"Causes" vs. "Worsens" vs. "Improves"

Common assumptions of related phenomena

Pathology is the study of disease. It’s often interested in concepts like etiology (the cause or causes of a condition), pathogenesis (how a condition progresses throughout time, and to a lesser extent how it responds to various inputs), diagnosis (the ability to connect particular measurable markers to a known disease and the mathematical confidence possible), and treatment (improving or managing the condition).

These concepts often have significant overlap in practice. If you know what causes a condition, you can often predict what will make it worse and what will improve it to some extent, as with weight gain. Weight gain is caused by an excess of calories in most cases, is exacerbated by a continuing excess of calories, and is improved by some restriction of calories. Or consider alcoholic liver damage: this is caused by significant alcohol exposure, exacerbated by ongoing alcohol exposure, and treatment generally requires reduction in alcohol consumption among more direct intervention.

Conditions affected by lifestyle choices have high salience for most of us - we’re aware of our own degree of control over the outcomes, and we see the clear connection between what causes the problem in the immediate sense and what makes the problem better, sometimes treating it altogether. Because we tend to have high awareness of conditions we’re expected to avoid or control, there can be some slippage in our thinking about cause and treatment - people often assume that causes and treatment are neatly connected, or at least that what makes a condition worse necessarily or likely causes a condition. Our intuitions here are often wrong.

Treatment is Not Causal Information

A lot of the time, perhaps most of the time, the overlap between etiology, pathogenesis, diagnosis, and treatment is a lot messier than it is for mild weight gain. Knowing what causes a condition is useful but insufficient for knowing how to treat it - that lung cancer is often caused by smoking may help with interventions that can reduce or prevent getting the condition in the first place, but other insight is necessary to develop treatments that can improve the lives of those diagnosed with lung cancer.

Similarly, something that worsens the course of a disease will not cause it to appear in otherwise healthy people. It’s slightly cheap of me to use a genetic condition here, but since those are relevant case studies in this framework, I’m doing it - getting a cold can worsen cystic fibrosis, but it’s not going to cause it in someone without the genetic signature for the condition.

Diagnostics can take advantage of etiological or disease-causing knowledge, especially in infectious disease: in order to determine whether someone has strep throat, the most common diagnostic tool is to swipe the throat and test or culture it for streptococcal bacteria. Here as elsewhere, the salience and simplicity of this model tends to have us assuming that if you know what causes a disease, you probably know how to diagnose and/or treat it, and if you don’t know what causes a disease, you’re significantly hampered in both.

While there’s some truth to this - endometriosis has a famously evasive and still to my knowledge unknown cause/etiology, and treatment and diagnostic tools are limited - it’s not universally true that lack of causative information hampers diagnosis and treatment. It’s especially not true that knowing what causes a condition always helps you diagnose it or treat it. There are several conditions we can diagnose and/or understand causality for that we don’t have adequate treatment for, such as Huntington’s. There are also conditions with known causes that can’t be diagnosed with laboratory precision. Tetanus, although known to be caused by exposure to an output of a particular kind of bacteria, is diagnosed with clinical signs - how a patient behaves - not with a lab test.

Case Studies: Sugar in Diabetes and Cancer

You’ll often see people claiming that eating a lot of sugar is bad for you, sometimes specifically claiming that it causes diabetes or cancer. There are obvious reasons for this claim. A lot of sugar tends to be bad for individual health in much the way that a diet high in fried foods tends to be bad for individual health. In excess, with low addition of nutritionally important components like protein or micronutrients, and especially if the person is eating in a caloric surplus consistently, these dietary choices have a consistent relationship with weight gain and the metabolic disruption that comes with it.

In short, eating lots of sugar and/or fried food tends to accompany eating food high in calories with few nutritional benefits, and by extension, by eating more calories than the body needs - this is easy to do when your diet is fun to eat and doesn’t fill you up. This is worth avoiding, no question, as is weight gain in general.

