Why are my allergy symptoms so severe but cannot be diagnosed?

Diagnosing allergies depends not only on immunology, but also on skilled techniques and treatment experience.

By Theresa MacPhail

Translation | Qin Qikai

Parikh has been practicing allergy for more than 10 years and is a clinical assistant professor of pediatrics at NYU Grossman School of Medicine. She specializes in asthma care and does research on childhood asthma, but she sees a different type of allergy patient in her Midtown office.

Allergy is a medical subspecialty that relies as much on the clinician's experience and intuition as it does on modern diagnostic tools and the patient's biomedical history. That's why Parikh likens her day-to-day work to that of a detective. Diagnosing allergies is never easy. In a way, it's like solving a medical mystery.

Like Tolstoy's unhappy family, every allergy sufferer is different. No two allergy cases are alike, and a formal allergy diagnosis can take hours, days, weeks, months, or even years. This is because allergies are biologically complex, test results can be inconclusive, and the most common symptoms of allergies are very similar to those of other medical conditions.

Parikh was puzzled that I had not seen an allergist, given my allergies and my father’s death from a bee sting. She stared at me with a friendly smile and said, “I really think you should make an appointment, get tested, and we should fix this.”

Like many people with allergies, I had been hesitant to see an allergist. Since my symptoms were generally mild and easily controlled with over-the-counter antihistamines, it was easy for me to procrastinate on seeking more professional care. But I knew Parikh was right, so I eventually took her advice.

Typical atypical diagnosis

When I returned to her office, a full year had passed and my sinuses were really bothering me. As instructed at the appointment, I had not taken my antihistamine for a week. After a brief consultation, Parikh called her nurse into the exam room and gave me a standard skin-prick test to test my reaction to specific allergens and a quick breathing test to see if I had mild asthma in addition to my allergies. After three measurements, the nurse told me that my numbers were well within normal range: no asthma at all.

The nurse returns to the room with 3 small blue plastic trays. The trays contain white plastic applicators that look like 8-legged insects. Each leg ends in a prong that, when pressed gently on the upper arm or back, slightly scratches the skin and releases a tiny amount of allergen extract just below the first layer of the dermis. Allergists prefer to do the test on the arm so that the patient can see the reaction for themselves, because seeing their own skin react is often the first step for patients to understand their own condition.

In total, I was tested for reactions to more than 50 different allergens, including tree and grass pollens, as well as common food allergens like egg and wheat. The test also included a negative control (saline) to which normal skin shouldn't react, and a positive control (histamine) to which it should react, to ensure the test was working and the results were accurate. The nurse, who had marked my arm with the corresponding number so Parikh could easily see the results, carefully pressed the applicator against my forearm and upper arm, gently rocking it back and forth. I felt the plastic tip of the applicator dig in. Then the nurse left the room, leaving me to stare at my skin for the next 20 minutes, the average time it takes for skin cells to react to each allergen.

I immediately felt the histamine positive control kick in. The skin beneath the small scratch began to itch—mildly at first, then uncontrollably. I had to resist scratching it. I stared at my arm and saw a raised pink rash, like a mosquito bite, where the histamine had been. A sensitive person’s skin immediately reacts to the allergen with an inflammatory response at the injection site, what allergists call a “rash and flushing reaction.” Histamine released from a patient’s mast cells is the primary driving force behind this reaction. Typically, a patient is considered positive for sensitivity if a rash is larger than 3 mm and the erythema is larger than 10 mm in diameter. However, if the positive control produces a rash and erythema smaller than 3 mm, that size may warrant evaluation for other rashes. Any size rash is considered evidence of allergic sensitization, although smaller rashes may not represent true allergies.

I watched for reactions at the other numbers, but all I saw were dried marks of allergen extract dripping onto my pale skin. After the allotted time, Parikh knocked on the door and peeked in. She examined my arms closely, gave a “hmm,” and told me that my skin hadn’t reacted to any of the allergens. “That doesn’t necessarily mean you’re not allergic to any of these things,” she explained. “It just means we have to dig deeper, pardon the pun.”

Typically, an intradermal test is done after a failed skin prick test. An intradermal test uses a traditional syringe to deliver a small amount of the allergen extract deep into the skin. Parikh's nurse returned with a metal tray filled with 20 different syringes. She wiped my upper arm with an alcohol pad to remove the pen marks and any remaining extract, then gently pinched my skin to inject. One by one, the needles pierced my skin. When the nurse was done, my skin looked terrible. There were small drops of blood and raised bumps at the sites of the needle pricks. Then, I waited another 20 minutes. This time, as I stared at my arm, I remembered that I had an aunt in my family who had severe allergies. I wondered to what extent my immune response would be like hers or not. But nothing happened except the needle pricks and itchy skin where the histamine was injected.

