Not Knowing Who To Trust Part 1 received a lot of positive feedback. It went over the ebbs and flows of trust and distrust that we have in ourselves and our doctors to pinpoint what on earth is happening to our bodies when we’re sick. Lyme: Not Knowing Who To Trust Part 2 is a follow up to that, and it’s aim is to help you become a better researcher and ultimately a super awesome advocate for your own health.
Trusting Your Research
Even after you have learned to ~trust yourself and ~divide autonomy over your health with your well-meaning cute lil’ 80 year old doctor, there is another problem: no one really knows how to treat Lyme Disease!
The sad reality is that even successful Lyme literate doctors have not had success with many-to-most of their patients. Each individual is different, and even if the protocol to a cure was written in stone, it wouldn’t help every complex case of Lyme Disease.
Lyme Disease is unexplored territory. Some of us have general practitioners who don’t know much about Lyme, while some of us see Lyme specialists. Either way I think it’s important to do independent research and learn about your disease. Like I said earlier, relying on “that one thing” or any one doctor can be practically, emotionally and financially hazardous. You may be seeing a Lyme expert, but no one will do customary research for your case like you can.
That is, if you do your research effectively.
At one point, the more research I did, the more confused I’d become. All my questions had been answered – and that was the problem. There were too many people, scientists and researchers answering my questions without questioning their answers.
My goal in the next few paragraphs is to make you more watchful when you read the latest news about Lyme on the internet. It may be a bit “high-school science class” but I’m going to review the ups and downs of research methods with you so that you have a better idea of why putting your eggs in any one guru- or treatment-basket is risky.
High-School Science Class
The reason there is so much information out there for us to get lost in is that people – AKA researchers – like to ask questions based on their observations. Researchers turn their questions into testable hypotheses and conduct research to test out their hypotheses. They build on their knowledge by asking more questions, forever conducting more research and testing their theories. This is the scientific process.
Some of the research is nearly accurate, some of it is obnoxiously bias, but either way, it is all tested by hairless, two-legged primates called humans: no research is 100% full proof. Sometimes, there is correlation between two points of research, yet this does not necessarily mean one thing causes another (there can be other factors). Sometimes researchers and test subjects go into experiments with biases and can’t help but read the results differently than they would have otherwise. There are lots of reasons that humans are still quite primate-ish (I could have just said primitive but, no, primate-ish) in their research and we will get into that more below.
Correlation Does Not Imply Causation
Correlational studies are a little different than lab experiments in regards to research methods.
Correlational studies, like lab experiments, look at how different variables affect each other, but in a less controlled environment. In fact, correlational studies tend to research people in natural environments instead of labs, or collect data from surveys. This is good, because we don’t live in controlled labs anyways (sure, i.e., lab rats live longer on less calories in labs, but they have lower immunities that may kill them sooner in the real world).They look for positive and negative associations between variables and can help us predict outcomes.
But they cannot guarantee outcomes. For example, studying may be associated with higher grades, although it does not guarantee them. And high television consumption may be associated with lower grades. We cannot know which variable causes the other: do higher grades cause students to study more or does studying more cause higher grades. There could be a third, forth, or fifth variable that we haven’t even considered. Quoting every science teacher ever, “correlation does not imply causation.”
Still, correlational studies help us make predictions, so these are worth investigating as well. Looking at studies that show positive or negative correlations between longer term antibiotic use and longer term remission can help us predict what would happen if we were to take antibiotics for a long time. However, these studies won’t prove for a fact that longer term antibiotics increase or decrease our chances at remission, since there are other variables we may not even think about.
Think about it, higher coffee consumption positively correlates with heart disease, however higher coffee consumption also positively correlates with consumption of junk foods. Recognizing this third variable, we can no longer predict that someone who drinks coffee but eats “healthy” otherwise will get heart disease. Imagine a correlational study on long-term antibiotic use: we might consider that people who do take long term antibiotics have more money to spend and possibly easier circumstances. These variables may increase their chances of getting better. We might also consider that people who do not take long term antibiotics but instead opt for a more natural approach may be more proactive in supporting other aspects of their health (sleep, diet, etc). Alternatively, some other people might take long term antibiotics because they are covered by insurance, some spend every last penny they can on antibiotics, and some others take antibiotics in conjunction with every holistic-trick in the book. Meanwhile, some people who do not take long term antibiotics don’t do much for their health except keep trucking through as their health deteriorates. Lots of third, fourth, fifth, factors get overlooked in correlational studies.
