A vulnerable population in research is any group of people who have a reduced ability to protect their own interests during a study, whether because they cannot freely consent, face pressure to participate, or are at heightened risk of being harmed or exploited. U.S. federal regulations single out specific groups for extra protections: children, prisoners, and pregnant women each have their own set of rules governing how and when they can be included in research. The 2018 revision of those regulations also formally recognizes individuals with impaired decision-making capacity and those who are economically or educationally disadvantaged as vulnerable.
Why These Protections Exist
Modern research ethics grew directly out of documented abuses. The Nuremberg Code, issued in 1947 after Nazi human experimentation, established that voluntary consent is absolutely essential. In the United States, two cases became turning points. The Tuskegee syphilis study, which ran for 40 years and deliberately withheld treatment from Black men with syphilis, led Congress to create the National Commission for the Protection of Human Subjects. That commission produced the Belmont Report in 1979, which laid out three core principles: respect for persons (treating people as autonomous and protecting those with diminished autonomy), beneficence (minimizing harm and maximizing benefit), and justice (distributing the burdens and benefits of research fairly). Meanwhile, from 1956 to 1971, researchers at Willowbrook State School for the Retarded deliberately infected children with hepatitis to study the disease’s progression.
These cases revealed a pattern: researchers gravitated toward people who couldn’t easily say no. Children, institutionalized people, racial minorities, and the economically desperate were enrolled in studies that offered them little benefit and exposed them to serious harm. The regulations that followed were designed to break that pattern.
Children in Research
Children cannot give legally valid consent, so the federal rules require two layers of agreement. Parents must grant permission, and the child must provide “assent,” a simpler form of agreement appropriate to their age, maturity, and psychological state. For studies that carry more than minimal risk without the prospect of directly helping the child, both parents typically need to give permission rather than just one.
The regulations sort pediatric research into four categories based on risk and benefit. At the lowest level, research must involve no more than minimal risk. The next tier allows greater-than-minimal risk only if the child may directly benefit. A third category permits higher risk with no direct benefit to the child, but only if the study is likely to produce important knowledge about the child’s specific condition. The fourth and most restrictive category covers research that doesn’t fit the other three and can only move forward through a special federal review process when it addresses a serious health problem affecting children.
The Belmont Report specifically flagged a concern about children: because they are easy to recruit, researchers risk overburdening them with study participation. The principle of justice demands that children not bear a disproportionate share of research risk simply because they are convenient subjects.
Prisoners in Research
Prisoners are considered vulnerable primarily because the controlled environment of a prison can make truly voluntary consent difficult. When your daily life is governed by institutional authority, an offer to participate in research can feel like an implicit requirement, or the compensation and medical access that come with a study can be disproportionately attractive compared to what’s otherwise available.
Federal rules limit the types of research that can involve prisoners to four narrow categories: studies of incarceration itself and its effects (at no more than minimal risk), studies of prisons as institutions or prisoners as a population (also minimal risk), research on conditions that disproportionately affect prisoners as a group (such as hepatitis or substance addiction), and research with a reasonable chance of improving the health of the individual participant. The last two categories require the federal government to consult outside experts in ethics, medicine, and penology before approving them.
Any ethics review board evaluating prisoner research must include at least one prisoner or prisoner representative among its members, and a majority of the board’s non-prisoner members must have no ties to the prison system. These composition rules exist to counterbalance the institutional interests that might otherwise dominate the review.
Pregnant Women and Fetuses
Research involving pregnant women must meet a series of conditions designed to protect both the woman and the fetus. When possible, researchers must first conduct preclinical studies and studies in non-pregnant women to assess potential risks before enrolling pregnant participants. Any risk to the fetus must either come from a procedure that could directly benefit the woman or the fetus, or, if there is no prospect of benefit, the fetal risk must be no greater than minimal and the research must aim to produce important knowledge that cannot be obtained any other way.
If a study could benefit only the fetus and not the mother, both the pregnant woman and the father must consent, with exceptions when the father is unavailable, incapacitated, or when the pregnancy resulted from rape or incest. Researchers are also prohibited from offering incentives to terminate a pregnancy and cannot play any role in decisions about the timing or method of ending a pregnancy or in determining a newborn’s viability. These firewalls separate the research from reproductive decision-making.
Notably, the 2018 revision of federal regulations removed pregnant women from the general list of populations considered inherently “vulnerable.” The shift reflects the view that pregnancy alone does not impair a person’s ability to make autonomous decisions. The specific protections under their own regulatory subpart still apply, but the framing changed to avoid unnecessarily excluding pregnant people from research that could benefit them.
Impaired Decision-Making Capacity
People with conditions that affect cognition, such as dementia, intellectual disabilities, traumatic brain injury, or acute psychiatric illness, may have difficulty understanding research procedures, weighing risks, or communicating their preferences. Federal policy recognizes them as vulnerable, but impaired capacity does not automatically disqualify someone from participating in research. Instead, it triggers additional safeguards.
The most common safeguard is involving a legally authorized representative (LAR), someone who can consent on the person’s behalf. The LAR is expected to use “substituted judgment,” meaning they should make the decision the participant would have made based on that person’s known values and preferences, not simply what the LAR thinks is best. The LAR’s consent is documented in the same formal way as if the participant were consenting directly. When possible, researchers also seek the participant’s own agreement to the extent they can provide it.
Economic and Educational Disadvantage
People living in poverty or with limited education are vulnerable not because they lack intelligence, but because their circumstances can distort the consent process. When a study offers free medical care, compensation, or access to treatments a person could not otherwise afford, the offer can become so attractive that the individual signs up regardless of the risks. Ethicists call this a “decision-impairing inducement”: the benefit is so enticing relative to the person’s baseline situation that it effectively overrides careful weighing of potential harms.
The concern is not that economically disadvantaged participants cannot understand study information. It is that what the study offers, whether money, healthcare, or other goods, holds outsized appeal when set against unjust background conditions like lack of insurance or chronic financial insecurity. This can render people, in the regulatory language, “relatively or absolutely incapable of protecting their own interests.”
Investigators and ethics review boards share the responsibility of preventing exploitation. That means distinguishing between a study that takes unfair advantage of someone’s poverty and one that offers fair compensation but happens to recruit from a disadvantaged community. Practical steps include careful attention to how study information is disclosed, ensuring that benefits to participants are proportionate to their contributions and risks, and avoiding compensation structures so large they crowd out rational decision-making.
The International Standard
Outside the U.S., the Declaration of Helsinki, maintained by the World Medical Association, provides the most widely recognized ethical framework. It defines vulnerability broadly: some individuals, groups, and communities face greater risk of being wronged or harmed because of factors that may be permanent or may shift with context. A person might be vulnerable in one setting and not another.
The Declaration adds an important counterpoint to protectionism. It warns that excluding vulnerable groups from research can perpetuate or worsen health disparities, because those groups then miss out on the medical advances that studies produce. Research involving vulnerable populations is justified when it responds to their specific health needs, when they stand to benefit from the results, and when the study cannot reasonably be conducted with a less vulnerable group. This framing pushes researchers to think carefully about both the harms of inclusion and the harms of exclusion, rather than defaulting to keeping vulnerable people out of studies entirely.