Bailey Abney and Brendan Kelley. Macomb Mathematics Science and Technology Center. AP Chemistry 12. Mrs. Hilliard, Mrs. Gravel, Mr. Acre, Mrs.

Transcription

1 Comparing the Effects of Artificial Sweeteners and Sugar on the Growth Rate of Lactobacillus Acidophilus Bailey Abney and Brendan Kelley Macomb Mathematics Science and Technology Center AP Chemistry 12 Mrs. Hilliard, Mrs. Gravel, Mr. Acre, Mrs. Tallman 17 December 2015

2 Comparing the Effects of Artificial Sweeteners and Sugar on the Growth Rate of Lactobacillus Acidophilus In America, one in three adults is considered obese. With a multitude of diseases and conditions directly related to obesity, many individuals find themselves desperately needing to lose weight. Artificial sweeteners have become a popular dieting aid, but their possible side effects are not fully understood. Artificial sweeteners have recently been linked to increased populations of Firmicutes, gut bacteria that increase fat absorption. An in vitro study was conducted to determine if artificial sweeteners directly affect the growth of Lactobacillus acidophilus. This bacterium is a type of Firmicutes commonly added to processed foods. The experiment tested two popular artificial sweeteners, sucralose (Splenda) and aspartame (Equal), against traditional sugar. These treatments were added to tomato juice agar and 18 dishes of each treatment were prepared. Positive and negative controls were also included in the experiment. A milk broth was inoculated with Lactobacillus and grown for 24 hours at 37 degrees Celsius. Each Petri dish was then inoculated with 0.5 ml of the broth and incubated for 24 hours. Afterwards, percent coverage was determined for each dish. An ANOVA was utilized to analyze the data and returned a P-value of Because this value of below the alpha level 0.05, two sample t-tests were also run. Sugar and Splenda as well as Equal and Splenda were found to have significantly different means. The average percent coverage for the artificial sweeteners was similar to that of sugar and greater than the average for the positive control. This suggests that artificial sweeteners increase the growth of Lactobacillus in a way similar to sugar, likely contributing to obesity and its associated health risks.

3 Table of Contents Introduction..1 Review of Literature 4 Problem Statement.11 Experimental Design..13 Data and Observations...18 Data Analysis and Interpretation...30 Conclusion.39 Acknowledgements 44 Appendix A: Aseptic Technique 45 Appendix B: Disposal Technique..46 Appendix C: Formulas and Sample Calculations..47 Works Cited...54

4 Abney-Kelley 1 Introduction In recent years, obesity has been a rising concern in the United States, but few realize just how chronic the problem has become. The obesity rate among adults older than 20 years of age has more than doubled since 1960, with one in three adults being classified as obese. Furthermore, one in six children from the ages of six to 19 are also considered obese. Those who are obese have an increased risk of developing a myriad of diseases including, but not limited to, type 2 diabetes, high blood pressure, heart disease, liver disease, stroke, and certain types of cancer. There are several factors that influence whether a person becomes obese, such as genes and eating habits ( Overweight and Obesity Statistics ); however, this research will focus on a seemingly unlikely culprit artificial sweeteners. Artificial sweeteners are synthetic sugar substitutes that provide sweetness without calories. Thousands of products contain artificial sweeteners, and the obesity crisis has only increased the demand for artificial sweeteners as people try to improve their health by losing weight. On the surface, artificial sweeteners seem like the perfect solution for reducing calorie intake without giving up sweet foods. Nevertheless, the reality is that the long term impacts artificial sweeteners have on the body are unknown (Yang). This raises the question: is there more to artificial sweeteners than the scientific community realizes? As the use of artificial sweeteners has increased, scientists have begun looking into the effects these sweeteners have on the body. In 2014, an Israeli team of scientists were able to link artificial sweeteners to the bacteria populations present in the gut microbiome of mice. Mice that were fed artificial sweeteners had greater populations of

5 Abney-Kelley 2 Firmicutes, bacteria known to increase fat absorption. These mice became obese while mice that were fed sugar remained healthy. This suggested that artificial sweeteners could be a contributing factor in obesity (Shell). These findings were expanded on by conducting an in vitro study designed to observe how gut bacteria interacts with artificial sweetener outside of the body. For this research, Lactobacillus acidophilus, a type of Firmicutes commonly added to processed foods, was subjected to Splenda, Equal, and sugar to determine how each of these sweeteners impact its growth. This was accomplished by preparing Petri dishes using treated tomato juice agar. The Lactobacillus was cultured in a milk broth for 24 hours before 0.5 ml was added to each dish. The dishes were then placed in the incubator for 24 hours, after which time percent coverage was determined. By conducting an in vitro study rather than an in vivo study, possible lurking variables such as the organism s diet, metabolism, and food-reward system were eliminated. Observing whether or not Firmicutes have increased growth when subjected to artificial sweeteners outside of the body provides a better understanding of the possible link between artificial sweeteners and obesity; that is, knowing if increased Firmicutes populations are directly caused by artificial sweeteners and consequently cause weight gain, as suggested by the Israeli study, or if lurking variables are at play. The link between artificial sweeteners and Firmicutes growth could change the way weight loss is approached. While artificial sweeteners appear to be a beneficial aid, it is possible that people are unknowingly hindering their ability to lose weight. Currently, dieticians do not consider artificial sweeteners to be risky. Sweeteners are often presented as an option for both diabetics and nondiabetics who seek to lose weight (Morfino). By understanding

6 Abney-Kelley 3 how artificial sweeteners impact the gut microbiome, truly optimal methods for weight loss can be determined. For instance, eliminating artificial sweeteners entirely could potentially decrease Firmicutes populations, resulting in decreased fat absorption. Looking beyond calories could be the key to lowering the obesity rate and creating conversant and healthier future generations.