After the allotted time, Parikh returned to the room, carefully examined my arm, and then sat down. "First," she said, "I want to emphasize that I believe you. I think you have clinical signs of an allergy. The problem is that your skin is 100 percent nonreactive. This happens all the time."

Parikh explains that in a small percentage of patients with significant respiratory allergies, skin cells are much more tolerant to allergens than sinus cells. In other words, I may have legitimate, convincing seasonal hay fever or perennial respiratory allergies, but it will never show up on any skin test. Cells on the skin and cells that make up mucous membranes may react very differently to the same allergen.

Parikh decided to give me a serological allergy test. In a serological allergy test, a patient’s blood serum is mixed with an allergen, and the resulting antibody response is checked. IgE antibodies are associated with atopy and are a predictor of allergic reactions; if they are activated after exposure to an allergen, the patient is considered sensitized to the allergen. (To complicate matters, standard diagnostic tools can only test for sensitivity and cannot always accurately predict whether a patient has or will develop an allergy.)

It took months for me to get my antibody test results, but by then my follow-up was online. When Parikh and I spoke again that May, it was during a particularly bad spring pollen season, and I was dealing with full-blown hay fever symptoms. My eyes itched and burned, sometimes spontaneously watering, as if I were crying. And despite taking allergy medication every day, my nose seemed perpetually blocked. I was desperate to know which trees or grasses might be responsible.

“You’re special!” Parikh announced at the start of the call, as if she were telling me I’d won a coveted prize. “According to these results, your blood showed no reaction at all. You had absolutely no reaction. In fact, your IgE antibody levels were very low. If I was looking at just these test results, I’d say you’re not allergic to anything.”

In that brief, silent moment, I felt a little crazy. If every test I had taken (skin prick tests, intradermal tests, blood tests for antibodies) came back 100% negative, did I really have allergies? Or had my itchy eyes and stuffy nose been my imagination all along? What was causing the significant nasal irritation that my ENT doctor had diagnosed years ago and that I experienced every spring, summer, and fall?

“I believe you have clinical symptoms,” Parikh said, as if she had read my mind. “I absolutely think you have allergies. It’s just that for some patients, their allergies are not mediated by IgE, and there is no easy test for that. Your body is reacting to something, that’s clear, but it’s not reacting through the IgE pathway. You have localized allergic rhinitis, and that’s my diagnosis.”

Basically, this means that the immune cells lining my nose and eyes react when they come into contact with allergens. For me, allergic reactions are targeted or "localized," not systemic or "systemic." My skin cells and their antibodies may not react to the pollen floating in the air in the spring, but the cells lining my nose and eyes do. Unfortunately, this also means there's no way to know which specific allergen is causing my symptoms. Technically, there's another approach we could try, but it would require placing each of the 50 allergens, one by one, in tiny amounts, directly on my eyes or nasal mucosa and waiting for the body to react. Unsurprisingly, neither Parikh nor I were willing to do that.

Parikh had exhausted all available avenues and still couldn’t crack this case: What triggered my allergies remained a mystery. She prescribed a daily dose of antihistamine nasal spray and eye drops and suggested that I stop taking oral antihistamines because they have side effects and my allergies are localized. If my allergies weren’t a systemic problem, there was no point in risking the side effects of allergy medications circulating throughout my body. It would be much better to treat the source of my symptoms, she suggested.

At the end of this months-long, very personal, but not uncommon, complicated allergy diagnosis story (filled with multiple negative allergy test results and a history based on patient self-report and clinical observation), my question to you is: Do I have a confirmed respiratory allergy?

The answer to this question depends on two things: the first is how we define what an allergy is and how we distinguish it from similar symptoms and medical conditions. Because my IgE levels are low and there is no evidence of a systemic immune system response, but my immune cells in my nose, eyes, and throat are definitely activated, then I have an allergy or type I hypersensitivity reaction, but I am not atopic.

The second is the different types of evidence we can accept to confirm an overactive immune response. If we only look at the results of clinical IgE skin and blood tests, there is no scientific "proof" that I am allergic. However, if we do further testing and have clear evidence of inflammation and irritation after exposure to pollen, then I have a localized allergic reaction.

My own story illustrates well (perhaps too well) that allergy diagnosis in the 21st century is a confusing puzzle. From the invention of the skin scratch allergy test in 1865 to the more recent development of fluorescent immunoassay tests for specific IgE antibodies, it has never been easy to diagnose or medically confirm an allergic reaction without actually observing symptoms. The milder or less obvious the reaction, the harder it is to detect, diagnose, or "prove" an allergy. Diagnosing allergies relies as much on skilled techniques and treatment experience as it does on immunology.

