The adult nobody protects: addiction, fluctuating capacity, and the losses that look like consumption

He is forty‑one. He receives a disability benefit, lives alone, and has been in and out of treatment for opioid use disorder for nine years. A man he calls a friend holds his debit card, "to help him budget." The rent is two months late. The account is emptied on the third of every month, and the cash machine that appears on every statement is a five‑minute walk from the friend's door. Nobody counts him. He is too young for elder‑abuse statistics, too lucid on a good day for a finding of incapacity, and too easy to blame: when the money is gone, everyone assumes he spent it himself. He is a composite, built from configurations that recur in court‑supervised practice. He is also a protection case the American system was not designed to see.

A population the numbers cannot find. In 2024, 48.4 million Americans aged twelve and over, one in six, met the criteria for a substance use disorder; 28.2 million had a drug use disorder, and 21.2 million adults lived with a substance use disorder and a mental illness at the same time. Among those who needed treatment, about one in five received it. The overdose curve, meanwhile, has finally bent. According to provisional counts released by the CDC in May 2026, an estimated 70,000 Americans died of drug overdoses in the twelve months ending December 2025, roughly 14 percent fewer than the year before and the lowest twelve‑month count since 2019, after a peak near 110,000 in 2023. That is the best news this field has had in a decade, and it has a consequence that nobody has budgeted for. Every death averted is an adult who goes on living with a disorder that fluctuates, a benefit that arrives on a fixed date, and, very often, someone nearby who knows the date.

The protective apparatus was built for a different person. In the most recent federal Adult Protective Services data, 81.9 percent of substantiated victims were sixty or older; adults aged eighteen to fifty‑nine were 16.8 percent. Self‑neglect accounted for 65.3 percent of victims, financial exploitation for 15.7 percent. And the national reporting tables provide no usable measure of substance use among victims or perpetrators. The federal definition of elder financial exploitation used by bank regulators starts at sixty. Yet the FBI's complaint data for 2025 show where the money actually goes: of 20.9 billion dollars in reported losses, 7.7 billion came from victims over sixty, and roughly 9 billion from victims under sixty, including 3.7 billion among people in their fifties and nearly 3 billion among people in their forties. The United States collects detailed fraud data by age. It collects far less systematic information on the financial exploitation of younger vulnerable adults, and almost none on its intersection with addiction.

Why this exploitation is invisible. Four reasons, and they compound.

First, the losses look like consumption. Cash withdrawals in round amounts, at odd hours, in the same neighborhood, are what a habit looks like on a bank statement. They are also what a "friend" who holds the card looks like. A dealer who extends credit, a partner who "manages" the money, a cousin who cashes the check for a fee: none of them produces a transaction that a fraud model is likely to flag, because the pattern resembles the disorder the victim is known to have.

Second, capacity can fluctuate, while legal and clinical systems usually assess it at a particular moment and for a particular decision. A person with a substance use disorder can be fully capable on Tuesday, unable to protect himself on Friday, and capable again after ten days in treatment. An assessment on a good day may find capacity and close the file; an assessment on a bad day may find incapacity and open a broad order. Either can misrepresent the person's actual pattern of functioning if a single assessment is treated as the whole picture rather than one point on a curve.

Third, the victim is blamed. The phrase is "lifestyle choice," and in practice it closes files. In a number of states, eligibility for adult protective services turns on age or a qualifying disability, and a substance use disorder on its own does not always open the door; where it does, the case tends to be filed as self‑neglect, which is what the person is doing to himself, rather than as exploitation, which is what someone is doing to him. The exploited adult becomes the author of his own exploitation.

Fourth, addiction sits on both sides of the ledger. In a National Institute of Justice study of 948 elder mistreatment investigations, 26.9 percent of alleged perpetrators had substance use problems, and perpetrator substance use was significantly associated with financial exploitation, physical abuse, and emotional abuse. The same family often contains the exploited parent and the exploiting child: the professional who takes the parent's case meets the child's addiction, and the professional who takes the child's case meets the parent's checkbook, or the dealer's ledger. Layer on top of this the instrument designed to protect benefits. The Social Security Administration projected that some 3.3 million adult beneficiaries would need a representative payee in 2025, and among disabled workers aged twenty‑five to thirty‑four, nearly a third already had one. Payees are usually relatives or acquaintances. Because they often hold direct control over benefit income, misuse inside that relationship is a particularly serious safeguarding risk in this population.

