You just got your lab test results back and your doctor tells you that all your tests were normal. Or maybe there were a couple of tests that fell outside the normal range. But what does “normal” mean and what does it mean if you’re outside this range? Read on to find out more about normal lab values, how normal reference ranges are created, why normal doesn’t always mean healthy, and why normal ranges differ between labs.
In most cases, normal lab values are a range of values that 95% of a healthy population falls into. However, there are some exceptions to this. For example, for markers of heart damage called troponins, 99% of the healthy population have values that fall within the normal range .
Normal ranges are commonly referred to as reference ranges or reference intervals by labs and healthcare providers. Doctors use them to interpret their patients’ lab test results, help make a diagnosis, or decide on treatment [1, 2].
In the context of lab test results, “normal” does not mean usual, typical, or ordinary. Instead, “normal” refers to how these values line up on a graph and form a range. The word “normal” here means a “normal distribution”, or set, of lab values. When we graph this, it looks like the symmetrical, bell-shaped curve seen below.
In a normal distribution, the average value lies in the center. Half of the population will have values that lie on the left side of this average; the other half will have values that lie on the right side. This accounts for most people (95%). Of course, in real life, the set of lab values taken from a population won’t fit this bell shape perfectly. But the similarity is close enough to be useful in creating reference ranges for most tests.
Once 95% of the values are accounted for, any remaining ones fall outside of the reference range. These are the 2.5% tails on either end of the curve above. Any lab value that falls in these areas is flagged as abnormal and brought to the attention of your healthcare provider.
In cases where only “low” values of the marker are of interest to doctors and point to a disease or disorder, the reference range will be one-sided on the left with a 5% tail as the abnormal range (instead of the 2.5% tails on either end). One example of such a test is sperm concentration, where only low values are likely to be of concern to doctors. In other cases, only “high” values may point to health problems, and the abnormal range will be the 5% tail on the right .
There are different ways to create a reference range, or reference interval, as it is commonly referred to by labs and doctors. The gold standard is to conduct a reference interval study .
A reference interval study analyzes a large number of test results from healthy individuals. These studies are often performed according to established international clinical and laboratory guidelines .
The first step is identifying a healthy reference population from which to take samples. The minimum number of people needed is 120, according to the guidelines. The absence of disease should be the main difference between this population and the patients for which the test will be applied .
Some tests need to have more than one reference range because of factors that affect the results. For example, men and women have separate ranges for total testosterone levels.
The main factors that determine whether multiple reference ranges are necessary are :
- Reproductive status (puberty, menstrual cycle, stage of pregnancy, and menopause)
- Race (e.g. prostate-specific antigen in African Americans)
- The time of day a sample is collected (e.g. random vs. first-morning spot urine sample)
Reference interval tests should be performed exactly the same way as in the lab. The time of day for sample collection, sample handling (such as the amount of time before the sample is analyzed), a person’s fasting status – must all be the same .
Reference populations are selected from volunteer blood donors, door-to-door contacts, medical students, and hospital outpatients. Potential participants take health questionnaires to exclude individuals by certain criteria such as physical activity levels, use of medications, medical history, and the presence of certain diseases. Samples are then taken from those included and the most extreme (outliers) test results removed .
The final step is to analyze the lab results and identify the upper and lower limits that 95% of the values fall within .
An alternative to reference interval studies is searching the existing patient data (collected by labs and hospitals and stored in electronic databases) and analyzing it. This type of analysis is called an “a posteriori” study, meaning that the research is done after (“post”) collecting the information. Since the data is easily accessible, these studies are less costly, less time-intensive, and can include a large number of patient samples. However, some samples may not come from healthy subjects, especially when it comes to samples from hospitalized patients .
Certain markers are assigned a decision limit (or multiple decision limits) that is [are] better than reference ranges in making diagnoses and treatment decisions. A decision limit is a cutoff point where values above or below are linked to an increased risk of developing specific diseases. Their purpose is to indicate when intervention is necessary to prevent disease. Decision limits may also be based on doctors’ clinical experience .