Why are skin tests not very accurate?

In the early 1980s, allergies were considered an underdeveloped field of medicine. In fact, medical students received little training in allergies. (It’s still the case today, with most interns spending only about two weeks studying allergic diseases.) “People didn’t even think of it as a science,” explains Sampson, the Kurt Hirschhorn Professor of Pediatrics at the Icahn School of Medicine at Mount Sinai and director emeritus of the Elliott and Roslyn Jaffe Food Allergy Institute in New York City. “They really didn’t think skin testing meant anything.” There’s a reason behind this lack of belief: Getting accurate results from a common skin-prick test is often a struggle.

First, the skin test must be done correctly, with both positive and negative controls. The negative control is the diluent used in the mixture, to which normal skin should not react; the positive control is histamine, to which normal skin will react, forming a rash. Second, skin and intradermal tests must be done precisely. For respiratory and food allergy skin prick tests, the applicator must penetrate deeply enough to deliver the allergen to the correct location in the skin. If the prick is too deep and the patient bleeds, it may be considered a false positive (especially if the intradermal puncture is too deep). If the scratch or injection is too close together, the results may be difficult to identify, as it may not be clear which specific allergen caused the reaction. This is much better if a high-quality standardized allergen extract is used, but this is more difficult than it looks.

Part of the problem with skin-test accuracy is that several different companies make extracts for skin-prick and intradermal testing, and these extracts can vary significantly in allergen concentration (how much allergen is in each dose) and composition (the type of solution in which the allergens are mixed). Because there is no standardization of commercial preparations of allergens for skin-prick testing, the amount of allergen actually injected can vary, making it difficult to know how much has penetrated the skin. Not enough or too much can affect the results. Sometimes, the inactive ingredients used in different extracts can themselves cause a reaction, resulting in a false positive. There is a high risk of injecting too much allergen in an intradermal test, which can result in a false positive or a more severe reaction. (In fact, all allergy skin testing must be done in a clinical setting in case a patient has a severe allergic reaction to one of the allergens.)

Today, the allergen extracts used in most skin tests are either single allergens or mixtures of similar allergens (e.g., a test for "grass" allergy may contain allergens from multiple grasses). This makes it difficult to accurately interpret the results, especially if the extract lacks a certain plant allergen that is common in the patient's geographic area. Skin test results are collected, averaged, and then used to standardize allergen extracts (this may seem like circular logic, but never mind), and to conduct epidemiological and pharmacological studies, which is one of the reasons why it is so difficult to get accurate figures on the number of people with allergies.

Even if everything is done correctly to produce a high-quality allergen extract, the reliability of skin prick and intradermal test results can be affected by “the skill of the person, the testing instrument, skin color, and potency of the extract,” as well as “testing site, age, body mass index (BMI), medications, allergen immunotherapy, circadian and seasonal changes, menstrual cycle, stress, and anxiety.” Skin test results can also be affected by taking antihistamines, steroids, antidepressants, sedatives, and other medications that affect immune system function. Because of this, allergists often ask patients to stop taking these medications for a few days to a week before testing. If skin testing is to be performed anyway, such as for a patient who cannot discontinue medication for medical reasons, all negative results must be considered likely false negatives, although positive test results are still considered positive.

Infants are also difficult to skin test. Their skin doesn’t show reactivity until around 3 months of age, and even after that, their results can be harder to read than those of adults and are considered more inconclusive. That’s why doctors in the early 20th century often defaulted to using the P–K test to test infant patients for allergen sensitivity.

Finally, and perhaps most importantly, there is no standardized or universally accepted system for interpreting skin tests or recording and collecting results. There are some general recommendations for clinicians, but each allergist makes his or her own decisions about how to best interpret the results of skin prick and intradermal allergen tests. This is why it is important to have a trained allergist, rather than a general practitioner, administer and interpret skin tests. It may take years of experience to accurately interpret skin test results.

Additionally, skin tests can only be performed on "normal" or currently non-reactive skin; otherwise, the allergy results are nearly impossible to interpret. As you can imagine, this makes it difficult for people with skin allergies to get accurate results.

Just do your best

When I spoke with Dr. Peter Leo, one of the top experts in atopic dermatitis (eczema), he explained that common skin prick tests are usually not suitable for patients with skin allergies. In his clinic, skin testing is very time-consuming. Leo will apply 80 to 120 stickers containing various allergens to the patient's back and leave them in place for 48 hours.