Full disclosure: a lot of info comes up when you search “correlational studies for long-term antibiotic use” on the internet. Researchers have correlated antibiotic usage with antibiotic resistance, and changes involving immunity, obesity and the microbiome. I am uncomfortable bringing up these studies, because of the many unconsidered variables. Lyme Disease patients are especially unpredictable, since so many of us are on a cocktail of drugs, deal with complex infections (sometimes antibiotics are the lesser of two evils), and take varying degrees of probiotics – among other variables.
The Optimal Way To Examine Cause And Effect: Controlled Experiments
Researchers look for cause-and-effect when conducting experiments. In other words, they want to know if one thing causes another thing. For example, they want to know if antibiotics eradicate Lyme, if eating fiber improves the GI composition, or if caffeine causes short-term stimulation. When trying to prove that one variable is the true cause of an outcome, the experiment should be well designed. If it’s poorly designed, then we cannot be as confident that the controlled variable being tested is the sole cause of the outcome.
Well-designed experiments depict cause-and-effect pretty effectively. Experiments involve manipulating one or more variables, observing and measuring those variables effects on other variables, and reducing or eliminating all other factors so that the cause-effect conclusions drawn can be as accurate as possible. The manipulated or controlled variable is also known as the independent variable. So for example, in an experiment where researchers would like to find out if using a cellphone while driving is dangerous, the manipulated variables in the experiment would be: one, putting a participant on a phone and, two, having them drive. Then researchers would remove all other variables (another person in the car, a headache, or anything they think of that could manipulate the results) to see if the effect would equate to dangerous or safe driving.
Controlling external factors help rule out alternative explanations. Other factors will change the results of the experiment, such as whether the participants are all in the same age group or range from new drivers to senile folks. Well-designed experiments are usually done in lab settings. The results of an experiment can be determined when particular combinations of variables produce distinct effects.
Threats To The Validity Of Research
As valid as some experiments are, there are factors that threaten the validity of research that we should consider when catching up on the latest in Lyme research.
Sometimes the controlled or independent variables being tested get mixed up with uncontrollable variables without the researchers being aware. Do you recall hearing years ago that a study “proved” that students performed better on exams after listening to Mozart? Playing Mozart was the independent variable in one group of students, while playing no music at all was the independent variable in another group of students. Later on, two researchers from Canadian universities decided to reframe the experiment – this time they made three control groups: one where they played happy music, one where they played sad music, and one where they played no music. They found that students tested best after listening to happy music, second best after silence, and worst after sad music. They realized that they had neglected to consider something in the previous experiment: the mood of the students. It was not necessarily that Mozart increased studying productivity, but rather that upbeat-Mozart increased studying productivity. This confounding of variables ruined the validity of the Mozart effect.
Another way to ruin validity is through placebo effect, where a participant’s expectation of receiving a treatment produces a change in behavior. You are probably familiar with this. What a lot of people forget is that placebo happens to researchers as well. Experimenter expectancy effect involves the “subtle ways that a researcher’s behavior influences participants to behave in a manner consistent with the hypothesis being tested.” Researchers can reduce experimenter expectancy effects, for example, by being kept blind to the hypothesis of the experiment. In a double-blind procedure, both the experimenters and participants are kept unaware of the research condition to which the participant has been assigned.
Then, as I mentioned earlier, we must consider that experiments done in lab settings do no always play out the same in real life. Rats, given antibiotics their entire lives, also known as “germ free rats,” can live longer in lab settings than in nature. Why? Because they can be kept away from germs forever in a lab, so their immunity (which it turns out, is terrible) is never challenged. The same applies to lab rats who are fed 60% less calories than rats on normal rat diets: they live longer in the lab but not outside of the lab. This is where a correlational researcher might jump in to tell you that their research positively correlates high chances of getting a cold one week after stopping a round of antibiotics.
To recap, experiments involve independent variables that are manipulated to measure their effects. Experiments are usually done in lab settings and are therefore optimal for examining cause-effect relations. We do, however, need to remember that lab settings can give us different results than natural settings, where people tend to exist. Experiments will control external factors to the degree that researchers are aware how to. Controlling external factors can rule out alternative explanations. Uncontrollable variables, placebo, and experimenter efficacy effects can undermine the validity of experiments.
Other Research Methods
Experimental research done in labs is typically the best research method for examining cause-effect relations, and I think it’s valuable for all of us to educate ourselves on the latest experimental research on Lyme. However, I think it is easier and more interesting to most people to study up on sources of info that use other – more fun and personable – research methods.
