Impairment
As a secondary outcome measure, allergy-related impairment in quality of life was assessed using the short form of the Rhinitis/Rhinoconjunctivitis Quality of Life Questionnaire. This is the German translation of the Mini Rhinoconjuctivitis Quality of Life Questionnaire (MiniRQLQ) 29. Through 14 items, the MiniRQLQ assesses the degree of impairment caused by allergy-related symptoms. Two items refer to practical problems (e.g., ”having to rub nose or eyes”) and three items each refer to limitations in activities (e.g., ”sleep [difficulty sleeping through the night and/or falling asleep at night]”), nasal symptoms, eye symptoms and other complaints (e.g., ”irritability”). Respondents indicate the degree of impairment for each item on a seven-point Likert scale (0 = “not at all” to 6 = “extremely”). The instruction was slightly adapted so that the items referred to the last 14 days, instead of the last 7 days, to ensure comparability with the primary endpoint. The MiniRQLQ is a reliable and valid questionnaire that is sensitive to symptom changes29,30. Cronbach’s alpha was α = .90 at T1, α = .91 at T2 and α = .90 at T3.

Medication Use

To assess whether participants took any medication against their allergic symptoms, they were asked to choose one of three options: 1) “I regularly take medication against my allergic symptoms”; 2) “I take medication against my allergic symptoms on demand”; 3) “I don’t take any medication against my allergic symptoms”.

Additional questionnaires

In addition, we also assessed participants’ expectations and hopes regarding the placebo treatment, the extent to which they felt informed about placebos, their actual knowledge about placebos, their self-efficacy beliefs, and their beliefs about a potential relationship between their allergic symptoms and COVID-19 (see supplement).

Statistical Analyses

After data screening, we performed an intention-to-treat analysis by estimating the missing values of the two persons who dropped out using the expectation maximization procedure according to methodological recommendations 31,32. For the main analysis, we conducted two separate analyses of covariance (ANCOVA), with the severity and frequency of allergic symptoms at T2 as the dependent variable (DV), treatment (OLP vs. TAU) as a between-subjects factor, and baseline allergic symptoms (T1) as a covariate. For the secondary endpoint, we performed another ANCOVA with T2 impairment as the DV and T1 impairment as a covariate. Subsequently, we added medication use (no medication vs. regular medication vs. medication on demand) as an additional between-subjects factor to the aforementioned ANCOVAs. Type-1 error levels were set at 5%. All analyses were conducted using IBM SPSS Statistics Version 27.