One Problem, Thousands of Names: Your Diagnosis Is a Label, Your Cells Are the Story

How many diagnosis codes are there?

There are 74,879 billable diagnosis codes in US medicine right now, and 190 more were added on October 1, 2026 alone (the CMS code files, linked under Resources). The number you will hear quoted is 36,000 disease names, so the real list is about twice that. Each code names where the trouble showed up. None of them names why. A diagnosis is a code in a copyrighted classification, the WHO's ICD, run in the US by the CDC and CMS. It is used to bill, to prescribe, and to schedule surgery for symptom management, and every one of those is paid. It is not an explanation, and it is not a plan. To see what a marker on your own report actually measures, start with the Lab Atlas.

Who decides when a disease definition gets wider?

Expert panels redraw disease definitions, and many of the people on them are paid by drug companies. When the line moves, millions of people cross it overnight. A study of guideline panels that widened common disease definitions found 75% of the panel members had ties to drug companies, and not one panel seriously weighed the harm of labeling more people sick [1]PubMed. The psychiatric manual most clinicians diagnose from was written by a panel where 60% of members took industry payments, $14.2 million in total, during the years they wrote it [2]PubMed. Across US guideline panels, 62% of members took at least $500 from industry, and 17% of those people declared no conflict at all [3]PubMed. We cover where the money sits in Health Is Not a Subscription.

Is renaming a disease data manipulation?

Yes. Change the name and you change the count, and that is data manipulation. It makes a trend disappear, or appear, without a single body changing. When the CDC tracked sudden infant deaths from 2015 to 2020, its own analysis said the 2020 rise in SIDS came partly from diagnostic shifting: the same deaths filed under a different heading [4]PubMed. Sudden infant deaths now sit under a wider umbrella, sudden unexpected infant death. Compare a year before the switch with a year after and the trend line tells you about the filing system as much as the babies. PCOS is the newest example. It is now published as polyendocrine metabolic ovarian syndrome, PMOS [5]PubMed. Defined simply, it is a preventable metabolic diagnosis: insulin resistance, raised androgens and irregular cycles, driven by inflammation and the gut [6]PubMed. Nutrition is the primary lever, not an add-on [7]PubMed. The new name finally says metabolic. It also resets the count. Every rename resets the count, and a reset count hides whether anything is getting better. Children carry the most of it, as we lay out in More Medicine, Sicker Children.

Do normal lab ranges keep moving?

Yes. Cholesterol, fasting glucose and blood pressure all had their lines moved, and each move put millions more people in the patient column without a single body changing. Before 1984, most US physicians treated total cholesterol only above 300 mg/dL, some above 250 (Kannel 1986, under Resources). Then came a national cholesterol program, the first statin in 1987, and targets that kept dropping. In 1997 the fasting glucose line for diabetes dropped from 140 to 126 mg/dL, and in 2003 the floor of impaired fasting glucose dropped from 110 to 100, creating a far bigger prediabetes category [8]Source, [9]Source. In 2017 the blood pressure line for hypertension moved from 140/90 to 130/80 [10]Source. That one document labeled about 31 million more US adults hypertensive (Muntner 2018, under Resources). And the drug at the end of the cholesterol line: across 21 statin trials, the absolute drop in death was 0.8 percentage points, sold as a 9% relative cut (Byrne 2022, under Resources). The numbers keep moving to fit a sick society, and every move makes more customers for more medication. That is the same data manipulation, done with a ruler instead of a name. The full timeline, with the documents, is in Cholesterol, Statins, and the Numbers That Changed. To read your own three numbers, use the Lab Atlas.

How much chronic disease is preventable?

Most of it. The American Heart Association's own Food Is Medicine advisory says diet is a major risk factor in 90% of US healthcare spending on chronic disease [11]PubMed. A review of cancer causes put only 5 to 10% on inherited genes and 90 to 95% on environment and lifestyle [12]PubMed. A 2026 analysis found that better diet quality and more movement each cut chronic disease risk, together by 46%, at any body weight [13]PubMed. 95% of diagnoses are preventable metabolic disorders. A preventable disease is one with a cause you can remove: too little of what a cell needs, too much of what it cannot clear. Deficiency and toxicity. When you are missing nutrients, the body cannot do its six jobs right, and that is what a symptom is. Your lab markers are where it shows first. If you want the basics in one sitting, read Nutrition 101, then see how the food system got us here.

