Minitab Standard Deviations
Topic 1: Energy Drinks – are they worth the risk?
Research Methods Report
Your report must consist of the sections described below.
Use this section to introduce your topic. In particular, address the following:
- Why were researchers studying this topic?
- Summarise the research questions that the papers address.
- How did the studies hope to address one or more of deficiencies (if any) in the current body of knowledge?
- Materials and Methods
In this section, address the following:
- What type of study designs were chosen by the authors to address research questions? Name these study designs.
- How were subjects chosen for the studies? How many subjects were there? What do we know about them (e.g. age)?
- What efforts (if any) were made to control for potential confounding factors?
- What variables were measured? Are there any issues which relate to the internal reliability and validity of the measurement processes used?
- What types of statistical analyses were employed?
- Outcomes and contributions
In this section discuss the following:
- What were the key results of the studies? What significance level was used?
- How do these results contribute to what is known about the area of study? Summarise briefly what the authors say in that regard.
- What are the factors which may influence or limit the drawing of meaningful conclusions from the studies? Summarise the key issues which relate to the internal and external validity of the
- For one of the statistical procedures employed in one of the papers, discuss power and effect size using appropriate Minitab output. Instructions on how to do this are given on page 7 of this Put full details of your power calculation in an Appendix (6 marks)
- How did the authors choose to present their results? Were their tables and displays easy to follow? Were they helpful? Comment briefly.
In this section draw some overall conclusions and provide your answer to the topic question. There is no word limit. As a guideline, one to two short paragraphs will be sufficient.
Power Calculation in Section 3 Instructions: Place full details of your calculation (Steps 1 to 5 below) in an Appendix. Also include a report
style discussion in Section 3 of the report.
First of all, check whether the authors themselves discuss their study power.
Step 1.Identify a two-sample t-test or a paired t-test of interest among the results in one of the papers (you can use more than one if you wish but it is not strictly required). Hint: Select a paper and a scenario in that paper that makes most sense to you and so will be easy to explain.
Step 2.Decide whether a two-sided or one-sided alternative was used / would have been appropriate. Step 3.Assume the significance level used by the authors, or else use α = 0.05.
Step 4.For a medium effect size (|d| = 0.5), use Minitab to calculate power for the chosen test at the sample size used by the authors:
- Use standard deviation calculated at baseline, corresponding to a control group, or pooled, whichever one is given or appropriate.
- Use that standard deviation and effect size to calculate the difference needed to produce the effect size.
Step 5.Comment on your result. Suggestions for Topic One: Energy drink effects on BP and ECG study (Shah et al 2016a) – you can use results from Table 1 or Table 2. Pick one of the variables and consider the comparison between the placebo and the energy drink group. Alternatively, you can verify the authors’ claim regarding power. Energy drinks and ginseng clinical trial (Shah et al 2016b) – you could try to use SBP or DBP for the comparison between placebo and energy drink at one of the time points. C-energy study (Kozik et al 2016) – consider changes from baseline reported in Table 1 or Table 2.
The research study by Shah et al. (2016a) intended to evaluate the effects of single and multiple caffeinated energy shots on hemodynamic and electrocardiographic (ECG) parameters. The authors noted that energy drinks are increasingly becoming available in the market under various brand names. An important element of these drinks is their combination of caffeine and nutritional elements. Despite these changes in energy drinks market trends, the article worryingly mentioned that consuming energy drinks results in health complications with the number of visits to emergency departments doubling in the five-year period from 2007 to 2011. Additionally, the article reported that other research studies had noted the link between caffeine drink consumption and adverse health effects. These studies have mentioned cardiovascular side effects as being common. Still, it is noted that there is limited of knowledge about the precise mechanism and extend to which energy drinks contribute to adverse health effects. As such, there is a need to collect more information on the effects of energy drinks consumption on blood pressure (BP) and ECG parameters (Shah et al., 2016a).
- Materials and Methods
The study presented in Shah et al. (2016a) was conducted as a randomized, double-blinded, placebo-controlled, crossover study in young active duty personnel. The inclusion criteria entailed volunteers being between 18 and 40 years of age. Persons excluded from the study were those with premorbid conditions, elevated BP (exceeding 140/80 mmHg), high QTc interval (exceeding 440 ms) or on medication that interacted with energy drinks. Prior to beginning of the study, the participants were subjected to BP and ECG readings as the baseline. Following that, participants were subjected to the first phase of the study. This entailed randomizing the participants and enrolling in two groups identified as placebo and energy drink group then asked to take a drink twice daily. The first phase was followed by a washout phase before commencing the second phase. The second phase entailed following the same timetable as the first phase. To avoid having confounding effects on the study, the participants avoided drinking energy drinks for at least seven days prior to the first phase commencing. Following the energy drinks consumption for the first and second phases, 26 participants were evaluated for hemodynamic and electrocardiographic changes with the data collected as presented in Table 2 (Shah et al., 2016a, p. 467). This included the systolic BP, diastolic BP, HR, PR-interval, QRS duration, QT interval, and QTc interval for base line at time when caffeine was injected, and one hour, three hours, five hours and at the end of the day. All missing data was handled using the carry forward method for 3.4% of the BP data and 9.3% of the missing ECG parameters. The statistical analysis performed consisted of a paired student’s t test that compared the two arms at each time-point (time matched) along with descriptive reporting of data. All data are reported as mean ± SD. A p value ≤ 0.05 was considered statistically significant. Intention to treat analysis was performed and reported. The maximum post-dosing values were also compared (Shah et al., 2016a).