"It's a little bit of a drag," Leo says. "On Monday, you put the stickers on the patient. On Wednesday, we take them off. Then, the patient comes back on Friday and we look at the results that appeared on the skin in the last 96 hours. It's more invasive for the patient, but it does give us important information." Once the final reading process is complete, Leo gives his patients a list of various products to avoid, based on any positive reactions. Sometimes triggers are hidden in shampoo, soap, or other everyday items. It can take a while to determine which allergens are actually causing the allergic reaction, as it can take up to two months for a patient's skin to "calm down" after they stop being exposed to those substances.

For a patient to be diagnosed with atopic dermatitis rather than simply having a positive skin test, three criteria must be met. First, the patient must have eczema, a rash, or inflammation of the skin, not just blisters or bumps. Second, the patient must have symptoms of itching. Third, the rash and itching must be chronic, or come and go, not just occasionally. Atopic dermatitis is mostly diagnosed in children and usually disappears as they reach adulthood, but it can also get worse in adult patients. Leo explained to me that current research may lead to the development of new diagnostic tests for atopic dermatitis subtypes based on immunophenotyping, a test used to study the different proteins expressed by each cell. But for now, patch testing is the only way he can determine possible allergic triggers for his eczema.

For respiratory allergies and food allergies, when the results of skin tests are inconclusive or inconsistent, there is the option of testing for specific IgE antibody reactions to allergens. When Sampson first began his career, allergists were also using radioallergosorbent testing (RAST) to check patients’ blood for IgE reactivity to different allergens. The test is a radioimmunoassay that uses a small amount of radioactive antigen mixed with the patient’s serum. If the patient is allergic to the antigen, then the patient’s IgE antibodies bind to the antigen; the free-floating antigen is measured by a gamma counter (the less free-floating antigen, the more active the IgE, and therefore the more sensitive the patient is to the antigen).

Today, RAST has been largely replaced by newer immunoassays, but in common parlance, the term "RAST" is used to refer to other blood tests, even by allergists. If you need a blood test like I did, your allergist will usually prescribe an enzyme-linked immunosorbent assay (ELISA) or the more popular and more accurate fluorescent enzyme immunoassay (FEIA).

In an ELISA test, an antigen and an enzyme-labeled antibody are mixed with the patient's serum to detect antibody responses to a specific allergen. ELISA tests are rapid and inexpensive, but require an allergist to test individual allergens or groups of allergens separately. They also require manual testing. The FEIA test uses a similar approach to RAST and ELISA, except that the antibody label used to measure antibody responses to a specific antigen is a fluorescent enzyme. The FEIA is fully automated, less prone to error, and can screen for multiple allergens at once. The advantage of the standard FEIA (commercially known as ImmunoCAP) is that it measures allergen-specific IgE (sIgE) rather than total serum IgE levels. It also reduces (but does not completely eliminate) the chance of false positives due to unexpected cross-reactions or genetic similarities of antigens (e.g., different nuts from the same family).

However, even if a serum test “works” and shows positive sIgE activity, it does not necessarily mean that the patient is allergic to a specific allergen, only that they are reactive to that antigen. Sampson reminded me that relying on blood tests to diagnose food allergies is a very bad idea. He noted that when people with positive blood test results are tested with oral food challenges, “the proportion of positive tests far exceeds the number of people who actually have a clinical reaction.” In fact, both skin and blood tests for food allergies have false-positive rates of 50% to 60%.

Decades passed before allergy researchers were finally able to show a strong correlation between the level of sIgE antibodies in blood tests, the size of the rash produced by skin tests, and the likelihood that a person would have an immune response if they ingested a specific food or were exposed to a respiratory or skin allergen. But this new understanding also created some confusion for patients: They often conflated the level of IgE antibodies in their blood or the size of the rash after a skin test with the severity of their allergy. On social media sites, patients often share photos of their skin tests to emphasize the severity of their allergies. In other words, they equated a test that only measured sensitivity or the likelihood of a reaction with being able to accurately assess the degree of allergic reaction they would experience if they were exposed to an allergen in normal life. Unfortunately, this is not the case.

“There is not a good correlation between the size of the rash or the antibody levels on a skin test and the severity of the reaction you’re going to have,” Sampson explained to me. “The only correlation is the likelihood of a reaction, not how bad the reaction is.”

This is why the gold standard for diagnosing food allergy (both then and now) is the double-blind placebo-controlled oral food challenge, commonly referred to as OFC.