What does a standard blood test not measure?

Lab markers should be where you catch it. A standard panel measures what your organs are putting out. It rarely measures the raw materials going in, or the toxins sitting on top. A blood test is a tool, not a verdict. We explain the measurement side in Nutrition Diagnostics.

• A thyroid panel does not measure iodine. In US adults, TSH and T4 did not track iodine status at all, so a normal panel tells you nothing about whether you have enough [14]PubMed. Iodine status is read from urinary iodine and thyroglobulin [15]PubMed. See How Do I Read My Thyroid Labs? and The Mineral Nobody Talks About.
• Almost nobody orders a urine glyphosate test. In a group of pregnant women in Indiana, 93% had glyphosate in their urine, higher levels went with shorter pregnancies, and none of their tap water had it, so the route was food [16]PubMed. Glyphosate turns up in urine, blood and breast milk worldwide [17]PubMed.
• It comes down fast. Families who ate fully organic for six days cut their urinary glyphosate by about 71% [18]PubMed. The produce with the most residue carries it on 96% of samples [19]Source. Start with the 10 everyday swaps that cut your toxic load.
• Pesticides are one load among many. The same blind spot hides formaldehyde, which we trace room by room in Where Is All the Formaldehyde Coming From?, and what comes out of the tap, covered on Water.
• Both directions, held together: a large cohort found that eating more fruit and vegetables was not linked to cancer at any residue level [20]PubMed. The answer is not fewer plants. It is cleaner plants.

Which nutrients are Americans most often low in?

Potassium, magnesium, vitamin E, vitamin C, iodine and selenium are the six with the deepest research and the widest shortfalls. Nature's love language is nutrition: minerals are the building blocks, vitamins run immune function, hormone production and antioxidant defense, and fats and proteins are the raw material for hormones and every cell wall. See Minerals for the full set. Most Americans are short. In 26,282 US adults, 84% ate less vitamin E than they need and 46% less vitamin C [21]DOI. An older national analysis put vitamin E shortfall above 90% [22]PubMed. Adults who met their needs for vitamin E, magnesium, potassium and fiber had lower death rates from all causes [23]PubMed. Four of the six are the Simple Stack: magnesium, vitamin C, vitamin E and potassium. The other two, iodine and selenium, work as a pair.

• Potassium. Over 18 years, adults with low potassium had 38% higher death from all causes [24]PubMed. In 416,000 people, more sodium relative to potassium meant more heart and overall deaths [25]PubMed. More in Stop Drinking Salt and Calling It Hydration.
• Magnesium. A magnesium deficiency score was linked to 30% higher death from all causes [26]PubMed. Across 4.3 million Danes, the lowest-magnesium tap water went with 22% more heart-attack deaths [27]PubMed. More in The Two-Point Mineral.
• Vitamin E from food. More vitamin E intake was linked to less heart disease and lower death [28]DOI, except in current smokers [29]DOI. Held together: very high blood levels went the other way in one survey [30]DOI, and another found no link at all [31]PubMed. The food story holds. The megadose story does not, and the form matters, as covered on Vitamin E and in They Sell You the Wrong Vitamin on Purpose.
• Vitamin C. Higher blood vitamin C went with lower death in 11,539 US adults [31]PubMed.
• Selenium. In 20 cohort studies, each step up in selenium meant 13% lower death from all causes and 15% lower cancer death [32]PubMed. Low selenium plus lead or cadmium carried a much higher death risk than either alone, deficiency and toxicity meeting in one person [33]PubMed. Held together: high intake was also tied to more type 2 diabetes [34]PubMed. More in Why Does Your Thyroid Need Selenium?.
• Iodine. Nearly half of pregnant American women fall below the requirement, even with prenatal vitamins [35]DOI. The full story is on the iodine page.

What should you eat after the six nutrients?