7 Abney-Kelley 4 Review of Literature The obesity rate in the United States has become alarmingly high, resulting in a greater number of individual s experiencing obesity-related health problems. Obesity is the result of a continuous energy imbalance, meaning that more calories are consumed than are expended on a regular basis. Obesity puts individuals at risk for health issues such as, but not limited to, type 2 diabetes, heart disease, high blood pressure, nonalcoholic fatty liver disease, and stroke ( Overweight and Obesity Statistics ). In an attempt to lose weight and improve their health, more and more individuals are turning to artificial sweeteners. Artificial sweeteners have become popular because they decrease calorie intake without removing sweet foods from an individual s diet. Artificial sweeteners are synthetic sugar substitutes. Currently, six artificial sweeteners have been approved by the Food and Drug Administration as food additives ( High-Intensity Sweeteners ). This does not include sugar alcohols, which are sweeteners that have been derived from plants and chemically altered. Despite being FDA approved, few studies have followed the long term effects of artificial sweeteners, and research relating artificial sweeteners to obesity has only begun in recent years. Two of the most common artificial sweeteners are sucralose and aspartame ( Artificial Sweeteners and Other Sugar Substitutes ). Sucralose has a chemical formula of C12H19Cl3O8 and is most often marketed under the brand name Splenda ( CID=71485 ). Aspartame on the other hand has a chemical formula of C14H18N2O5 and is commonly known as Equal or NutraSweet ( CID= ). In comparison, sucrose, more commonly known as table sugar, has a chemical formula C12H22O11 ( CID=5988 ). While sugar contains 16 calories per teaspoon ( Basic Report: 19335, Sugars,

8 Abney-Kelley 5 Granulated ), artificial sweeteners are nonnutritive, meaning that they have no nutritional value, and have negligible calories. This is because artificial sweeteners range from 300 to 7,000 times sweeter than sugar, allowing minuscule amounts to be used to sweeten products and calories to be essentially nonexistent. Splenda and Equal were included in this research because they are most commonly found in diet soda; soft drinks are consumed in large quantities in America and are one of the most prevalent sources of added sugar, so people trying to lose weight or control sugar intake tend to switch to diet soda (Benton). Aside from diet soda, artificial sweeteners can be found in a multitude of other products ranging from frozen meals and baby food to yogurt and toothpaste. Because they are not always clearly labeled on products, consumers likely encounter artificial sweeteners without being aware of their presence (Yang). While artificial sweeteners are added to products with good intentions to decrease added sugar they may not be as harmless as they seem. Evidence is accumulating in the scientific community that suggests artificial sweeteners may not only inhibit weight loss but may actually be contributing to weight gain. For example, studies have observed that individuals who consume artificial sweeteners often overcompensate for their decreased calorie intake. They justify eating more because they saved calories by replacing sugar with artificial sweetener. Unfortunately, people often consume more calories than they eliminate through the use of artificial sweetener, which ultimately leads to weight gain (Yang). Also, artificial sweeteners do not affect the brain in the same way as sugar. The food reward system in the brain, which fuels the desire to eat, responds to food consumption similarly to pleasurable activities such as sex and drug use. It also exhibits

9 Abney-Kelley 6 behavioral traits associated with other addictions such as binging, withdrawal, and craving (Yang). When food is consumed, especially sugar, dopamine is released as a pleasure response. Along with dopamine, a hormone called leptin is released to control appetite. Leptin gradually reduces the activation of dopamine, lowering the initial reward value of sugar. This not true in the case of artificial sweeteners. While dopamine is still released when artificial sweetener is consumed, leptin is not. Lack of satisfaction causes the brain to continue craving sweetness, which can also lead to excessive calorie consumption (Kirkwood). This alone shows that while artificial sweeteners taste similar to sugar, they are processed differently by the body. In the past, several studies have been conducted in an attempt to relate artificial sweeteners to weight gain and/or changes in the gut microbiome. For instance, in 1979 the San Antionio Heart Study examined the relationship between artificially sweetened beverages and long-term weight loss. Approximately 5,100 participants were chosen at random from the San Antonio area. The initial body mass index, BMI, was recorded for each participant, their dieting status, as well as their reported artificially sweetened beverage consumption. This consumption was based on how often each participant claimed to drink sugar free soda and/or add artificial sweetener to their coffee or tea. The study also accounted for factors such as exercise frequency, race, gender, smoking status, and socioeconomic standing because there was a possibility these factors were lurking variables that may have affected long-term weight loss. The data collected showed that people who were dieting were more likely to consume artificial sweeteners. When a follow up was conducted with roughly 3,600 participants after a period of 7-8 years, it was observed that baseline dieters had a greater increase in BMI than non-dieters.

