Question 6
Yes, at the beginning of the study, the participants were not included in developing the outcome measures or setting the research questions. The respondents did not participate in the development and implementation of the study or the interpretation of the results. In their study, Lyall et al. (2017) developed a prospective study to determine the correlation between active commuting and causes of mortality, including cancer and cardiovascular disease (CVD). Making the respondents free of the outcome enhances the credibility and reproducibility of the result are essential.
Question 7
Yes, the outcomes were measured reliably and validly. Lyall et al. (2017) employed Cox proportional hazard models to measure the correlation between health outcomes and active commuting. These models were valid and reliable since they are developed to precisely determine the association between a patient’s survival time and more than one variable. Data from both active and non-actively commuting participants were recorded on an electronic questionnaire. This method enhanced the objectivity of the feedback as it minimized the researchers’ influence on the participants’ responses. The use of electronic questionnaires also provided anonymity to the participants, which might have positively affected the nature of the responses. Also, since the study involved thousands of respondents, electronic questionnaires allowed the researchers to reach all the participants easily. During the study, all the analyses were adjusted for confounders, including comorbidities, Townsend deprivation index, ethnicity, age, and sex. The analysis of the body mass index (BMI) was also coded based on the classification developed by the World Health Organization (Lyall et al., 2017). Subsequently, Schoenfeld residuals were used to test for the proportional hazard assumption (Lyall et al., 2017). The researchers then conducted all analyses on STATA 14 statistical software. Since the software is the latest version, its results are more reliable; thus, improving the validity of the research findings.
The data used was drawn from the UK Biobank and is a representation of the population. However, it may have a selection bias to healthy volunteers because it did not capture the prevalence of comorbidities and obesity among the participants. Nonetheless, the study has face validity since the electronic questionnaire used seemed to meet the research objectives. Additionally, the study had high reliability since it had a follow-up time of up to five years and sourced information from the National Health Service and Scottish morbidity records (Lyall et al., 2017). Overall, the outcomes were measured in a reliable way, which improves the reproducibility of the findings.
Question 8
Yes, follow-up time was sufficient to allow for outcomes to occur. In the research, the average follow-up period was five years for cancer mortality and CVD, while cancer and incident CVD was 2.1 years (Lyall et al., 2017). Also, a total of 2430, 1110, and 3748 respondents died, developed incident CVD, and cancer, respectively, during the follow up (Lyall et al., 2017). In this study, follow up was conducted as a regular part of the research and to ensure it was achieving its targeted goals.
Question 9
Yes, the follow-up was complete. For England and Wales, mortality had been registered until February 17, 2014, while in Scottland, the data available up the end of 2012 (Lyall et al., 2017). In this regard, follow-up for mortality analysis was censored at these dates or any time before that deaths occurred. For English and Scottish respondents, hospital records were available until the end of June 2012, whereas in wales, the last entry was at the beginning of March 2012 (Lyall et al., 2017). It was during these dates that the researchers ended follow-up for CVD incidences. Therefore, the researchers reported a loss of the following.
However, there were no differences among exposure groups after the researchers lost follow-up due to the death of some respondents. According to Lyall et al. (2017), the 10th revision of the international classification of disease (ICD-10) coded 164, 163, 161, 160, or121 was used to categorize CVD incidents. On the other hand, ICD-10 code D3.7-9, and C0.0-C9.9 were used to measure cancer cases. From the study, the follow-up in all cases ended at the date of the last hospital entry record. During the follow-up, the researchers used median values to categorize participants’ weekly commuting distance. The researchers also employed a questionnaire to assess the baseline physical characteristics of the respondents. The participants had also worn a tri-axial accelerometer on their wrist, which helped in accessing physical activity during follow-up. Accelerometer and cardiorespiratory data from a subset of 54, 378, and 39022, respectively, were assessed subsequently analyzed (Lyall et al., 2017). Since there was a small number of respondents in the subset, the data were not included as covariates during the development of outcome models. Nonetheless, the researchers adjusted for occupational and leisure-time physical activities, such as strenuous sport, and walking for pleasure.
Question 10
Yes, strategies to address incomplete follow-up were utilized. In the study, the researchers performed an analysis of the 2430 respondents who died during follow-up. It is through this analysis that they discovered that cycling and mixed-mode of commuting were directly associated with reduced risk of all causes of death and cancer.