One test with multiple decision limits is total cholesterol. The upper limits for total cholesterol levels accepted by most labs and doctors are 200 mg/dL and 240 mg/dL. These limits were established by an expert panel for the National Cholesterol Education Program (NCEP), a program managed by the National Institutes of Health. Levels between 200 mg/dL and 240 mg/dL are associated with a moderate risk of heart disease, while levels above 240 mg/dL are associated with a high risk. These limits inform doctors on which treatment options to use, such as lifestyle and diet interventions or prescribing statins and other cholesterol-lowering drugs .
Because there is no universal reference range for most lab tests, ranges will vary from lab to lab. This means that it is possible to get a normal result from one lab and an abnormal result for the same test from another lab, and vice versa.
Reference ranges should be established for each marker by every lab. But the reality is that few labs carry out their own reference interval studies. Recruiting a healthy reference group and getting their informed consent is expensive and time-intensive, so most labs opt to instead use the reference ranges provided by the test manufacturers. This is preferable because the lab has to use only 20 sample tests to verify that the manufacturer’s range is accurate. Yet, some labs may skip this step [9, 10].
Labs are faced with numerous challenges when conducting reference interval studies: creating reference ranges for different subpopulations, rare sample types (e.g. cerebrospinal fluid), and tests that require multiple measurements .
Sometimes labs will use ranges established from previously published reference interval studies or even use ranges established by other labs .
Reference ranges are useful because they provide fixed values from reasonably healthy populations that healthcare providers can compare patient lab results to. This allows them to make informed diagnostic and treatment decisions.
Reference ranges are also appealing to patients (even if they may not know exactly what they mean) because they can see where their results fall in relation to upper and lower limits.
Reference ranges have some drawbacks. First, they do not take into account the results of large population research. This research can reveal different limits for an increased risk of mortality and disease. Decision limits are sometimes better for this reason.
Also, reference ranges do not leave much room for nuance and grey areas. Health is not a black and white phenomenon but instead exists on a spectrum.
Reference ranges do not take into account the uniqueness of and day-to-day, year-to-year variation in each person’s biology, environment, and genetics. Instead, they are arbitrary cutoffs based on how lab values are spread out across a “healthy” population, which is hard to define. A reference population may still contain people with an undiagnosed disease or condition that affects their lab test results.
Just because your lab value is outside the normal range does not mean that you necessarily have a disease or disorder. Indeed, and by definition, 5% of healthy individuals will have levels outside of the normal range.
Conversely, a normal lab value does not guarantee the absence of a disease or disease process.
It is important that your healthcare providers examine your lab results in relation to one another. They should also take into account your medical and family history as well as any previous test results (to identify trends).
It’s also important that the lab result is interpreted in light of the reason for requesting the test, such as a routine health check, managing a disease, or making sense of your overall symptoms.
For example, a provider may interpret the same total cholesterol value of 241 mg/dL differently for two patients if one has no history of heart disease, and the other previously suffered a heart attack and is on statins. The provider may be inclined to retest the patient with no past history of heart disease, but continue or increase the current statin dosage for the patient with heart disease.
Another important consideration is how far outside the normal range your lab test result is. For example, sodium concentrations in the blood are kept in a tight range. Lab results that are even slightly too high or too low can be dangerous even in the short-term .
For most tests, if your result is slightly outside of the normal range or the abnormal result doesn’t match the rest of your results, your provider may order additional tests or repeat the same test to confirm it. Factors that may cause abnormal values despite the absence of a disease or disorder include errors in processing and testing the sample, poor patient compliance in pre-test preparation (e.g. not fasting or stopping the use of prescription medications), and random fluctuations in the patient’s biology .
A normal reference range sometimes doesn’t give insight into the optimal range for a lab test. There are a couple of reasons for this. The most obvious reason is that by its very nature and design, a reference range is not intended to capture the optimal range. It is simply an interval that is based on samples from a pre-defined population.
Another reason reference ranges don’t provide any information on optimal values is because they rarely take into account research on the risk of developing certain diseases.
Lastly, the reference population may not have optimal levels. If the reference population isn’t sufficiently healthy, then this is reflected in the reference range.
Read more about optimal ranges here.
It is important that you mention any medications you may be taking to your healthcare provider as well as any other factors you think may affect your lab results.
To ensure consistency in your test results, you should use the same lab to perform your tests. If you are forced to switch labs, keep in mind that different labs may use different methods of testing your sample or have different reference ranges.