10 Abney-Kelley 7 Overall, it was found that the average change in BMI for those who drank artificially sweetened beverages was 47% higher than those who did not or rarely consumed artificial sweeteners. This study was unable to prove causation; however, since the BMI change for those who consumed artificially sweetened beverages was much higher than those who did not, it was greatly implied that artificial sweeteners may in some way attribute to long-term weight gain (Fowler et.al). In 2014, a team of Israeli scientists were able to observe the effect that artificial sweeteners have on the gut microbiome through an in vivo study, a study conducted on a living organism. The gut microbiome refers to the bacteria populations that exist on skin, in saliva, and in the intestinal track. While bacteria are often given a negative connotation, some are actually a crucial part of healthy bodily functions. For example, gut bacteria often aid the body in the digestion of complex carbohydrates (Jabr). The Israeli team was specifically able to link artificial sweeteners to increased populations of the bacteria phylum called Firmicutes. In the experiment, mice were given water that was either laced with natural sweeteners or artificial sweeteners. After 11 weeks, it was observed that the mice given artificial sweetener had developed glucose intolerance. Glucose intolerance means that the body experiences difficulty absorbing glucose from the blood. This condition can lead to diabetes, heart disease, and a variety of other health complications. The mice that received the sugar treatment, however, were found to be in normal health. In addition to developing glucose intolerance, it was found that the Firmicutes populations had flourished in the mice that received artificial sweetener. This did not occur in the mice that consumed sugar, which suggested that the artificial sweeteners had in some way directly increased the growth of Firmicutes (Shell).

11 Abney-Kelley 8 The results of the Israeli study are supported by previous research that linked increased Firmicutes populations to the absorption of fat. This relationship was first observed in zebrafish. Zebrafish are translucent when they are very young, so scientists were able to track lipid absorption through the use of florescent dye. It was determined that the fish which had Firmicutes present in their intestinal track absorbed more fat than zebrafish that were raised in a germ-free environment and were fed the same diet. This means that the bacteria present in the gut microbiome can alter the amount of calories that are absorbed from food, resulting in different calorie intakes from the same diet. It was also found that the types of food consumed affect the populations of gut bacteria. A study on mice found that those that consumed a diet higher in fat had larger populations of Firmicutes than the mice that ate a low fat diet. These studies can be related to humans because significant overlap has been found between the bacteria present in the gut microbiome of animals and that of humans (Stromberg). The relation between Firmicutes and fat absorption and obesity was also observed in a previous study by Jeffery Gordon, a physician and biologist at Washington State University. In his study, increased Firmicutes populations were observed in genetically obese mice. His team investigated the role of Bacteroidetes and Firmicutes bacteria in obesity, for these bacteria make up 90 percent of the gut microbiome. He and his team concluded that genetically obese mice had 50 percent fewer Bacteroidetes and 50 percent more Firmicutes than regular mice. Furthermore, when Firmicutes from the obese mice were transplanted into normal mice, the normal mice too became fat (Shell). This can be explained by the wider variety of carbohydrateactive enzymes supplied by Firmicutes, which allow the body to break down complex carbohydrates, causing the body to extract

12 Abney-Kelley 9 more energy, or calories, from food. There may however be additional ways that gut bacteria react to artificial sweeteners that contribute to fat absorption (Griffiths). The findings of these previous studies were expanded on by conducting an in vitro study, meaning outside of a living organism. For the purposes of this research Lactobacillus acidophilus was subjected one of three treatments: sugar, Splenda, or Equal. Again, these sweeteners were chosen because of their popularity. The treatment levels were based on a combination of serving size and the FDA allowed daily intake. Lactobacillus was selected because it is a type of Firmicutes that is commonly added to processed foods such as yogurt. By designing an in vitro study as opposed to an in vivo study, conclusions can be made as the direct impact of artificial sweetener on the growth of gut bacteria that increase fat absorption. Based on the results of the Israeli study, it a suggested that the artificial sweeteners should increase the growth of the bacteria more than sugar. By comparing the growth of Lactobacillus when subjected to artificial sweeteners to the growth when subjected to sugar, whether or not the bacteria respond to artificial sweeteners the same way outside of the body as inside the body can be observed. Also, an in vitro study also allowed variables such as metabolism and the foodreward system to be removed from the experiment. Definitively linking artificial sweeteners to increased Firmicutes populations would show that artificial sweeteners contribute to increased fat storage, weight gain, and obesity. This in combination with the chemical processes in the brain could potentially reveal that artificial sweeteners do not in fact possess the perceived benefit of aiding weight loss. As more pieces of the puzzle come together, the suggestions being made about artificial sweetener become increasingly alarming. While artificial sweeteners are

13 Abney-Kelley 10 approved by the FDA as a safe food additive, little is known about the long term effects of consumption. It has not been until recently that scientists have linked artificial sweetener to increased Firmicutes populations, through studies such as those conducted by Jeffery Gordon and the Israeli team. These newfound observations have uncovered the very possible relationship between artificial sweetener and obesity (Shell). It is becoming more and more evident that artificial sweeteners may be fueling the obesity epidemic, the very crisis they are being implemented to avoid. By better understanding how the body processes artificial sweetener, society will be able to better fight the rising obesity rate, better educate the public about artificial sweetener and obesity, and encourage healthier choices among individuals.