Real food, real fiber and fewer chemicals. Fiber feeds the trillions of bacteria in your gut. In 39 of 42 studies, whole-grain fiber made the gut community more diverse [36]PubMed. Fiber- and polyphenol-rich eating shifted both the microbes and the switches on your DNA in a healthier direction, and the Western diet pushed both the other way [37]PubMed. That diversity reaches the mind: teenagers with depression carry less diverse gut bacteria [38]PubMed. A more diverse gut handles stress better, mental and emotional included, and the emotional side is its own input, which you can check with Love or Fear. None of this is sexy. It does not come in a bottle with a launch party. It is also the part with the most research behind it. Build a week of it in the Diet Builder.

How does a drug look 95% effective when it helps 1 person in 100?

With the same data manipulation, aimed at you instead of the statistics. The labels, the renames and the moving lines decide how many people get counted sick. Relative risk decides how good the fix looks. It is deceptive marketing, and it is legal. Here is the math with a made-up example: if 2 out of 100 people have a heart attack without a drug and 1 out of 100 with it, the ad says the drug cuts heart attacks by 50%. The real change is 1 person in 100. That means 100 people take it for 1 to benefit, and all 100 carry the side effects. Ask for the absolute number every time. The full method is in How to Read a Health Study, Number Needed to Treat, and how pharma studies get dressed up.

• Doctors fall for it too. Shown the same trial as relative risk instead of absolute risk, physicians judged the drug far more effective [39]PubMed.
• Across 35 studies, the relative number felt bigger and more persuasive every time [40]PubMed. The same 50% cut can mean almost nothing or a lot, depending on the starting risk [41]DOI.
• Statin trials used exactly this framing to make small absolute benefits look large [42]DOI.
• The COVID-19 mRNA trials reported relative risk reductions above 94% and left out the absolute reduction, 0.7 to 1.1% [43]DOI. The relative figure kept overstating it in 2023 and 2024 [44]DOI. That 95% became "safe and effective" on every screen. One paper argues trial-period absolute numbers understate real-world effect [45]DOI; the public was still only ever shown the relative one.
• That campaign is now in court. In the Netherlands, vaccine-injured plaintiffs sued 17 defendants, including Bill Gates, Pfizer CEO Albert Bourla and the Dutch state, in July 2023. Their position: the mRNA genetic injection was never proven safe, people were coerced into it, and on the evidence presented there is no difference between it and a bioweapon, which they charge as crimes against humanity. The court did not throw it out. On October 16, 2024 the Leeuwarden court ruled it has jurisdiction over Gates and ordered him to pay the costs of his failed objection, and the case moves to a hearing on the merits, listed for October 22, 2026. The court is now weighing the evidence; it has not yet ruled on the charges (court records under Resources).
• Who paid for the study is half the story. See Who Funded the Study That Shaped Your Plate?.

Can medical care itself cause harm?

Yes, at scale. Medical error is described in the research as the third leading cause of death in the US [46]PubMed, [47]PubMed. One analysis ranks prescription drugs third on their own, and about half of those deaths came from drugs taken exactly as prescribed [48]PubMed. Vaccine injuries are undercounted the same way. A federally funded project at Harvard Pilgrim found that fewer than 1% of vaccine adverse events are ever reported to VAERS, the national reporting system [49]Source. Providers were told the shots were safe, and an injury everyone assumes is impossible rarely gets written up. On October 8, 2026, HHS finally moved: it will reimburse doctors for reporting suspected side effects, require electronic health records to capture them, modernize VAERS, and opened an NIH clinic to study vaccine injuries (HHS initiative, under Resources). That is how you get accurate data. That is not a reason to avoid care. It is a reason to ask what caused the problem before you accept a lifetime refill for it. The coverage side of that bill is in Insurance Is a Scam.

Why is there a drug for everything except nutrition?

Because nobody bills for the potassium you never ran short of. Medicine is a business, and the business is paid per diagnosis, per visit and per refill. Children's medicine is one of its biggest earners, so there is little money in a healthy kid. There is a drug for almost every label and no drug for the root, because the root is food, minerals and what you keep out of your body. Do not hand your health to institutions that take money to shape the rules. You can eat well enough to move your own labs and get on with the rest of your life. Read your markers in the Lab Atlas, cover the basics with the Simple Stack, and browse every free tool.