- Outcomes and contributions
Shah et al. (2016a) reported that there were statistically significant results noted for the different hemodynamic and electrocardiographic parameters for both the first and second phases of the study. In fact, the results of the single energy shot from the first phase of the study indicated that the systolic BP was significant in the third and fifth hour while diastolic BP was significant in the first and fifth hour. The HR, PR-interval, QRS duration, QT interval, and QTc interval were not significant. In the second phase where post chronic consumption effects were recorded, it was noted that none of the parameters reported significant differences. The implication is that energy drinks consumption (whether intermittently or regularly) has an immediate effect on systolic and diastolic BP without any notable long-term effects. The researchers included 26 participants in the study, with 13 participants recruited into the test group (consuming energy drink) and the remaining 13 participants recruited into the placebo group. The researchers had estimated that between 16 and 34 participants would be required, thus ensuring that the 26 recruited participants were within the estimate. This guaranteed at least 80% of the study to detect, at 5% level of significant (alpha of 0.05), a difference of between 4 to 6 mmHg in the blood pressure with an SD estimated at 8 mmHg. With a sample size of 26, difference of 4 mmHg and SD estimated at 8 mmHg, the power value was calculated as 0.688. This suggests that the study had a 68.8% probability to correctly reject the null hypothesis that energy drinks consumption had no effect on the hemodynamic and electrocardiographic parameters, should the population differences be at least 4 mmHg with standard deviation 8 mmHg. The ideal power of 0.95 would have been achieved by increasing the sample size from 26 to 54 even as all other factors remain the same. This implies that the study was NOT sufficiently powered as confirmed by the Minitab output presented in the Appendix (Shah et al., 2016a).
Shah et al. (2016a) presents the study results in two tables to show the outcome of the single energy shot in the first day, and the outcome of chronic consumption in the seventh day. The figures showed the quantitative variables of mean, SD and p-values when showing the difference between single energy shot and chronic consumption. These were the results of the paired student’s t test for the Hemodynamic and electrocardiographic parameters (Shah et al., 2016a).
Shah et al. (2016a) noted some limitations in the study. Firstly, the results were affected by the pre-study caffeine consumption. This could have affected BP response. Secondly, the assessment was conducted for peripheral BP parameters, and yet central BP parameters would have produced superior assessment and predictive results if used. Thirdly, the study was conducted over one week, but a longer study period would be required to present more complete results. Thirdly, the study included young and healthy participants thereby making the results less valuable older persons as well as those with hypertension and cardiac rhythmic abnormalities. Fourthly, some of the data and this affected the results. Fifthly, the caffeine concentration in the energy drinks was not verified thus making it difficult to ascertain whether the noted effects were from caffeine or another ingredient. Finally, the small sample size introduced the possibility of type I error occurring (Shah et al., 2016a).
In an experimental setting, Shah et al. (2016a) showed that consuming energy drinks does not have a significant effect on hemodynamic and electrocardiographic parameters, except for the significant effect for systolic and diastolic BP for single energy shot. While the authors failed to demonstrate that single and multiple energy shots affected BP and ECG parameters, their results are valuable to energy drink consumers, particularly in the face of increasing production and marketing (Shah et al., 2016a).
Shah et al. (2016b) similarly applied an experimental approach in noting how energy drinks consumption affected electrocardiographic and blood pressure. The results noted that energy drinks caused an increase in systolic blood pressure, going on to add that the increases was also noted in QTc interval (Shah et al., 2016b). Kozik et al. (2016) applied an experimental approach and noted that consuming energy drinks increased systolic blood pressure, altered electrolytes, and resulted in repolarization abnormalities. Overall, these studies all applied experimental approaches and reported that energy drinks increased blood pressure thus validating the results presented in Shah et al. (2016a).
Kozik, T. M., Shah, S., Bhattacharyya, M., Franklin, T. T., Connolly, T. F. … &Pelter, M. M., ‘Cardiovascular responses to energy drinks in a healthy population: The C-energy study ‘, American Journal of Emergency Medicine (2016) 34, 1205-1209.
Shah, S. A., Dergush, A. E., Potts, V., Lee, M., Millard-Hastings, B. M. … & Lacey, C. S., ‘Effects of Single and Multiple Energy Shots on Blood Pressure and Electrocardiographic Parameters’, The American Journal of Cardiology (2016a) 117, 465-468.
Shah, S. A., Occiano, A., Nguyen, T. A., Chan, A., Sky, J. C. … & Nguyen, N. N., ‘Electrocardiographic and blood pressure effects of energy drinks and Panax ginseng in healthy volunteers: A randomized clinical trial’, International Journal of Cardiology (2016b) 218, 318-323.
Table 1. Minitab power calculation for assessment of differences in hemodynamic and electrocardiographic parameters for single energy shot and chronic consumption at sample size of 26
Figure 1. Power curve for assessment of differences in hemodynamic and electrocardiographic parameters for single energy shot and chronic consumption at sample size of 26
The Minitab outputs in Table 1 and Figure 1 indicate that with 26 participants (13 in the test group and 13 in the placebo group), the study had a 68.8% probability to correctly reject the null hypothesis that energy drinks consumption has no effect on the hemodynamic and electrocardiographic parameters, should the population differences be at least 4 mmHg with standard deviation 8 mmHg.
Table 2. Minitab power calculations showing that the ideal power 0.95 can be achieved using a sample size of 54
Figure 2.Power curve for assessment of differences in hemodynamic and electrocardiographic parameters for single energy shot and chronic consumption at sample size of 26 and 54
The Minitab outputs in Table 2 and Figure 2 indicate that with 54 participants, the study has a 95% probability to correctly reject the null hypothesis that energy drinks consumption has no effect on the hemodynamic and electrocardiographic parameters, should the population differences be at least 4 mmHg with standard deviation 8 mmHg.