Should You Believe a Statistical Study?

ShouldYou Believe a Statistical Study?

MårildK, Størdal K, Bulik CM, et al. (2017). Celiac Disease and AnorexiaNervosa: A Nationwide Study. Pediatrics,139(5),e20164367

Inconducting the study, the researchers’ goal was to determinewhether celiac disease (CD) is associated with the diagnosis ofanorexia nervosa (AN) because previous studies had suggested anassociation between the two conditions (Mårild, Størdal, Bulik, etal., 2017). The researchers used a population of 17,959 women with CDand a control population of 89,379 (Mårild, et al., 2017). The studywas a cohort and case-control that followed the two groups of womento determine whether they would or not develop AN. In addition, thestudy is a quantitative research that dealt mainly with statisticaldata.

Allthe seven researchers were involved in the study and carefullyexamined the historical records of all the study subjects chosen todetermine the development of AN. The possibility of having bias issignificantly reduced because all the researchers involved werequalified and justified to take the research, given that they wereall from health departments in respective institutions (Rapaport,2017).

Thesample used in the study was obtained from medical histopathologyrecords of 28 pathology departments in Sweden placed since 1989(Mårild, et al., 2017). There was a 20-year follow up in the cohort.Consequently, there are little chances of bias because theresearchers did not have prior knowledge of the subjects rather theywere obtained from already completed medical treatments.

Thesample used in the case might have some bias because it reflects thecase in Sweden and not any other country and region in the world. Itmight be difficult to make a generalized conclusion about therelationship between AD and CD, given that the small population ofwomen chosen in Sweden cannot possibly represent the entirepopulation of women in the world.

Indefining and measuring the variables of interest in the study, someproblems could have occurred. For instance, by using past records,the researchers could have made errors because they were dealing witha population of individuals they did not know. In fact, any errormade by the departments was likely to affect the results. In otherwords, they were using secondary data instead of going for their ownnew data.

Apossible confounding factor could be giving birth or pregnancy. Forinstance, when women become pregnant or give birth, they are likelyto develop eating disorders such as AN. In addition, pregnancy andbirth giving are also likely to affect the development of CD(Lebwohl, Ludvigsson &amp Green, 2015). Consequently, this factorcould be a confounder, thus cause bias in the results of the study.

Theresearchers did a good job in data presentation as they used bothstatements and visual aspects such as graphs and charts, which madeit possible for the audience to understand the content.

Teresearchers conclude that the bidirectional association between thetwo diseases is possible and needs attention because bothmisdiagnosis and under diagnosis can cause unnecessary morbidity.This conclusion is logic and reasonable, given the high morbidityrate in the diseases. In addition, the results make some practicalsignificance because they call for additional studies to furtherexamine the relationship between AD and CD in more advanced researchprojects.

References

Lebwohl,B., Ludvigsson, J. F., &amp Green, P. H. (2015). Celiac disease andnon-celiac gluten sensitivity. BMJ,351,h4347

Mårild,K., Størdal, K., Bulik, C. M., et al. (2017). Celiac Disease andAnorexia Nervosa: A Nationwide Study. Pediatrics,139(5),e20164367

Rapaport,L. (2017). Anorexianervosa diagnosis more common in women with celiac disease.Retrieved fromhttp://www.reuters.com/article/us-health-celiac-anorexia-idUSKBN1752EG