The law is moving. The method is not. States have begun opening legal doors. In California, Senate Bill 43 rewrote the definition of "gravely disabled" to include a severe substance use disorder; counties could defer implementation until January 1, 2026, and Los Angeles County, the largest, adopted it on that date. An adult whose addiction leaves him unable to provide for his personal safety or necessary medical care can now be placed under a conservatorship that, three years ago, was reserved for mental illness. In Florida, the 2024 reform of the Marchman Act consolidated the two petitions previously required for involuntary substance use treatment into one and allowed witnesses to appear remotely; state officials told legislators in November 2025 that the reforms were paying dividends. These are real changes. They are also doors without a corridor behind them.

A conservatorship for a forty‑one‑year‑old with opioid use disorder is not a conservatorship for an eighty‑five‑year‑old with dementia. Capacity will come back, partially and repeatedly. Recovery research reinforces the importance of avoiding assumptions of permanent incapacity: the National Recovery Study, led by John Kelly at Massachusetts General Hospital and Harvard Medical School, found that roughly 22.3 million Americans, more than nine percent of adults, had resolved a significant substance use problem, and subsequent research found that resolution often involved more than one serious attempt, with a median of two and a mean of five. For a conservator, that possibility of recovery matters: the task is to protect the person through periods of impaired functioning without treating those periods as proof that autonomy can never return, to protect the person through the trough without confiscating the peak. There are two familiar ways to fail at it: the paternalistic conservator who turns money management into punishment and pushes the person toward the dealer's credit, and the hands‑off conservator who returns the card on the first good week and is surprised on the third of the month. Neither failure is a failure of law. Both are failures of method.

What protecting this adult actually requires. Adults with substance use disorders have been a recurring part of my caseload in Seine‑Saint‑Denis, and the second of my books is devoted to them. What follows is practice, described so that it can be adapted to any state's framework.

Document capacity as a curve, not a snapshot. At every contact, record the person's observed state, the time, and who else was in the room. Eight weeks of that log tells a court something no single evaluation can: when this person decides, and when he cannot. A measure calibrated to the curve, lighter at the peaks and firmer in the troughs, is more consistent with a fluctuating disorder than a fixed response that ignores periods of regained functioning.

Separate consumption from exploitation in the flows. Consumption is irregular and follows the person's state. Exploitation is regular and follows someone else's calendar: the day the benefit lands, the geography of another person's address, a card used while the account holder was in treatment, "loans" documented after the fact. Three months of statements, read with those questions, can reveal patterns that an interview alone may miss. In my experience, the question that opens the most files is not "how much did he spend?" but "who else knows the date the money arrives?"

Grade the influence. The dealer who has become a creditor, the partner who holds the card, the helpful acquaintance, the relative with an addiction of his own: these are not the same risk, and treating them as one produces either paranoia or blindness. The LAPA Coercive Influence Assessment Matrix grades influence as a gradient, so that the response is proportionate and the reasoning is written down.

Manage the money so that it protects sobriety. Secure the benefit at the source. Pay the essentials directly. Release the remainder weekly rather than monthly, so that a bad Friday costs a week and not a month. Leave a sum the person controls without accounting for it, and record, on documented stable days, what he wants done with his own money, in his own words. The objective is not abstinence, which is not a guardian's job. The objective is a person who is not robbed on the day he is weakest. A 2026 study from the same Harvard group, examining long‑term relapse among people with alcohol use disorder, found that the warning signs cluster around stress, isolation, exposure to high‑risk people and environments, and the moment when recovery slips behind other priorities. In my practice, with adults whose disorders involve other substances as well, an emptied account has been a powerful relapse trigger for reasons that look very much like those: it removes the rent, the food, and the sense that recovery has preserved anything tangible.

And in the first forty‑eight hours, do what the Flash 48h checklist does for any protected adult: cards, online access, powers of attorney, the bank alerted to outgoing transfers, one coordination lead, every contact dated. The instrument does not care how old the person is. The system should not either.

The overdose curve is bending. Seventy thousand deaths is still a catastrophe, but it is some forty thousand fewer than three years ago, and each of those forty thousand is an adult who will need something the system does not yet count: protection that assumes he will be capable again, and makes sure there is something left when he is.