14 Abney-Kelley 11 Problem Statement Problem: To determine if the growth rate of Lactobacillus Acidophilus is greater when subjected to a treatment of artificial sweetener, either aspartame (C14H18N2O5) or sucralose (C12H19Cl3O8), in comparison to a conventional sugar treatment, in this case sucrose (C12H22O11), when grown at 37 degrees Celsius for 24 hours as it relates to obesity. Hypothesis: Both the artificial sweetener treatments, aspartame and sucralose, will result in a significantly greater average percent coverage of Lactobacillus Acidophilus, in comparison to the sucrose treatment, when grown at 37 degrees Celsius for 24 hours. Data Measured: In this experiment, the independent variables were the treatments added to the tomato juice agar. These treatments included aspartame, sucralose, and sucrose. For the aspartame treatment, grams of aspartame was added to the agar. Then for the sucralose treatment, grams of sucralose was added to the agar. Finally, for the sugar treatment, grams of sucrose was added to the agar. Eighteen Petri dishes of each treatment were then inoculated with 0.5 ml of a milk broth that had been inoculated with Lactobacillus Acidophilus and placed in an incubator set at 37 degrees Celsius. The dependent variable of the experiment was the percent coverage calculated after a 24 growth period. The data collected for each treatment was compared using an ANOVA. Based on the results on the ANVOA, a series of two-sample t-tests was used to determine

15 Abney-Kelley 12 if any of the treatments resulted in a significantly greater number of colonies than the other treatments.

16 Abney-Kelley 13 Experimental Design Materials: Lactobacillus Acidophilus live sample (66) 60 x 15 mm Polystyrene sterile Petri dishes Digital scale ( precision) 100 g Casein powder 100 g Peptone 11 g Nutrient agar powder 1.9 L Distilled water 100 g Powdered skim milk 5 g Yeast 500 ml Tomato juice (4) Weigh boat, small Weigh boat, large Scoopula (6) 1 cm Stirring magnet Hot plate Hot mitt g Splenda g Equal g Domino sugar Bunsen burner Striker Inoculation loop (5) 500 ml Erlenmeyer flask 1000 ml Erlenmeyer flask 250 ml Erlenmeyer flask Pipet with disposable tips Translucent grid Wet Erase marker Digital scale ( precision) Incubator, 37 C Procedures: Note: Please read Appendix A: Aseptic Technique before carrying out any of the following procedures. It is crucial that aseptic technique is properly utilized throughout the duration of this experiment to ensure optimal results and to avoid contamination. Begin each lab session with clean hands that have been washed using soap and warm water and sprayed with ethanol. Always spray the lab surface with ethanol before beginning. Treated Tomato Juice Agar Preparation: 1. Add 250 ml of distilled water to a 500 ml Erlenmeyer flask. 2. Measure 100 ml of tomato juice using a 100 ml graduated cylinder. Add to the Erlenmeyer flask of distilled water. 3. Use a scoopula and weigh boat to measure out 2.5 grams of casein powder. 4. Rinse the scoopula and use a second weigh boat to measure out 2.5 grams of peptone. 5. Rinse the scoopula and use a third weigh boat to measure out 2.75 grams of nutrient agar powder.

17 Abney-Kelley Add the casein, peptone, and nutrient agar to the Erlenmeyer flask of distilled water and tomato juice. Tip the flask slightly and gently slide a stirring magnet into the solution. 7. Place the flask onto a hot plate. Set the hot plate to its highest heat setting. Set the stirring speed to medium. 8. Allow the agar to come to a boil. Once the boiling point has been reached, use a hot mitt to remove the beaker from the hot plate. Ensure the agar does not bubble over. 9. Allow the agar to come to room temperature (when the flask is cool enough to touch) then add the intended treatment. For the sugar treatment add grams of sugar to the agar. For the aspartame treatment, add grams of Equal. For the sucralose treatment, add grams of Splenda to the agar. 10. Return the flask to the hot plate with low heat to prevent it from solidifying. Again set the stirring speed to medium. Allow the agar to mix for approximately one minute. Remove the flask from the hot plate with a hot mitt. 11. Clam open one sterile Petri dish to avoid complete air exposure (See Figure 1). Pour just enough agar to cover the bottom of the dish. Prepare 20 dishes using the treated agar. 12. Allow the agar to set. The agar is completely set when it is no longer a liquid and has formed a gel like substance. 13. Label the lid of the dish with the treatment, date, dish number, and group name. Set two dishes aside from each treatment for negative controls. These dishes will not be inoculated. Refrigerate until inoculation. 14. Repeat steps one through 13 for the two remaining treatments. 15. Repeat the steps one through 13 a fourth time, but skip steps nine and ten. This untreated agar will serve as controls. Prepare only six Petri dishes using this agar. Label these dishes untreated with the dish number, date, and group name. Four of these dishes will be inoculated and serve as positive controls, and two will not be inoculated and serve as negative controls. Milk Broth Preparation and Lactobacillus Acidophilus Dilution: 16. Using a graduated cylinder to measure, add 800 ml of distilled water to a 1000 ml Erlenmeyer flask. 17. Use a clean scoopula and a weigh boat to measure out 100 grams of powdered skim milk. Add to the 800 ml of distilled water. This will be solution A.