Complex food allergy diagnosis

Although OFCs are the best way to confirm a food allergy, they are the tests least likely to be performed. The reasons for this vary, but some of the most common reasons are: OFCs are expensive because the tests need to be performed in a hospital or other health care setting that is capable of caring for patients experiencing allergic reactions; the time required to complete an OFC because each allergen needs to be tested separately and the number of tests is increased over days or weeks; and there are risks because these tests can cause severe reactions in patients, especially small children. OFCs are particularly stressful for parents and can cause a lot of anxiety in children. In the absence of an OFC, most food allergies are diagnosed through a detailed history, physical examination, skin prick tests, and sIgE blood tests. (Tests that are not recommended include: intradermal tests because they can cause severe reactions; serum total IgE assays, which only measure the presence of generalized, not specific, allergic reactions; IgG assays because everyone will have an IgG reaction to food proteins; or any other test that claims to evaluate for food allergies.) Generally speaking, an experienced allergist can accurately diagnose most food allergies. However, without an OFC, there is no way to confirm with absolute certainty whether someone has a typical food allergy.

In addition to these challenges, Sampson points out that not enough testing has been done on adults. Most allergy studies, especially those related to food, are done in young children (which makes sense, since most patients first develop food allergies during infancy or early childhood). This makes interpreting the findings in adults more difficult and can lead to confusion.

Diagnosis of food allergy is further complicated because its main symptoms are very similar to other gastrointestinal diseases or diseases that are completely unrelated to allergies. There are also some food-related diseases that are not mediated by IgE at all, such as food protein-induced enterocolitis syndrome, food protein-induced proctocolitis syndrome, and eosinophilic esophagitis. Enterocolitis syndrome is an immune-induced inflammation of the small intestine, usually triggered by milk or grains, which can cause vomiting and diarrhea. Food protein-induced proctocolitis syndrome is an immune-induced inflammation of the colon, usually caused by milk, which can cause blood in infants' stools. Eosinophilic esophagitis is an inflammation caused by an excess of eosinophils (a type of white blood cell) in the esophagus and is triggered by specific foods. These rare immune-mediated diseases (affecting approximately 0.5%, 0.12%, and 0.0005% of the population, respectively) usually appear in infancy or early childhood but are not driven by the action of IgE antibodies. "Unfortunately," Sampson explains, "there are no good tests for these diseases."

Part of the problem with food allergy diagnosis, and allergy diagnosis in general, is that we still don’t really understand the immune mechanisms behind many allergies, Sampson told me. And, as allergy rates continue to rise, that means we don’t have the diagnostic tools to keep up with the magnitude of the problem, either.

The skin prick test is a great example. It remains the most common, easily available, and cheapest test for initial allergy diagnosis. But 8% to 30% of people who have a positive skin test (or develop a rash) do not display any allergy symptoms. Still, skin test results are an important indicator of allergies, as studies show that 30% to 60% of patients who are sensitized to a particular allergen will develop an allergy. If you can remember only one thing, let it be this: blood and skin tests only show sensitivity to a specific allergen; they never confirm an allergy. Suffice it to say that any skin or respiratory allergy should be diagnosed by an allergist based on the patient's medical history and the symptoms that occur when the patient is exposed to the allergen in a natural setting.

The objective science of allergy diagnosis is fraught with subjectivity. Many allergists rely on their intuition, informed by years of clinical experience, to interpret skin test results and diagnose allergies. As Parikh says, in the 21st century, interpreting allergy test results is both a science and an art.

This article is authorized to be excerpted from Chapter 2 "How are allergies diagnosed (and not diagnosed)?" of The Truth About Allergies (CITIC Press·Nautilus, April 2024 edition), with some deletions and subheadings added by the editor.

About the Author

Theresa MacPhail: Medical anthropologist, writer, and associate professor in the School of Humanities, Arts, and Social Sciences at Stevens Institute of Technology. She received her Ph.D. in medical anthropology from the University of California, Berkeley. She is a member of the American Anthropological Association and a 2018 NEH (National Endowment for the Humanities) public scholar. Her research and writing focus on global health, biomedicine, and disease.

About the Translator

Qikai Qin is a postdoctoral researcher at Massachusetts General Hospital and Harvard Medical School. He received his Ph.D. from ShanghaiTech University (School of Life Sciences and Technology and iHuman Institute). He has translated "The Elegant Guardian".

Special Tips

1. Go to the "Featured Column" at the bottom of the menu of the "Fanpu" WeChat public account to read a series of popular science articles on different topics.

2. Fanpu provides a function to search articles by month. Follow the official account and reply with the four-digit year + month, such as "1903", to get the article index for March 2019, and so on.

Copyright statement: Personal forwarding is welcome. Any form of media or organization is not allowed to reprint or excerpt without authorization. For reprint authorization, please contact the backstage of the "Fanpu" WeChat public account.