could justify closer clinical evaluation, depending on the measurement and the surrounding medical evidence.
The opposite situation is also possible. A person may repeatedly receive a mildly abnormal result because their usual biological pattern sits near or beyond a population reference limit. That finding should not automatically be dismissed, but neither should every repeated deviation be interpreted as evidence of disease without considering the wider clinical picture.
This illustrates the distinction between population-level normality and individual-level change. The population reference range asks where a measurement sits relative to other people. An individual trajectory asks whether the measurement is behaving as expected for that particular person and whether its direction or rate of change may be clinically important.
Neither question is sufficient on its own. Their combined interpretation can offer a more nuanced picture of health.
Individual Health Trajectories: Understanding the Body Over Time
An individual health trajectory describes how a person’s physiological measurements, symptoms, functional abilities and other health indicators evolve over time. Instead of treating each medical appointment as an isolated event, this approach considers the relationship between successive observations.
A trajectory can reveal gradual changes that are difficult to identify from a single measurement. Repeated blood pressure readings, for example, may help clinicians distinguish a temporary fluctuation from a persistent upward trend. Changes in kidney function, blood counts or metabolic markers may also become more meaningful when compared with a person’s previous results and clinical circumstances.
Longitudinal assessment is particularly valuable in chronic disease management. Conditions such as diabetes, cardiovascular disease and chronic kidney disease often develop or progress over extended periods. Monitoring trends can help clinicians assess whether a treatment is working, whether a risk factor is worsening and whether further evaluation is appropriate.
However, a trajectory is not simply a line connecting measurements. Results can change because of illness, medication, hydration, exercise, laboratory variation or other temporary influences. Clinicians must determine whether a change is persistent, meaningful and relevant to the patient’s health rather than interpreting every fluctuation as a warning sign.
The objective is to understand the direction and context of change, not to create an expectation that healthy bodies must remain perfectly constant.
Personalised Reference Intervals and the Future of Laboratory Medicine
Personalised reference intervals represent one developing method for bringing individual history into the interpretation of laboratory results. Instead of relying exclusively on population-derived limits, these approaches estimate an individual’s expected range using repeated measurements, biological variation and relevant population information.
Research published in Scientific Reports in May 2026 examined an approach for developing individual reference intervals by combining personal measurements with information from comparable individuals. The researchers reported potential applications in detecting changes associated with chronic disease and monitoring disease progression. The study illustrates how personal health histories may complement conventional laboratory interpretation.
A separate 2025 study published in Clinical Chemistry explored a statistical method for estimating personalised reference intervals using population information alongside an individual’s available test results. Such methods are important because a patient may have only a limited testing history, especially early in life or when receiving care for the first time.
Personalised reference intervals are not intended to replace every existing clinical threshold. Certain conditions require established diagnostic criteria, and some laboratory results demand urgent attention regardless of a person’s previous baseline. Instead, individualised interpretation can provide additional context where longitudinal patterns are clinically useful.
The long-term challenge is to determine which tests benefit most from personal reference intervals, how much historical data is necessary and how clinicians can distinguish genuine physiological change from random variation.
The Role of Artificial Intelligence in Detecting Health Patterns
Artificial intelligence is creating new opportunities to analyse longitudinal health information. Electronic health records contain laboratory results, prescriptions, diagnoses, clinical notes and other observations collected throughout a patient’s medical journey. When these records are sufficiently complete and reliable, computational models can help identify patterns that would be difficult to recognise from isolated measurements.
AI systems may estimate a person’s usual range for a particular biomarker, detect unusual changes, compare current findings with previous measurements and identify combinations of indicators that warrant clinical review. Machine learning can also examine relationships between measurements and later health outcomes across large patient populations.
The potential value lies in combining two perspectives. Population data provide information about patterns observed across many people, while individual histories help establish what is usual for a particular patient. A model that uses both sources may offer a more useful assessment than one that considers either in isolation.
Nevertheless, an algorithmic prediction is not a diagnosis. Models can produce misleading results when records are incomplete, measurements are inconsistent or the training data do not adequately represent the patient being assessed. Changes in laboratory equipment, healthcare access and clinical practice can also affect model performance.
AI should therefore support clinical reasoning rather than replace it. Physicians must evaluate algorithmic findings alongside symptoms, examination results, medical history and established diagnostic guidance.
The Risk of Over-Personalising Healthcare
Moving beyond population averages may appear to offer an obvious improvement, but excessive personalisation introduces its own risks. A patient’s previous measurements are not necessarily a reliable definition of healthy functioning. Historical data may contain undiagnosed disease, temporary abnormalities or changes caused by medication and other interventions.
If a system treats every departure from an individual’s historical baseline as abnormal, it may generate unnecessary warnings and investigations. Conversely, if a harmful change has become persistent, a model might gradually treat it as the person’s new normal and fail to recognise the underlying problem.
Research published in 2026 on longitudinal blood biomarkers highlighted this challenge. The study, available through PubMed Central, examined a large collection of laboratory measurements and warned that purely individualised approaches can overfit personal data. Its proposed framework combined patient history with population-level information to improve the interpretation of biomarker patterns.
This finding reinforces an important principle: individualisation should not mean abandoning shared clinical evidence. A person’s history provides valuable information, but population data help determine whether a pattern is plausible, clinically meaningful and associated with known health risks.
The most effective approach is likely to be a balanced system in which personal trajectories, population reference ranges and condition-specific clinical thresholds work together.
