A recent study by Cecilis-Morales et al. (2017) focuses on the main outcome measures of incident CVD (cardiovascular diseases) and cancer in geriatric patients. The research question of the analyzed article aims to explore the relationship between active commuting and cases of cancer, CVD, and all-cause mortality. A closer examination of the research shows low internal validity as the study fails to describe the exposure procedure and strategies used to control the confounding variables.
The study design chosen for this research is a population-based cohort. The initial examination of the scope of the research procedures, methodology, and objective suggests that the design utilized is appropriate. According to Cade and Hutchinson (2015), population-based prospective cohort study designs allow investigating associations between a phenomenon and its outcome. In the clinical setting, such research design works best for determining the onset of a particular health condition. That’s why Celis-Morales et al. (2017) rightfully derived a representative of approximately 2,900 participants aged 49-60. Such sample helps to represent the addressed population in the study – vulnerable geriatrics.
To evaluate the internal validity of the study, five criteria featured in the Critical appraisal checklist for cohort studies by Joanna Briggs Institute (2017) were utilized. First, the two groups used in the study appear to be similar, being recruited from the same population. In a setting of UK Biobank, all the participants came from 22 sites with a fairly equal gender distribution: 52% females and 48% males. Inclusion criteria were explicitly stated in the study while sociodemographic factors were identified as confounding variables. Second, the research does not contain a detailed description of the exposure procedure, failing to provide an in-depth overview of the measures taken for the experimental and control group. Such a conclusion compromises the internal validity of the study by minimizing the amount of data available regarding the clarity and rationality of the procedures taken. Furthermore, it is not possible to determine whether exposure of interest occurred.
Third, though little information is provided regarding the procedure itself, the exposure appears to be measured in a reliable way. The research clearly indicates the method of exposure measurement while establishing a golden standard for comparison based on the past findings. Intra-observer reliability is mentioned in the study, yet there are no mentions of interobserver reliability as such. Forth, Celis-Morales et al. (2017) determined the confounding factors in the study (baseline characteristics) to point the possibility of the intervention exposure effect and confirmation bias. This step is crucial for establishing the internal validity of the study because it rules out the distinction between the comparison groups. Fifth, even though the authors clearly stated the confounding factors, the strategies applied to deal with their effect have been explicitly described neither in the data analysis nor in the study design. The article does not include evidence of stratifying or sampling matching to balance the consequences of the confounding factors.
In conclusion, the study by Celis-Morales, C. A. et al. (2017) does not satisfy high standards for internal validity, failing to include an in-depth description of exposure procedures and strategies to control the confounding variables. Such finding compromises the reliability of the results along with their interpretation. Further research should be conducted to rule out the inconsistencies identified in this critical appraisal to achieve the initial objective.
Cade, J. and Hutchinson, J. (2015) ‘Study design: population-based studies’, in Lovegrove, J. A. et al. (eds.) Nutrition research methodologies. Hoboken: Wiley, pp. 13-27.
Celis-Morales, C. A. et al. (2017) ‘Association between active commuting and incident cardiovascular disease, cancer, and mortality: prospective cohort study’, BMJ, 357, pp. 1-7.
Joanna Briggs Institute. (2017) Critical appraisal checklist for cohort studies. Adelaide: Joanna Briggs Institute.