Is obesity causing diabetes and other illness, and not just correlated with?

Question

According to the World Health Organisation website: (emphasis mine)

What are common health consequences of overweight and obesity? Raised BMI is a major risk factor for noncommunicable diseases such as:

• cardiovascular diseases (mainly heart disease and stroke), which were the leading cause of death in 2012;
• diabetes;
• musculoskeletal disorders (especially osteoarthritis – a highly disabling degenerative disease of the joints);
• some cancers (including endometrial, breast, ovarian, prostate, liver, gallbladder, kidney, and colon)

The risk for these noncommunicable diseases increases, with increases in BMI.

However, according to my observations in life, people with similar eating habits, sometimes tend to react differently and have different body sizes. My reasoning is as follows:

They did experiment on an obese population, using their BMI (a lame calculation based on weight and height) as a discriminating factor. Within that obese population, people with bad eating habits (such as junk food, nibbling, or eating disorders) will be over-represented. And those bad eating habits could just as well be themselves the cause of diabetes and other illnesses (as opposed to obesity alone). In other words, this is a major confounding factor.

Now I would like to know if there's any study based on an overweight population that stuck to healthy eating habits (wholesome food, 3 meals per day of normal quantity), and if that group had more risks of diabetes or other disease than thin people with the same regime. If the above mentioned illnesses are really due to obesity, then this should be the case. But I'm skeptical.

Were two populations with similar food regime, one obese and one not, ever compared?

My point skepticism comes from my suspicion that junk food is the problem, not the fact that some people have different metabolisms.

Answer 1

It's not completely clear from your question whether you want to know if obesity causes illnesses, or you want to know if diet causes illnesses even after controlling for obesity. I'm going to assume you are mainly concerned with whether obesity has a causal relationship to illnesses, or is just correlated with them.

This question has been addressed by the medical literature, with a 2017 meta-analysis by Franks and Atabaki-Pasdar, titled Causal Inference in Obesity Research, published in the Journal of Internal Medicine.

The main conclusions are:

• probable causal relationships between adiposity (another term for obesity) and bone health/disease, cancers (colorectal, lung, and kidney cancers), cardiometabolic traits (blood pressure, fasting insulin, inflammatory markers and lipids), uric acid concentrations, coronary heart disease and venous thrombosis (in the presence of pulmonary embolism)
• possible causal relationships between adiposity and gray matter volume, depression and common mental disorders, oesophageal cancer, macroalbuminuria, end-stage renal disease, diabetic kidney disease, nuclear cataract and gall stone disease
• no evidence for causal relationships between adiposity and Alzheimer’s disease, pancreatic cancer, venous thrombosis (in the absence of pulmonary embolism), liver function, and periodontitis

This study is a systematic meta-analysis of research on causal relationships between obesity and other factors. This means the authors examined many different studies, assessed the quality of the evidence offered by each study, and summarized the findings in aggregate.

Disentangling correlation from causation is much more complicated than the old adage "Correlation does not necessarily imply causation", and there has been a big change in statisticians' attitudes toward it over the last two decades. We now have a family of techniques called 'causal inference' that allow us to say whether certain factors do or do not cause certain outcomes, given all the other factors that might be confounding them, without necessarily using a double-blind controlled trial. The technique that seems to be widely used in obesity research is Mendelian Randomization Studies. The quality of evidence from these studies can vary. This is why the paper includes some findings of possible or probable causal relationships.

If you are skeptical of the authors' analysis, the paper contains both supporting tables that summarize the quotes from relevant papers on which they based their assessments, and of course, you can always follow up with the sources themselves (often meta-analyses or analysis against very large datasets, sometimes with tens of thousands of subjects).

Answer 2

Type 2 diabetes is caused by consuming too much easily digestible carbohydrates (for example sugar). Consuming too much carbohydrates will also tend to make you obese.

I was also told that being obese makes it harder for your body to cope with diabetes (it makes many things harder), but obesity and type 2 diabetes are most likely caused by the same root cause.