Diabetes

It’s also true that because diabetes interferes with the body’s blood sugar management system, eating an excess of sugar, especially simple sugars that require little breakdown for the body to use or move around, tends to make the condition worse. There’s nuance here, of course - insulin can be used to offset sugar intake even in diabetics, and a high carbohydrate diet can mean several different things, with different implications for diabetics.

All of the above is often defended as a shorthand way to suggest that eating sugar, especially particular high sugar items, is causally linked to diabetes, and that if you want to avoid diabetes, you should avoid or intentionally minimize dietary sugar. To be clear, it’s wise to avoid, say, a diet that gets 80% of its average daily calories from candy, but you knew that. Dietary sugar - even in the form of refined table sugar, candy, or desserts - is not likely to cause diabetes in a healthy weight person eating an otherwise reasonable diet. To be more specific, dietary sugar is not likely to have negative metabolic effects in a healthy person who isn’t eating too many calories while eating a diet balanced between protein, fat, and carbohydrates.

Context is everything, in diets as well as elsewhere. In a person who is eating a large amount of fat every day - say, north of 40% of calories coming from fat - high dietary sugar is a much bigger problem. A diet with a high percent of calorie intake from sugar and fat is definitionally a diet low in protein and fiber. Both protein and fiber make a big difference in satiety and fullness, or how satisfied the average person feels by that diet. A diet that is long term (i.e. for more than a week or even a month) high in both fat and carbohydrates, especially simple carbohydrates, is usually going to result in a long term calorie surplus, weight gain, and eventual metabolic dysregulation.

You can eat sugar and stay healthy. You can even eat really sugary desserts and be a healthy person provided that the majority of your calories have a reasonable macro balance and you’re not consistently eating too many calories. It’s wise to avoid eating a diet that is high in both sugar and fat in the long run. If you really like different kinds of dietary fat (avocado, fish), you can eat more of them by eating fewer carbs, and if you really like carbs, you can eat more of them by eating less fat. Carbohydrates and even concentrated sugar is not a cause of diabetes, an overall poor diet with consistent caloric excess is.

Cancer

It’s also true that cancer cells use sugar differently than healthy cells do. In healthy cells, glucose and oxygen do paired work to provide energy to the cell as long as there is a reasonable amount of local oxygen (and glucose) to provide it. You’ve probably heard of aerobic versus anaerobic exercise. Aerobic exercise puts your cells to work churning through glucose and oxygen to power your body, and when your cells run out of local oxygen in this process, they switch to anaerobic energy production. Aerobic energy production is really effective and efficient, while anaerobic outputs much less energy for a similar amount of work and specifically, a similar amount of glucose input. In other words, anaerobic energy production is very sugar hungry compared to aerobic.

Cancer cells primarily use this anaerobic approach to creating energy even in the presence of oxygen, a phenomenon sometimes referred to as the Warburg Effect. This is a pretty cool and weird feature of cancer cells, one that we’ve been able to put to diagnostic use. You can give a patient with suspected cancer an injection of glucose where each molecule has a tag that will show up with imaging. In a patient with cancer, this tagged glucose will cluster on a scan to check for the molecular tags, something called positron emission tomography (PET) scanning.

Here as in diabetes, you’ll often see people claiming that this connection between sugar and cancer means sugar causes cancer and/or that it “feeds” or worsens cancer. Unlike diabetes, however, it’s not nearly so obvious that a high-sugar diet exacerbates the effects of cancer, even given the molecular differences in cellular metabolism. As with diabetes, obesity itself is a risk factor for multiple forms of cancer, including endometrial and colorectal cancers. Therefore, as with diabetes, for both prevention and management of cancer, interventions that avoid or mitigate weight gain are a better line of defense than particular interventions that target sugar.

Conclusion

It’s not always straightforwardly true that what makes a particular condition worse will give you the condition in the first place, help you diagnose it, or help you treat it. This is an error that crops up across people’s first approximations of medical conditions, but it shows up elsewhere too - it’s tempting to think that we can solve anything if we know why it happened, improve any political outcome if we can conclusively say x makes it worse. Unfortunately, in medicine, personal life, and politics, this isn’t always true - be sure to do your due diligence before assuming that identifying an issue means identifying a cure or a cause.

Related Posts: