First, let me discuss natural antibodies. We often think of these a poisons targeting particular microbes, but they are not. All natural antibodies do is mark an organism for attack by other components of immune systems. If all antibacterials we call antibiotics behaved this way there would be fewer problems with them. Except for monoclonal antibodies, for which drug names typically end in -mab, common antibiotics are derived from natural defensive measures one kind of microbe uses against others, not host antibodies. One irony in this is that many of these are toxins produced by fungi, mycotoxins. You can find other threads on sensitivity to mycotoxins in ME/CFS patients. Others are toxins produced by beneficial bacteria to fight competition. It is far too easy for the immune system to become convinced this sudden increase in bacterial toxins results from overgrowth of harmless bacteria. There is still another way immune systems decide which of a slew of possible microbes is causing the problem. After a successful immune response kills off large numbers of microbes, and cell signalling indicates the infection is resolving, fragments of the defeated invaders are carried to lymph nodes, where many kinds of immune cells learn what invaders are responsible for the attack. If you accidentally kill off the wrong bacteria, the odds are weighted toward immune response against these in the future. People who grow up on farms are surrounded by microbes, but there is a difference in types: most of these are not specifically adapted as human pathogens. Their immune systems generally do better at distinguishing real human pathogens from harmless microbes. For people who grow up in cities things are quite different. A high percentage of microbes they encounter will be human pathogens, because there aren't too many other hosts around. Non-human hosts in this unnatural environment are likely to be classed as vermin, and those that are particularly successful have a long history as vectors of human disease. The historical misunderstanding which is still with us is that antimicrobial action is simple chemical toxicity. When you get exactly the right antibodies the selectivity is like magic, with very low toxicity. The earliest antiseptics were not just toxic, but often caustic. It took a long time to realize that putting them in wounds slowed healing. The research on this took place during WWI. You can decide how long it took doctors to see the obvious. The first antibacterials against treponema pallidum, like Salvarsan, actually contained arsenic. Paul Ehrlich thought of chemical agents as magic bullets (actually "Zauberkugeln" or "magische kugel". Ehrlich was merely talking about highly specific chemical agents, using a metaphor derived from hunting with a shotgun.) Much better antibacterial agents for this were eventually found which contained no toxic heavy metals at all. The first patented antibiotic, Prontosil, was developed from dyes, based on the incorrect idea that binding a toxic chemical to bacteria was necessary to get rid of them. It was later found that a chemical which had been known so long it could not be patented, sulfanilamide, was the active ingredient, and more effective by itself. Sulfanilamide has exceptionally low toxicity, which was a good thing when it was liberally sprinkled on open wounds during WWII. Its mechanism of action is much more subtle. In a disease like AIDS, where the patient's immune system fails, various antimicrobial agents are of little use unless they stop the activity of the specific virus responsible. In most other cases, it is the patient's immune system which ultimately causes the infection to resolve. This doesn't stop doctors from trying to poison whatever they think is causing a problem, usually based on totally inadequate evidence, and frequently poisoning patients. We don't know the long-term consequences of such ignorance, but these can't be good. One reason research on chronic infectious disease has been severely neglected is that it wouldn't do to discover that common illnesses of unknown etiology are iatrogenic.