What Looks Like Ordinary Decline

Graham Young, the “Teacup Poisoner,” killed his stepmother with thallium in the early 1960s and, after eight years in Broadmoor, poisoned his factory coworkers the same way, causing two more deaths before his own diary gave him away; the case produced Britain’s Poisons Act.¹ Thallium fits the mechanism precisely: its early symptoms — gastrointestinal distress, peripheral neuropathy, hair loss that appears only after several days — mimic ordinary viral illness, which is why Young’s earliest victims were treated for suspected natural sickness before anyone considered poison.²

Julia Lynn Turner poisoned her husband, a Cobb County police officer, in March 1995, and her boyfriend, a Forsyth County firefighter, in 2001, both with ethylene glycol, the sweet, odorless active ingredient in antifreeze, allegedly mixed into Jell-O, sweet tea, or soup. Her husband’s death was formally certified as natural, attributed to an irregular heartbeat, and stood on record for years until her boyfriend’s poisoning prompted investigators to reopen the case. She was convicted in 2004 and again in 2007, motivated by roughly a hundred fifty thousand and thirty-six thousand dollars respectively in life-insurance and pension payouts.³

Lakhvir Kaur Singh laced her former lover’s leftover curry with aconitine from Aconitum ferox in January 2009, killing him and nearly killing his fiancée, who survived only after a medically induced coma. Singh drew a life sentence with a twenty-three-year minimum⁴ — the first aconitine murder prosecution in Britain since 1882, a rarity that itself made the poison harder for physicians to suspect in time.⁵

Charles Cullen is this chapter’s most complete case: not a single dramatic poisoning but slow, repeated, low-dose administration producing a pattern of decline attributed to existing illness rather than deliberate harm. A registered nurse, Cullen murdered patients with insulin and digoxin overdoses across multiple New Jersey and Pennsylvania hospitals over sixteen years ending in 2003.⁶ Because his victims were already hospitalized with genuine underlying conditions, each individual death was, for years, explainable by those conditions alone. What broke the pattern was not any single suspicious death but a statistical one: Somerset Medical Center staff noticed unexplained medication-overdose deaths and IV bags with far more needle punctures than normal, and New Jersey’s state epidemiologist was brought in to analyze the cluster. Cullen emerged as the common factor across cases with abnormally elevated insulin or digoxin levels, and was arrested in December 2003.⁷ He pleaded guilty to twenty-nine murders and received eighteen life sentences without parole; some estimates place his actual victim count considerably higher.⁸

Three further substances deserve separate treatment, because their value to a killer lies in something more specific than resembling an ordinary illness: they resist detection by routine post-mortem toxicology altogether.

Succinylcholine, a muscle relaxant, breaks down into succinic acid and choline — both substances the body already produces, so routine toxicology has nothing distinctive to flag. Genene Jones, a pediatric nurse, used it to murder infants at a Texas hospital in the early 1980s; she was convicted in 1984 for the death of a fifteen-month-old whose body had to be exhumed and specifically tested for succinylcholine metabolites before the poisoning could be confirmed — detection was possible only once investigators knew precisely what to look for, not through any routine screen.⁹

Insulin poses a related problem: it is a natural hormone, and a fatal overdose leaves no distinctive marker by itself. The tell is a biological pairing — the body produces insulin and C-peptide together in a fixed ratio, but injected insulin arrives with none, so an abnormally skewed insulin-to-C-peptide ratio in postmortem blood becomes evidence of external administration.¹⁰

Potassium chloride is, in the forensic literature’s own phrase, “a forensic medicine enigma.”¹¹ Potassium naturally rises in blood after death regardless of cause, which makes a fatal injected overdose extraordinarily hard to distinguish from the body’s normal postmortem chemistry. One case cuts against any temptation to treat pattern-based toxicology as automatically reliable: an Italian nurse, Daniela Poggiali, was convicted of murder by potassium chloride injection in 2016 on a statistical model predicting expected postmortem potassium levels — a model later found flawed, resulting in her acquittal at a second trial with more complete analysis.¹²

Every disguise here shares one weakness: it depends on each death being read alone. Read the deaths together and the plausible-illness story stops holding. Poggiali’s overturned conviction marks the limit of that method: the same statistical lens that catches a poisoner can convict an innocent one, which is why the pattern has to be built from real markers, not inference alone.

One thing has to be said plainly to the reader this chapter could frighten, because the chapter’s own logic requires it. The overwhelming majority of illness is illness, and the overwhelming majority of slow decline is exactly what it appears to be — an ordinary disease, or a body, running its course. Deliberate poisoning dressed as natural decline is real, documented, and prosecuted, which is why it sits in this book; it is also rare, and rarer still against any particular person, because it demands a motive, a sustained access, and an appetite for risk that almost no life contains. What separates the rare true case from ordinary misfortune is never the feeling that something is wrong — that feeling keeps company with grief and real disease and anxiety alike — but the two things that convicted the poisoners above: a forensic marker a laboratory can name, such as the skewed insulin-to-C-peptide ratio or the metabolite found only once someone knew to test for it, and a pattern across many deaths that no single illness explains. A fear that you are being poisoned is a hypothesis, and it is answered the way every claim in this book is answered — against evidence outside your own body, by the discipline Chapter Eleven sets out — not confirmed by how unwell you feel. Most often that test returns the ordinary answer, and the ordinary answer is not a lapse in vigilance. It is the truth, and it frees you to treat what is actually wrong.