18 Abney-Kelley Tip the flask slightly and gently slide a 1 cm stirring magnet into the flask. Place the flask on a hot plate and turn on the heat to the highest setting. Set the stirring speed to medium. 19. Allow solution A to reach boiling point, then quickly remove the flask from the hot plate using a hot mitt. 20. For solution B, use a graduated cylinder to measure 100 ml of distilled water. Add this to a 500 ml Erlenmeyer flask. 21. Using a graduated cylinder to measure, add 100 ml of tomato juice to the 500 ml Erlenmeyer flask for solution B. 22. Rinse the scoopula and use another weigh boat to measure 5.0 grams of yeast and add this to the 500 ml Erlenmeyer flask for solution B. 23. Tip the flask slightly and gently slide a 1 cm stirring magnet into the flask. Place the flask on a hot plate set to the highest heat setting. The stirring speed should be set to medium. 24. Allow solution B to reach boiling and then quickly remove the flask from the hot plate with a hot mitt. 25. Loosely cover solution A and solution B with plastic wrap. Allow both solutions to cool completely. Then carefully add solution B to solution A. This combined solution will be referred to as the milk broth. 26. Add 150 ml of the milk broth to a sterile 250 ml Erlenmeyer flask. 27. Set up a Bunsen burner and flame an inoculation loop. Allow the loop to cool completely. 28. Open the live sample of Lactobacillus Acidophilus, flame the rim, and place the inoculation loop into the sample. The sample should form a film/bubble across the loop. 29. Place the transfer loop into the 150 ml of milk broth and move the inoculation loop back and forth to transfer the bacteria. 30. Repeat steps 27 through 29 two more times. 31. Loosely cover the Erlenmeyer flask with plastic wrap and place in an incubator for 24 hours at 37 degrees Celsius.

19 Abney-Kelley 16 Petri Dish Inoculation: 32. Remove the milk broth from the incubator and discard the plastic wrap. Flame the rim of the flask. 33. Place a disposable tip on the end of the pipet without touching the tip to avoid contamination. Set the pipet to 500 microliters. 34. Hold down the button on the top of the pipet and place the tip of the pipet into the milk broth and slowly release the button. 35. Clam open the appropriate Petri dish. While moving the tip of the pipet in a circular motion beginning at the center of the Petri dish and moving towards the outer edge, slowly press the button on the pipet to add the milk broth to the dish. Tilt the dish back and forth to cover the entire surface. 36. Label the dish with the date of inoculation and bacteria type. Repeat steps 35 through 37 for the remaining Petri dishes. Do not inoculate negative controls. 37. Once all of the appropriate Petri dishes have been inoculated, place all of the dishes in an incubator at 37 degrees Celsius. This includes the positive and negative controls in addition to the inoculated dishes. Determining Percent Coverage: 38. After 24 hours has elapsed, remove the Petri dishes from the incubator. 39. Print a grid onto translucent paper. Cut the grid to fit inside the lid of an unused Petri dish and secure with tape. 40. For each dish, replace the lid with the grid. Use a wet erase marker and color the areas where bacteria are present in the dish. See Figure Estimate the number of squares that are colored and divide by the total number of squares to determine percent coverage. Record this value for each of the Petri dishes, including controls. 42. After data collection is compete, refer to Appendix B for proper disposal procedures.

20 Abney-Kelley 17 Diagrams: Figure 1. Clamming a Petri Dish Figure 1 shows how to clam open a Petri dish. Only open the dish as much as is necessary to minimize air exposure. Figure 2. Percent Coverage In the above figure, the image on the left shows an uncovered Petri dish. The right image shows the Petri dish with the grid. The right image also shows where a wet erase marker (shown in red) was used to color the area where bacteria growth was present.

21 Abney-Kelley 18 Data and Observations Figure 3. Determining Percent Coverage In the above figure, the image on the left shows an uncovered Petri dish. A separate Petri dish lid with a grid taped inside was then placed on top of the dish as shown in the image on the right. Also pictured in the right image, a wet erase marker (shown in red) was used to color the area where bacteria growth was present. The number of colored squares was then estimated and used in the percent coverage equation shown in Figure 3. Mature or Distinct Colonies Distinct Premature Colonies Very Small Premature Colonies Figure 4. Premature Vs. Mature Colonies. Figure 4 highlights the difference between distinct colonies and premature colonies. The darker, beige splotches circled in the first picture are examples of the

22 Abney-Kelley 19 mature or distinct colonies. These colonies were easily counted. The second image shows premature colonies. These were indentations on the surface of the agar where mature colonies would have formed given more time in the incubator. If these premature colonies were large enough, they could still be colored and counted using the grid method. However, some premature colonies were too small to count accurately. An example of this is shown in the third image. Figure 5. Percent Coverage Equation Percent Coverage = Squares Bacteria Squares Total 100 The above equation details how percent coverage was calculated for each of the petri dishes after a 24-hour incubation period. The number of squares colored on the Petri dish lid grid template was divided by the total number of squares on the template, that being 90.5 squares. This quotient was then multiplied by 100 to find the percentage of the agar surface that was covered by bacteria. A sample calculation can be found in Appendix C. Table 1 Sugar Treatment Data Plate Number Order of Inoculation Sugar Treatment Date of Inoculation Date of Measurement Percent Coverage Observed /28/ /2/ /28/ /2/ Control 10/28/ /2/2015 N/A /28/ /2/ Control 10/28/ /2/2015 N/A /28/ /2/ /28/ /2/ /28/ /2/

23 Abney-Kelley 20 Plate Number Order of Inoculation Sugar Treatment Date of Inoculation Date of Measurement Percent Coverage Observed /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ Average: The above table includes the percent coverage recorded for each of the sugar treatment trials. It is important to note that only 18 of the 20 plates were inoculated with bacteria, for plates three and five served as controls. The controls were not inoculated. The data for these controls can be found in Table 4. All of the plates were inoculated on 10/28/15, and all of the plates were analyzed for percent coverage on 11/2/15. It can be seen that all but two of the inoculated plates showed more than 10% coverage. The minimum percent coverage was recorded for plate 10 at 7.73%, and the maximum percent coverage was recorded for plate 6 at 30.90%. This gives a range of 23.17%. The average for the sugar treatment trials was found to be 17.43%.

