Soda Tax? Does it Make Sense?

Wow, we are in an ever changing world where customs are every day being challenged such as smoking outside, burning a fire in a fire pit or even believe it or not, if one can burn a wood fire in a home fireplace! I thought I had seen it all but I caught this CNN news story out of NY (remember the trans-fat ban?) about the governor trying to implement a soda tax here!

I will talk about the soda tax’s rational and if it makes any sense or not. I found the underlying study cited and will provide a link for those of you who do not want to take the press release or reporters word on what it really means. Also a note of caution, I found at least one widely reported dissenting opinion from a group backed by the food and beverage industry. It is hard enough to figure out if a study has any merit and I don’t want to get into other industry rebuttals (although they too make some good points). Here is the governor comparing sodas impact on children to smoking.

Interesting argument huh? I guess he thinks that only soda increases obesity in children? At least smoking decrease did seemingly decrease lung cancer in those that were genetically predisposed. My quick take would be that if soda and other junk foods were removed from public schools and children were given the option of eating healthy foods in school cafeterias, children would reduce their overall calorie intake during the school year.

Does taxing just soda’s make any sense? Well if the goal is to increase the states tax revenue and not the health of the children then I suppose so. Forgive me of being cynical here but I just don’t see how taxing sodas and then having a public education campaign (assuming the tax dollars really went to that) would work better than removing the soda, candy and junk food from schools. Would then an adult have to prove no kids were consuming the soda to avoid the tax (the logical assumption but I’m sure the tax would apply to all in the name of the kids!)?

Take a look at the actual study published in the American Journal of Clinical Nutrition in August 2006 (84:274-288) here. First off for those of you who depend on reporters to dissect and opine on these studies, you can read parts of these studies to make your own conclusions as to the validity and if the researchers are making any sense. Most studies are narrowly focused on a question and usually depend on their own research or others’ research. One thing that happens often is a study is misreported by the media. One of my pet peeves is referring to the actual study to draw my own conclusions.

The parts of a study that I look at are the abstract, introduction, methods, discussion, and conclusion. This study was actually a meta study (meaning a study of previous studies) that included the results of 30 studies. The study starts out by describing the issue:

Consumption of sugar-sweetened beverages (SSBs), particularly carbonated soft drinks, may be a key contributor to the epidemic of overweight and obesity, by virtue of these beverages’ high addes sugar content, low satiety, and incomplete compensation for total energy.

So you can see the researchers are attempting to find out if sodas contribute to obesity. This study focused on children and they initially believe the association of increase SSBs to weight gain is unclear. They make the case of increased obesity in recent years (easy to do) and start to speak of how they designed their studies of the 30 studies. They initially looked at 264 studies and paired that down to studies that listed varabiles they were interested in such as the age and recorded soda intake.

As I understand this type of meta analysis, this is a correlational study, meaning although they had 5 experimental studies in the group of 30 studies, this research did not use an experimental and control group. Correlational studies are far harder to ever prove and this study does not prove anything but rather offer one indication of a possible link between soda consumption in children to childhood obesity.

In the discussion chapter of this study the researchers point out the problems as they see it with this meta analysis.

Overall, results from our review support a link between the consumption of sugar-sweetened beverages and the risks of over-weight and obesity. However, interpretation of the published studies is complicated by several method-related issues, including small sample sizes, short duration of follow-up, lack of repeated measures in dietary exposures and outcomes, and confounding by other diet and lifestyle factors.

Well if you have glanced at this study you can see of evaluating a meta analysis of other studies can take some time. I am quite unwilling to pull each of the underlying studies so I will take these researchers at their word (they are seemingly being very through). What I will briefly explore before jumping to the conclusion is the small sample sizes and this notion of confounding by diet and other lifestyle factors.

For those of you who have taken a college level statistics course you remember that the sample first has to be representative of the general population (the whole population that you are studying), the sample has to be randomly sampled for there to be any validity and the sample size is determined by the population size you are sampling.

For instance I would not sample women’s candy preferences and state it is applicable to men. This study seems to focus on children so that seems ok. Another example would be the underlying studies using a random sampling of the overall population. My guess is this meta-study does not meet this criterion because it is composed of smaller studies that have different problem statements. The other issue with sampling is the size of the sample that is randomly picked will not work mathmatically unless it conforms to a formula. One can use larger sampling sizes to slightly increase the accuracy of the analysis but smaller sample sizes are not accurate. As the researchers are citing small sample sizes I would just say the results are then suspect.

Now for the ”confounding by other diet and lifestyle factors” issue. This is really the big issue here. How can one attribute that sodas account for X number of weight gain without looking at the whole diet? I mean if a kid eats mainly junk food such as candy bars and sodas two meals a day and the parental type tries to funnel in one good meal a day, what is the culprit here? The soda, candy bar, lifestyle? Who knows? In an experimental study there would have been two sample populations chosen randomly where one would eat a good nutritionally balanced diet and the other one would eat a good nutritionally balanced diet plus sodas! That is the only way I can see a study can claim one food is the root of all evil (ok – I’m getting carried away but you can see the logical problem).

The study’s conclusion (paraphrasing as it is three paragrahs long) say there is strong evidence for a SSB link that increases childhood weight and obesity. However the researchers also say (as they plainly must – my emphsis) that more study is needed. They also say that public health strategies should support water, low-fat milk and real fruit juice intake and discourage sodas for children and adolecents.

Thanks for hanging with me while listening to my take on this importaint study. My opinion is that reducing sodas as well as candy and other ”junk foods” in children (and heck even adults) would probably be good for the individuals concerned but making a logical leap to taxing sodas as part of a solution is clearly not supported. Why this study does not prove anything! I could have told the researchers that soda is not really a food group and children could easily lesten their overall calorie intake if they drank less sodas.

If the governer really is concerned with public health of children as he states, I think he should remove all junk food from the New York public school system. What do you think?