Allergen immunotherapy works for hay fever, modestly, and only over years
A 160-study meta-analysis found real but moderate symptom reductions while patients are on treatment. The evidence for lasting benefit after stopping is much thinner, and rests mainly on grass.
| Group | Value (SMD) |
|---|---|
| Symptom score | 0.53 (0.42 to 0.63) |
| Combined symptom and medication | 0.49 (0.3 to 0.69) |
| Medication score | 0.37 (0.26 to 0.49) |
Allergen immunotherapy — repeated controlled exposure to the thing a patient is allergic to, by injection or under the tongue — produces a real but moderate improvement in hay fever symptoms while a patient is on it. The pooled effect across 160 studies is a standardised mean difference of -0.53 for symptom scores (95% confidence interval -0.63 to -0.42) and -0.37 for medication use (-0.49 to -0.26) [s1]. It is the only treatment for allergic rhinoconjunctivitis that targets the underlying mechanism rather than the symptoms, and the only one with any claim to modifying the disease [s2]. The evidence that the benefit persists after treatment stops is much weaker than the evidence that it exists during treatment.
What the systematic review found
The review that informed the European Academy of Allergy and Clinical Immunology's guideline searched nine international biomedical databases for published, in-progress and unpublished evidence, screened 5,960 studies and included 160 [s1]. The primary outcomes were symptom scores, medication scores and combined symptom-and-medication scores, pooled with random-effects meta-analysis [s1].
While on treatment, all three moved in the same direction: symptoms -0.53 (-0.63 to -0.42), medication -0.37 (-0.49 to -0.26), combined -0.49 (-0.69 to -0.30) [s1]. These results were robust to the sensitivity analyses the reviewers had specified in advance [s1].
A standardised mean difference of around 0.5 is conventionally described as a moderate effect. It is not the difference between having hay fever and not having it. It is the difference between a bad season and a manageable one, on average, across a population that includes people who responded well and people who did not.
On what happens after treatment ends, the review is explicit that the body of evidence is "more modest", and suggests a benefit for symptom scores only [s1].
What the guideline made of it
The EAACI guideline, developed under the AGREE II framework and informed by that review, recommends both subcutaneous and sublingual immunotherapy for seasonal and perennial allergic rhinitis on the basis of short-term benefit [s2]. Three things in its wording matter more than the recommendation itself.
First, it calls for a product-specific evaluation of the evidence rather than treating immunotherapy as one intervention [s2]. Different manufacturers' extracts have different evidence behind them, and a pooled estimate across 160 studies does not transfer automatically to a particular product.
Second, the strongest evidence for long-term benefit is documented for grass immunotherapy, and especially the grass tablets [s2]. That is a narrow claim about one allergen, not a general property of the treatment.
Third, to achieve long-term efficacy the guideline recommends a minimum of three years of therapy [s2]. This is the practical fact that decides most cases: the treatment is a multi-year commitment, and stopping early forfeits the part of the benefit that distinguishes immunotherapy from an antihistamine.
The guideline closes by naming the gaps in its own evidence base, particularly around long-term benefit and use in children [s2].
Asthma is a different question with a weaker answer
The companion review on allergic asthma included 98 studies [s3]. Short-term symptom scores were reduced, SMD -1.11 (95% CI -1.66 to -0.56), and short-term medication scores reduced, -1.21 (-1.87 to -0.54) [s3]. Both estimates are larger than the rhinitis figures — and for both, the reviewers report evidence suggestive of publication bias, which is the standard signal that small negative trials are missing from the literature [s3].
Short-term combined symptom and medication scores showed no reduction at all, SMD 0.17 (-0.23 to 0.58) [s3]. On the outcomes clinicians use to judge asthma — asthma control, exacerbations, lung function, non-specific airway hyperreactivity — there were no consistent effects [s3]. Subcutaneous immunotherapy improved quality of life and reduced allergen-specific airway hyperreactivity; sublingual immunotherapy did not [s3].
That is a considerably more equivocal picture than the rhinitis literature, and it is worth separating, because "immunotherapy works" is often stated without specifying for what.
Safety
The asthma review found a modest increased risk of adverse events overall [s3]. Systemic adverse events were relatively uncommon but more frequent with the subcutaneous route than the sublingual one, and no fatalities were reported across the included studies [s3]. That asymmetry — injections carrying more systemic risk than drops or tablets — is the reason subcutaneous immunotherapy is given in supervised settings with observation afterwards.
The limited cost-effectiveness evidence was mainly available for sublingual immunotherapy, and suggested it is likely to be cost-effective [s3].
What this leaves undecided
Who benefits most is not answered by these reviews. They pool across allergens, products, routes, ages and severities, and the guideline's response to that heterogeneity is to push the evaluation down to individual products [s2]. Whether a person with moderate hay fever controlled adequately on medication should undertake three years of treatment for a moderate additional improvement is a preference question, not an evidence question — the evidence gives the size of the effect, not the value of it.
What is reasonably settled: the treatment does something, the something is moderate, it takes years, the long-term evidence is strongest for grass, and the asthma case is weaker than the hay fever case [s1] [s2] [s3].
Immunotherapy is a prescribed treatment with a dosing schedule and a monitored safety profile. Decisions about starting, continuing or stopping it belong with a clinician who can assess the specific allergen, product and patient.
Sources
- Allergen immunotherapy for allergic rhinoconjunctivitis: A systematic review and meta-analysis — Allergy, 2017-05-11
- EAACI Guidelines on Allergen Immunotherapy: Allergic rhinoconjunctivitis — Allergy, 2017-09-23
- Allergen immunotherapy for allergic asthma: A systematic review and meta-analysis — Allergy, 2017-05-23
Sources
- Allergen immunotherapy for allergic rhinoconjunctivitis: A systematic review and meta-analysis — Allergy , May 11, 2017
- EAACI Guidelines on Allergen Immunotherapy: Allergic rhinoconjunctivitis — Allergy , September 23, 2017
- Allergen immunotherapy for allergic asthma: A systematic review and meta-analysis — Allergy , May 23, 2017
More on
Half the lengthening of North American pollen seasons is attributed to human forcing
A detection-and-attribution study across 60 pollen stations found seasons 20 days longer and concentrations 21% higher since 1990 — and separated how much of that trend is anthropogenic.
Only about 3% of children follow the 'atopic march' from eczema to asthma to hay fever
Two birth cohorts totalling 9,801 children were sorted into eight developmental patterns. The classic progression was one of them, and one of the smallest.
Local honey did not relieve hay fever in the trial designed to test it
A randomised trial gave one group unfiltered local honey and found no relief beyond placebo. A later trial found a benefit — using a dose of about 70 grams a day alongside an antihistamine.
The older antihistamines are not gentler. They are the ones with the safety record
A European task force argued the first-generation drugs sold over the counter should not be. On symptom relief the newer ones differ from each other too — and the most familiar performed worst.