24 Abney-Kelley 21 Table 2 Splenda Treatment Data Plate Number Order of Inoculation Splenda Treatment Date of Inoculation Date of Measurement Percent Coverage Observed /28/ /2/ /28/ /2/ /28/ /2/ Control 10/28/ /2/2015 N/A /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ Control 10/28/ /2/2015 N/A /28/ /2/ /28/ /2/ Average: Table 2 includes the data collected for the Splenda treatment trials. Plates four and 18 were selected to be controls and were not inoculated. The remaining 18 dishes were all inoculated on 10/28/15, percent coverage was analyzed on 11/2/15. These trials produced a range of 23.21%; the minimum percent coverage was recorded for plate 12 and plate 14 with 3.31%, and the greatest percent coverage was recorded for plate five at 26.52%. Slightly more than half of the plates showed a percent coverage greater than 10%. The calculated average for the Splenda treatment was found to be 11.48% coverage.

25 Abney-Kelley 22 Table 3 Equal Treatment Data Plate Number Order of Inoculation Equal Treatment Date of Inoculation Date of Measurement Percent Coverage Observed /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ Control 10/28/ /2/2015 N/A /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ Control 10/28/ /2/2015 N/A /28/ /2/ /28/ /2/ Average: Shown above is the data collected for the Equal treatment trials. Of the 20 dishes, plates nine and 18 were used as controls. These plates were not inoculated, however, the remaining dishes were inoculated. Each of the dishes were inoculated on the same day, 10/28/15, and the percent coverage was analyzed on the same day, 11/2/15. The minimum percent coverage was recorded for plate 11 at 6.63%. This is the only plate that did not show at least 10% coverage. The maximum percent coverage was recorded for

26 Abney-Kelley 23 plate 10 at 26.52%. This gives the data for the Equal treatment a range of 19.89%. The average was found to be 17.31% coverage. Table 4 Controls Not Inoculated Data Controls - Not Inoculated Treatment Plate Number Date of Measurement Percent Coverage Sugar 1 11/2/ Sugar 2 11/2/ Equal 1 11/2/ Equal 2 11/2/ Splenda 1 11/2/ Splenda 2 11/2/ No Treatment 1 11/2/ No Treatment 2 11/2/ The above table shows the results of the control dishes that were not inoculated. These included two controls for each of the following: the sugar treatment, the Splenda treatment, the Equal treatment, and untreated agar. All of these plates were analyzed for percent coverage on the same day, 11/2/15. None of these dishes showed any bacteria growth whatsoever. This suggests that the experiment was designed and carried out in such a manner that prevented contamination of the plates.

27 Abney-Kelley 24 Table 5 Controls Inoculated Data Controls - Inoculated - No Treatment Plate Number Date of Inoculation Date of Measurement Percent Coverage 1 10/28/ /2/ /28/ /2/ /28/ /2/ /28/ /2/ Average: 8.84 Table 5 includes the results of the control dishes which contained untreated agar and were inoculated. Each of the dishes were inoculated on the same day, 10/28/15, and analyzed for growth on the same day, 11/2/15. Plate 1 showed the least percent coverage at 6.63%, and Plate 3 showed the greatest percent coverage at 12.15%. Only plate 3 showed more than 10% coverage. The average was found to be 8.84% coverage. These controls serve to show that the treatments used in the experiment sugar, Splenda, and Equal did indeed affect the growth of the bacteria. Observations: Table 6 Sugar Treatment Observations Sugar Treatment Plate Number 1 2 Observations Several colonies along with several small premature colonies present. Mature colonies are close together and are concentrated along the outer edge of the dish. Two somewhat small, distinct colonies. Several very small premature colonies. 3 Control; see Table 9. Some mature colonies concentrated along the outer edge of the dish. 4 Two large premature colonies and several very small premature colonies.

28 Abney-Kelley 25 Plate Number Sugar Treatment Observations 5 Control; see Table 9. Nearly half the plate appears to be covered in bacteria. All of the 6 colonies are concentrated together starting at the edge of the plate and moving inward. Very few immature colonies. 7 Three small, distinct colonies. Many small premature colonies. One large colony toward the center of the plate with four small, distinct 8 colonies toward the edge of the plate. Many very small premature colonies Several somewhat large colonies, and very few premature colonies. The premature colonies that are present are extremely small. No mature colonies, however somewhat large, distinct premature colonies are present. Some bacteria growth at the outer edge of the plate. Many very small premature colonies. Some bacteria growth at the outer edge of the plate. A few very small premature colonies, and one large distinct premature colony. 13 Several somewhat large colonies. Many very small premature colonies A roughly even spread of mature and premature colonies across the plate, none of which are notably large. Concentrated bacteria growth along the outer edge of the plate with some very small, distinct premature colonies in the center. Somewhat mature bacteria growth toward the outer edge of the plate. Very few premature colonies. Distinct bacteria growth covering roughly one third of the plate, starting at the edge of the plate and moving inward. Few premature colonies. Some small mature bacteria colonies along with several small premature colonies across the plate. Small mature colonies along the outer edge of the plate. One larger premature colony also along the outer edge. Three somewhat large bacteria colonies along with several small premature colonies. The above table shows all of the observations for the sugar treatment trials. These observations primarily include the size and maturity of the colonies present after the 24-

