I am thinking of doing a small study regarding blood pressure and extra virgin olive oil. I did notice a study " olive oil and reduced need for antihypertensive medication" in google scholar. I am afraid my sample size will be around 10-15. My hypothesis is extra virgin oil supplementaion will lower the blood pressure. I plan to ask male adult participants to take 4 tablespoons of extra virgin olive oil, 3 spoonfuls in females, 2 tablespoons in teens. Unfortunately, I will not be able to control other aspect of their diets. I plan to check their BMI, BP, HR after 2 months and see whether there is any significant difference. I plan to create control group and ask them to follow low fat diet. I plan to randomly assign subjects to control or intervention group. Study will not be blinded.
I have few problems and will need advice.
1. I expect effect size small and with limited sample size, I may not be able to find significant difference.
2. Any suggestions to improve robustness of study?
3. Should I use crossover design?
4. How to do randomization?
5. I am not knowledgeable about statistics. How should I do analysis of data and what tests should I use?
Thanks
Blood pressure and extra virgin olive oil
Moderators: AmyCowen, kgudger, bfinio, MadelineB, Moderators
-
SciB
- Expert
- Posts: 2068
- Joined: Fri Feb 01, 2013 7:00 am
- Occupation: Retired molecular biologist, university researcher and teacher
- Project Question: I wish to join Scibuddies to be able to help students achieve the best science project possible and to understand the science behind it.
- Project Due Date: n/a
- Project Status: Not applicable
Re: Blood pressure and extra virgin olive oil
Hi and welcome to Scibuddies.
You have proposed an interesting and worthwhile study that should allow you to test your hypothesis that a diet including daily extra virgin olive oil (I would also choose organic olive oil if you can get it.).
I can see that you are aware of the problems of doing such a study on human volunteers--small sample size, inability to control for other factors that could influence BP and relatively short duration of the study. Yes, a large sample of 1000s of individuals monitored over a decade or two would give you statistically improved data, but that's not possible.
Your independent variable is the daily dose of olive oil and the dependent variable is BP, comparing a baseline reading at the beginning of the study with a final reading after 2 months. You also indicated that the study would compare adults and teenagers and male and female volunteers. Age and gender are two factors that can affect BP.
I'm not so sure about the olive oil doses, however. For adults, I would not give females a smaller amount than males, but I question whether it is realistic to expect a person to ingest 3-4 measured tablespoons of olive oil per day. I have olive oil dressing on my salad that probably amounts to a Tb of oil and I have a salad 3 or 4 times a week, but not every day. I sometimes pour a Tb on pasta with garlic and Parmesan, and I often use it for sauteing, but drinking a quarter cup of olive oil straight would not be palatable to me. How did you expect the volunteers to ingest that much olive oil every day?
BP varies widely depending on a person's age, activity level, general health, medications, how much caffeine they consume, how stressful there lives are, and more. I suggest that you do some research online about how to take accurate BP readings and then draw up a standardized protocol that you use for each measurement, such as having the person sit quietly for 15 minutes before making the reading, and doing it twice and averaging the two readings. Some BP monitors that you can buy will automatically do two measurements and average them for you. Also record the heart rate.
If you can get volunteers whose age differs by no more than ten years, your results may be more accurate.
Since you are planning to use human volunteers, you need to be sure that the rules of your school science fair allow you to do that. You will need to write up a proposal for your experiment detailing how you will select the volunteers, administer the olive oil, take the readings and do statistical comparisons. In clinical trials, volunteers have to read and sign a consent form before they can be enrolled in the study and the study design has to be approved by a review board. Find out from your teacher what the rules are for using human volunteers in a science project.
I'm sure you will have more questions, so don't hesitate to post them on Scibuddies. We are here to guide you in doing an interesting and useful project.
Sybee
You have proposed an interesting and worthwhile study that should allow you to test your hypothesis that a diet including daily extra virgin olive oil (I would also choose organic olive oil if you can get it.).
I can see that you are aware of the problems of doing such a study on human volunteers--small sample size, inability to control for other factors that could influence BP and relatively short duration of the study. Yes, a large sample of 1000s of individuals monitored over a decade or two would give you statistically improved data, but that's not possible.
Your independent variable is the daily dose of olive oil and the dependent variable is BP, comparing a baseline reading at the beginning of the study with a final reading after 2 months. You also indicated that the study would compare adults and teenagers and male and female volunteers. Age and gender are two factors that can affect BP.
I'm not so sure about the olive oil doses, however. For adults, I would not give females a smaller amount than males, but I question whether it is realistic to expect a person to ingest 3-4 measured tablespoons of olive oil per day. I have olive oil dressing on my salad that probably amounts to a Tb of oil and I have a salad 3 or 4 times a week, but not every day. I sometimes pour a Tb on pasta with garlic and Parmesan, and I often use it for sauteing, but drinking a quarter cup of olive oil straight would not be palatable to me. How did you expect the volunteers to ingest that much olive oil every day?
BP varies widely depending on a person's age, activity level, general health, medications, how much caffeine they consume, how stressful there lives are, and more. I suggest that you do some research online about how to take accurate BP readings and then draw up a standardized protocol that you use for each measurement, such as having the person sit quietly for 15 minutes before making the reading, and doing it twice and averaging the two readings. Some BP monitors that you can buy will automatically do two measurements and average them for you. Also record the heart rate.
