The study is of particular interest as it studies a novel risk factor for a prevalent disease considered to have a multifactorial aetiology. However, we can make a number of considerations to assess its contribution:

1. The study has a cross-sectional epidemiological design, i.e. disease and exposure are measured at the same time. Therefore, there is no temporal relationship and it is not possible to establish proof of causality between sodium intake and disease. This would require cohort studies, first the exposure (sodium intake) and then the occurrence of the disease (atopic dermatitis). 
2. The disease is measured in several ways: prevalent, active, severe, with long time periods related to many risk factors. In addition, atopic dermatitis does not have a laboratory diagnosis, its diagnosis depends on medical consultation, and there may be biases in access to medical care.  
3. Being a cross-sectional design, it does not measure disease risks, only associations from odds ratios. These resulting odds ratios are modest. The sizeable sample size of the population studied means that significance is achieved. However, with regard to the results of the univariate analysis (table 2) there are no significant associations, and in the sensitivity analysis there are no associations with active atopic dermatitis either.  
4. No risk factors were considered in the logistic regression models, including allergic diseases such as asthma, allergic rhinitis, food allergy, atopy, atopic dermatitis in childhood, family history of atopic dermatitis, history of respiratory infections, pneumonia, sinusitis or tobacco use.  

In conclusion, this is an exploratory study with an interesting hypothesis and, given its limitations, the importance of sodium intake in the aetiology of atopic dermatitis deserves to be studied in more depth with cohort or experimental studies.