29 Abney-Kelley 26 hour incubation period. It is notable that most of the mature colonies grew along the edges of the Petri dishes. Table 7 Splenda Treatment Observations Plate Number Splenda Treatment Observations Some of the milk broth splashed on the lid during inoculation. Three distinct colonies with some very small premature colonies. Bacteria covering most of the agar surface. Some distinct colonies but mostly small premature colonies. Approximately seven small distinct colonies with small premature colonies covering the majority of the agar surface. 4 Control; see Table 9. Bacteria growth covering about one half of the agar surface with few 5 premature colonies Three large, distinct colonies with many small premature colonies covering the majority of the agar surface. Approximately seven small distinct colonies with very small premature colonies across the plate but not entirely covering the surface. A few somewhat distinct colonies but many premature colonies covering the surface of the agar. Almost no distinct growth. Small premature colonies covering the agar surface. Many very small premature colonies across the surface of the agar but not entirely covering it. 11 Three large premature colonies and several very small premature colonies. 12 Some very small premature colonies. Very little plate coverage. 13 Some small premature colonies. Little plate coverage. 14 Some very small colonies along with many premature colonies. 15 Two distinct small colonies with many very small premature colonies Two distinct small colonies at the outer edge of the plate. Several small and very small premature colonies. Great amount of plate coverage. Distinct bacteria growth along the outer edge of the plate. Mainly premature colonies covering the agar surface.

30 Abney-Kelley 27 Plate Number Splenda Treatment Observations 18 Control; see Table 9. Some milk broth splashed on the lid of the dish during inoculation. 19 One distinct colony. Some very small premature colonies across the surface. 20 Small premature colonies covering the agar surface. Table 7 includes the observations for the Splenda treatment trials. Observations include size, type, and location of the bacteria growth. Premature colonies were common for this treatment. Plates one and 19 also noted that some of the milk broth splashed onto the lid when the dishes were inoculated prior to incubation. Table 8 Equal Treatment Observations Plate Number Equal Treatment Observations Some distinct bacteria growth along the outer edge of the dish. Few premature colonies. Some distinct growth along the outer edge of the plate. Some small premature colonies. Small colonies concentrated along the outer edge of the plate. Some small premature colonies. Great plate coverage. Great amount of plate coverage. Mainly small, somewhat distinct colonies. Some small distinct colonies across plate accompanied by many very small premature colonies. Distinct bacteria growth covering approximately one fifth of the plate. Growth has occurred across the plate, not just along the outer edge. Several very small premature colonies across the plate. Several small colonies along with some small premature colonies across the plate. 8 A few distinct colonies with very few premature colonies 9 Control; see Table 9. Great amount of plate coverage. Small distinct colonies covering 10 roughly one third of the plate. Many very small premature colonies.

31 Abney-Kelley 28 Plate Number Equal Treatment Observations Many very small premature colonies covering the surface of the agar. One large distinct colony and two smaller distinct colonies. Some very small premature colonies. Some distinct growth but mainly small premature colonies covering the surface of the plate. Small colonies along the edge of the plate. Few very small premature colonies. Large colony at the edge of the plate with several small premature colonies across the agar surface. Several small colonies across the surface of the agar with several small premature colonies across the surface of the agar. Three large colonies close to the outer edge of the plate. Many small premature colonies covering the agar surface. 18 Control; see Table 9. Great amount of plate coverage. Small colonies across the surface 19 of the agar and many small premature colonies across the surface of the agar. 20 Some distinct growth at the outer edge of the plate and some large premature colonies. The above table depicts the observations that were recorded for the Equal treatment trials. These observations include descriptions of the size, maturity, and location of the colonies.

32 Abney-Kelley 29 Table 9 Control Not Inoculated Observations Controls - Not Inoculated Treatment Plate Number Observations Sugar 1 No bacteria growth present. Sugar 2 No bacteria growth present. Equal 1 No bacteria growth present. Equal 2 No bacteria growth present. Splenda 1 No bacteria growth present. Splenda 2 No bacteria growth present. No Treatment 1 No bacteria growth present. No Treatment 2 No bacteria growth present. Table 9 includes the observations for the control dishes that were not inoculated. It is shown that none of the dishes showed any signs of bacteria growth. Table 10 Controls Inoculated No Treatment Observations Plate Number Controls -Inoculated - No Treatment Observations 1 Small distinct colonies. Few premature colonies. 2 Small distinct colonies primarily along the outer edge. Some large distinct colonies along with small distinct colonies along 3 the outer edge of the plate. Few premature colonies. 4 Small distinct colonies across the plate with some premature colonies. The above table includes observations made concerning the size, location, and maturity of the bacteria growth for the control dishes that contained an untreated agar and were inoculated.

33 Abney-Kelley 30 Data Analysis and Interpretation In this experiment, Lactobacillus acidophilus was grown on Petri dishes using a tomato juice agar and sweetener treatments. After 24 hours of growth in the incubator, a transparent sheet of grid paper was used to measure the percentage of the plate covered with the bacteria. This percentage was used to see if there was a significant difference using an Analysis of Variance and multiple two-sample t tests. The data found was considered reliable because it was a simple random sample. A simple random sample ensures that every trial has the same chance of getting the correct results. Another point of notice is the use of controls. In this experiment, the controls were plates that were not inoculated. No bacteria grew on these plates, which shows that there was no contamination on the treated plates. These preventative measures ensure that the data is valid and replicable. Before running any of these statistical tests, however, certain assumptions had to be met. These assumptions included that the data be normal. As shown below in Figure 6, there is little to no skewness for each treatment type. There are some outliers however, so care must be taken when making conclusions. Another assumption was that the trials were randomized. Each plate was selected in the order the random integer setting of the TI-Nspire calculator gave as an output. This makes the experiment a simple random sample, or SRS. These trials were also run independently. This means that one trial did not depend on another trial s results. For example, the previous Petri dish had no effect on the successive Petri dish. Also, the standard deviations of each of the populations is unknown; the sample standard deviations are very close to each other, with a range of