If you can get volunteers whose age differs by no more than ten years, your results may be more accurate.
Since you are planning to use human volunteers, you need to be sure that the rules of your school science fair allow you to do that. You will need to write up a proposal for your experiment detailing how you will select the volunteers, administer the olive oil, take the readings and do statistical comparisons. In clinical trials, volunteers have to read and sign a consent form before they can be enrolled in the study and the study design has to be approved by a review board. Find out from your teacher what the rules are for using human volunteers in a science project.
I'm sure you will have more questions, so don't hesitate to post them on Scibuddies. We are here to guide you in doing an interesting and useful project.
Sybee
Re: Blood pressure and extra virgin olive oil
Thanks for your reply.
So I will be comparing the pre and post intervention(olive oil) blood pressures of the subjects. I am still not sure how to use statistics in this case. Which test should I use?
So I will be comparing the pre and post intervention(olive oil) blood pressures of the subjects. I am still not sure how to use statistics in this case. Which test should I use?
-
SciB
- Expert
- Posts: 2068
- Joined: Fri Feb 01, 2013 7:00 am
- Occupation: Retired molecular biologist, university researcher and teacher
- Project Question: I wish to join Scibuddies to be able to help students achieve the best science project possible and to understand the science behind it.
- Project Due Date: n/a
- Project Status: Not applicable
Re: Blood pressure and extra virgin olive oil
Hi,
The basic test would involve taking the average of all the baseline readings (before olive oil) for each group and comparing it to the average for the later readings. In order to establish that the difference is statistically significant, you need to perform a Student's t test. There are lots of explanations and tutorials online on how to do this and the calculations can be done automatically by just inserting values for the mean and standard deviation. Here are some websites to get you started:
https://en.wikipedia.org/wiki/Student%27s_t-test
https://www.youtube.com/watch?v=pTmLQvMM-1M
I'm sure you will have more questions, so I will be here to help.
Sybee
The basic test would involve taking the average of all the baseline readings (before olive oil) for each group and comparing it to the average for the later readings. In order to establish that the difference is statistically significant, you need to perform a Student's t test. There are lots of explanations and tutorials online on how to do this and the calculations can be done automatically by just inserting values for the mean and standard deviation. Here are some websites to get you started:
https://en.wikipedia.org/wiki/Student%27s_t-test
https://www.youtube.com/watch?v=pTmLQvMM-1M
I'm sure you will have more questions, so I will be here to help.
Sybee
Re: Blood pressure and extra virgin olive oil
Thanks Sybee.
I think I can get 25 volunteers for the study. 20 of them will be in age group 40-45yrs(10 out of this group will be male). Remaining 5 of them will be 13 yrs old.
So there are 3 distinct groups.
1. 10 male volunteers with age 40-45 yrs
2. 10 female with age 40-45 yrs
3. 5 female 13 yrs old
How should I allot control group? Should I take out 2-3 out of each distinct group as control?
I plan to recommend low fat diet to control group.
Do you want me to use paired student t test for each distinct group?
I think I can get 25 volunteers for the study. 20 of them will be in age group 40-45yrs(10 out of this group will be male). Remaining 5 of them will be 13 yrs old.
So there are 3 distinct groups.
1. 10 male volunteers with age 40-45 yrs
2. 10 female with age 40-45 yrs
3. 5 female 13 yrs old
How should I allot control group? Should I take out 2-3 out of each distinct group as control?
I plan to recommend low fat diet to control group.
Do you want me to use paired student t test for each distinct group?
-
SciB
- Expert
- Posts: 2068
- Joined: Fri Feb 01, 2013 7:00 am
- Occupation: Retired molecular biologist, university researcher and teacher
- Project Question: I wish to join Scibuddies to be able to help students achieve the best science project possible and to understand the science behind it.
- Project Due Date: n/a
- Project Status: Not applicable
Re: Blood pressure and extra virgin olive oil
You do want to use the paired t test for comparison as it is more powerful.
As for controls--the individuals who do not take olive oil--we usually do equal numbers, so 5 of each for the male and female adults. Can you recruit 6 teens? Try to do that so you can compare the means of 3 control individuals to 3 test subjects.
If you tell the control group to eat a low-fat diet and they do that, then you have introduced another independent variable that you would have to control for, meaning more subjects would have to be recruited. I would suggest that you tell all the subjects to eat a low-fat diet so there is no difference between the test groups except the olive oil.
I was going to suggest that you also use a questionnaire to get life-style information from your subjects, but this would be very complicated to analyze and personal health information is subject to a great many privacy regulations, so now I would say don't do that. The age and gender are the two demographics that you are most concerned with now.
Good luck!
Sybee
As for controls--the individuals who do not take olive oil--we usually do equal numbers, so 5 of each for the male and female adults. Can you recruit 6 teens? Try to do that so you can compare the means of 3 control individuals to 3 test subjects.
If you tell the control group to eat a low-fat diet and they do that, then you have introduced another independent variable that you would have to control for, meaning more subjects would have to be recruited. I would suggest that you tell all the subjects to eat a low-fat diet so there is no difference between the test groups except the olive oil.
I was going to suggest that you also use a questionnaire to get life-style information from your subjects, but this would be very complicated to analyze and personal health information is subject to a great many privacy regulations, so now I would say don't do that. The age and gender are the two demographics that you are most concerned with now.
Good luck!
Sybee