From Reactive Diagnosis to Predictive and Preventive Care
Traditional healthcare often responds when a patient develops symptoms, crosses a recognised threshold or receives an abnormal test result. Although this approach remains essential, longitudinal analysis may help identify concerning changes earlier in the disease process.
A gradual deterioration in a relevant biomarker, for example, could prompt a clinician to review a patient’s medication, investigate possible causes or recommend appropriate follow-up before the change becomes more pronounced. The significance of such a trend depends on the specific marker, the patient’s risk factors and the strength of the supporting evidence.
Preventive care may also benefit from tracking patterns in sleep, physical activity, weight, blood pressure and metabolic measurements. When these observations are interpreted together, clinicians may gain a more comprehensive view of how lifestyle, treatment and underlying conditions interact over time.
However, prediction must be distinguished from certainty. A statistical model may identify an increased probability of a future condition, but that does not mean the condition will necessarily develop. Clinical follow-up should be proportionate to the level of risk, the reliability of the evidence and the potential benefits and harms of further testing.
The goal is not to turn every minor variation into a medical problem. It is to identify meaningful changes early enough to support appropriate decisions while avoiding unnecessary anxiety and intervention.
Continuous Monitoring and the Changing Definition of Normal
Wearable devices, home monitoring equipment and digital health applications are expanding the amount of information available about individual physiology. Smartwatches can record heart rate and activity patterns, while connected devices can track measurements such as blood pressure and glucose in appropriate settings.
These technologies make it possible to observe changes that traditional appointments may miss. A clinician might review patterns across weeks or months instead of relying exclusively on measurements collected during occasional visits. This can be particularly useful when a condition requires ongoing monitoring or when symptoms are intermittent.
Yet continuous measurement creates new interpretive challenges. Devices differ in accuracy, some measurements are affected by movement or environmental conditions, and consumer wearables do not provide the same level of diagnostic evidence as every clinical instrument. A larger volume of data does not automatically produce better medical decisions.
Healthcare systems must determine which signals are clinically useful, how frequently they should be collected and what action should follow an alert. Without these safeguards, continuous monitoring may overwhelm patients and clinicians with inconsequential fluctuations.
A more sophisticated definition of normal will therefore depend not only on collecting more data but also on distinguishing reliable patterns from noise.
Privacy, Equity and Ethical Challenges
Individualised healthcare depends on access to sensitive personal information. Longitudinal medical records may reveal changes in physical health, medication use, reproductive health and other private circumstances. Combining these records with wearable data or genetic information can make the resulting profile even more sensitive.
Healthcare providers and technology developers must protect such information through appropriate security, access controls, governance and informed consent practices. Patients should understand how their information is being used, who can access it and whether it may be used for purposes beyond their immediate care.
Equity is another important concern. Patients who receive frequent testing may have extensive medical histories, while those with limited healthcare access may have few records. If personalised systems work best for people with abundant data, they could unintentionally widen existing disparities in healthcare quality.
Models must also be evaluated across diverse populations to ensure that they do not systematically misinterpret the health patterns of particular groups. Personalised medicine should improve the relevance of clinical decisions without making high-quality care dependent on access to expensive devices or extensive testing.
The success of individualised healthcare will ultimately depend on scientific validity, responsible data use and the ability to deliver its benefits fairly.
What the Future of Healthcare May Look Like
The future of medical interpretation is likely to combine population-level evidence with a more detailed understanding of individual change. Standard reference ranges will remain necessary for screening, diagnosis and clinical consistency, while personal health trajectories may provide additional insight into subtle changes and treatment responses.
Electronic health records, AI-assisted analysis and improved laboratory methods could help clinicians build a more complete picture of each patient’s physiological history. Over time, these tools may support earlier recognition of selected health changes, more informed treatment monitoring and more carefully targeted preventive interventions.
Realising this potential will require prospective clinical validation, clear standards for evaluating personalised models and evidence that the new approaches improve patient outcomes. Researchers must establish when individualised reference intervals are beneficial, when conventional thresholds remain preferable and how uncertainty should be communicated to clinicians and patients.
Healthcare professionals will also need practical tools that integrate these insights into everyday workflows without creating unnecessary alerts or administrative burdens. A useful system should clarify clinical decisions rather than make them more complicated.
Ultimately, precision healthcare is not about finding a unique number for every individual. It is about interpreting measurements in context, understanding biological variation and recognising when change carries genuine clinical significance.
Conclusion
The movement from population averages to individual health trajectories represents an important evolution in the way healthcare understands normality. Population-based reference ranges provide essential scientific benchmarks, but they cannot fully describe the biological history or changing circumstances of every person.
By examining longitudinal measurements alongside population evidence, clinicians may gain a clearer understanding of individual baselines, meaningful changes and treatment responses. Personalised reference intervals, artificial intelligence and continuous monitoring offer promising tools for this work, but they must be applied carefully and supported by reliable clinical evidence.
The central challenge is to avoid two extremes: assuming that one standard range describes every person perfectly, or treating every individual variation as a sign of disease. A balanced approach recognises both shared patterns and personal differences.
As healthcare becomes more data-driven, the definition of normal may increasingly depend on more than a single result. It may also involve understanding where a person started, how their health has changed and what those changes mean in the context of their overall condition. That shift could help medicine become more precise, preventive and responsive without losing the scientific standards on which safe healthcare depends.