Forums, Surveys And Questionnaires
One source of “research” that people like to get their info is from forums. There are tens of thousands of people with Lyme Disease communicating over forums about what treatments are working for them and what treatments have set them back. These forums are great because they allow questions to be answered by a much larger population than a typical experiment. They can give us information about the broader population.
Surveys involve questioning participants. They tend to study a sample of people drawn from the larger population and they can be great for collecting a lot of info, revealing the opinions, experiences and lifestyles of the masses. Surveys are similar to forums in that they yield information from a broader population. They, too, involve participants answering questions.
The trouble with both forums and surveys is that the results may mislead you: surveys rely of self-reports, which can be skewed. Do not mistake answered questions on forums with proof. When people are a legend in their own minds or, more commonly, are affected by something called social desirability bias – the desire to say/do/believe what your friends would admire most or what would get the most sympathy votes – surveys and forums lose validity.
I’m guilty of coming to conclusions about diets, foods, supplements, books, restaurants, etc, due to reviews that I read online. It’s easy to, because we are hearing the stories of real-life, likeable, reasonable people. But it’s important to remember that everyone is different, and everyone is prone to overlooking research variables sometimes. Drawing conclusions from these info sources can also mislead us to generalize and assume that a particular consensus would apply to the general population. Remember, these sources of research can be greatly unrepresentative. Finally, there are no controlled variables in surveys, so we cannot draw conclusions about cause and effect through this type of research.
Internet questionnaires – such as the one Joey and I used to help us gather feedback for the making of our book – have their own problems. Researchers do not have much control over the quality of the data they collect in this way. The people who participate in these questionnaires can easily lie if they want to. The anonymity of online surveys, forums and the like can also fashion an air of maliciousness that real life conversations probably wouldn’t involve – and that’s just annoying. Say it to my face, bro!
Keeping in mind the drawbacks of getting your info from sources like forums, I think they are still an essential tool to Lyme patients. They are good places to go and ask the most bizarre of questions. They connect us to other people going through similar things. Taken with a grain of salt, forums give you a loose idea of what people think is working and not working for them personally. I may prefer to get my info from science journals that go over the results of well-designed experiments, but I prefer to get my info from forums over doctors who are going to profit off me whether their offered treatments work or not. I like to see how real life people feel during a treatment that I am thinking of starting for myself.
Case studies examine specific people, groups or events and how they change over time. Any of the above research methods (correlational studies, experiments, and surveys) can be characterized as case studies if the variables in question are documented as they progress. If there were more studies specific to Lyme Disease patients out there, then the outcomes of treatments would be more predictable, and we would be more confident about the effects of independent variables on Lyme survivors and relapsers.
Treatment trials on Lyme Disease are unfortunately rare. I don’t think this means you should wait to treat it until an appropriate amount of research has been done. But if you can acknowledge that Lyme research is lacking and therefore Lyme “experts” are – in some ways – over their heads and guinea-pigging their patients, then you may be a little more cautious in how you go about your treatments and research. It brings me back to my point on not putting your dozen eggs in one basket. Definitely do not cling on to your doctor’s advice like gold during a recession. The reality is, no “expert” out there has cured 100% of their patients, or knows the “verdict” on every treatment protocol. We all need some patience for more Lyme-specific case studies to be done before experts like that can exist.
Ah, and to finish off this section on research, confirmation bias is the tendency to look for information that reinforces our own beliefs. If you only read from one newspaper, if you only go on websites that are “pro-paleo,” or if you search for “the health benefits of gingko” or “long-term antibiotics and Lyme success” then you have experienced confirmation bias.
The internet is a big place. Whatever you want to find, you probably will. If you want to believe your treatment protocol or diet is “superior” to other protocols or diets, you will find lots of information and educated-opinions out there to confirm this. On the other hand, if you want to be open to information that disproves your own theories, you might also search for “the health risks of gingko” or “long-term antibiotics and Lyme failures” on the internet.
I can’t be sure, but I’d imagine keeping an open mind to the possibility that you, your doctor, and your research sources are wrong positively correlates with higher chances of remission!
All At Once
There are, for lack of a better word, sketchy treatments that many of us are curious to try because of what we’ve heard or read. Can you be open to hearing what you don’t want to hear, be brave enough to try alternative treatments and be prudent enough to look at other angles all at once? Lyme Disease is a costly illness. Inevitably we have to proactively do research and constantly seek new truths.
The absence of proof is not the proof of absence, sure. But the presence of proof does not discount the possibility for future studies to discount it.