34 Abney-Kelley 31 around These assumptions allowed an ANOVA and 3 two-sample t-tests to be run and ensure that the data is unbiased. Before running any tests, box plots were made to compare the shape and spread of each treatment type and to check for normalcy. Figure 6 below shows the boxplots. Figure 6. Percent of Plate Covered Boxplots Figure 6 above shows the boxplots for each treatment. The boxplot on top is trials treated with sugar, the middle boxplot is the trials treated with Equal, and the bottom boxplot is for trials treated with Splenda. As can be seen from Figure 6, the sugar and Equal boxplots are very similar. When comparing Splenda to both of these, however, Splenda has much smaller percentages. These boxplots are almost normally distributed with no skewness, but both sugar and Splenda have some outliers. These outliers may affect the results of the t test, so the results may not be completely reliable. These boxplots imply a significant difference between Splenda and the others, but a statistical inference test is necessary to be sure.

35 Abney-Kelley 32 First, an ANOVA (Analysis of Variance) should be run to see if the populations are significantly different from one another. ANOVA tests check if the population means are significantly different based on the amount of variance in each sample. The hypotheses used for an ANOVA are: H 0 : μ 1 = μ 2 = μ 3 H A : Not all μ 1, μ 2, μ 3 are equal In these hypotheses, µ is the population mean and the subscripts 1, 2, and 3 signify each treatment type s population. Since all of the assumptions were met, the ANOVA could be run. The mathematics The results of the test are shown below in Figure 7. Figure 7. ANOVA Results Figure 7 shows the results of the ANOVA run of the three treatments tested in this experiment. The F statistic was 6.048, which is very high, assuming the null hypothesis is true. The corresponding P-value is very low at This means that the null hypothesis is rejected because the P-value is significantly lower than the chosen alpha level, α = This statistical test shows that there is significant difference between the

36 Abney-Kelley 33 three populations, and by looking at the boxplots in Figure 6, it seems that Splenda is causing this result. Because the ANOVA found that there is significant difference between the populations, two-sample t tests were run on each pair of populations to further investigate this. In a two sample t-test, two population means are compared to see if they are equal or not equal. The two hypotheses used in this experiment are as follows: H 0 : μ 1 = μ 2 H A : μ 1 μ 2 In this case, µ is the mean of each population of sweetener type, and the ones and twos signify which sweetener was used. These hypotheses were used for every t test run. As stated above, all of the assumptions were met, so each test was run without flaw. The mathematics of the t test is found in Appendix C. The first pair tested was sugar and Equal. The results are shown below. Figure 8. Sugar and Equal t test Results

37 Abney-Kelley 34 Figure 9. Sugar and Equal P-Graph Figure 10. Sugar and Equal Confidence Interval Figure 8 above shows the results of the t test performed on the trials for sugar and for Equal. The t-value was 0.065, which means that the difference of the means is standard deviations away from 0, assuming the null hypothesis is true. The P-value was 0.948, which is extremely high. This means the null hypothesis failed to be rejected because the P-value is significantly higher than the chosen alpha level, α = This t test and the boxplots are strong evidence that sugar and Equal have nearly the same amount of growth of the bacteria used in this experiment. There is a 94.8% chance that

38 Abney-Kelley 35 these results were obtained by chance alone. Figure 9 above shows the P-graph of the two-sample t test. This graph has a large shaded area, which means that there is a large chance that these results were obtained by chance. Figure 10 shows the confidence interval of the two-sample t test. This shows that there is 95% confidence that the true mean difference of the sugar and Equal populations is between -3.7% and 3.9%. These numbers contain 0, which shows that there is not much difference between the populations. The second pair tested was sugar and Splenda. The results are below in Figure 11. Figure 11. Sugar and Splenda t test Results Figure 12. Sugar and Splenda P-Graph

39 Abney-Kelley 36 Figure 13. Sugar and Splenda Confidence Interval Figure 11 above shows the results of the t test performed on the trials for sugar and Splenda. The t-value was 2.899, which means that the difference of the means is standard deviations away from 0, assuming the null hypothesis is true. The P-value was , which is fairly low. This means the null hypothesis is rejected because the P- value is lower than the chosen alpha level, α = This t test is strong evidence that sugar and Splenda have significantly different amounts of growth of Lactobacillus. There is a 0.65% chance that these results were obtained by chance alone. Figure 12 above shows the P-graph of the two-sample t test. This graph has an extremely small shaded area, which means that there is a very small chance that these results were obtained by chance. Figure 13 shows the confidence interval of the two-sample t test. This shows that there is 95% confidence that the true mean difference of the sugar and Splenda populations is between 1.8% and 10.1%. These numbers do not contain 0, which shows that there is a larger difference between the populations. The last pair tested was Equal and Splenda. The results are below in Figure 14.

40 Abney-Kelley 37 Figure 14. Equal and Splenda t test Results Figure 15. Equal and Splenda P-Graph Figure 16. Equal and Splenda Confidence Interval