The Federal Regulations at 45 CFR 46 Provide Additional Protections for Subjects Who Are Workers/Employees to Protect Them From Potential Risks of Harm: What the Rule Actually Requires

The additional protections are not a single clause. They sit in Subpart B (pregnant women, human fetuses, neonates), Subpart C (prisoners) and Subpart D (children), each of which addresses a specific vulnerability that employment can create or compound, plus the Common Rule's general requirements for informed consent and IRB review at 45 CFR 46.109 and 46.116. Workers are not named as a protected class anywhere in the regulation. The protections reach them because a worker who is pregnant, incarcerated, or a minor falls inside one of those subparts — and because the general consent rules forbid the "you'll be fired if you don't enroll" dynamic that makes workplace research coercive.

That distinction matters more than it first appears. A manager who assumes 45 CFR 46 contains an "employee protections" section will not find one. A manager who understands how the subparts and general requirements interact will know exactly which IRB questions to expect, which recruitment tactics to abandon, and where the real legal exposure sits.

Where the Rule Actually Puts Employee Protections: Subparts, Not a Standalone Section

Subpart A, the Common Rule, was revised with an effective date of January 21, 2019, with a compliance date of January 20, 2020 for most provisions. It governs all federally funded human subjects research and covers consent, IRB review, and continuing review. It says nothing specific about employment.

Subpart B covers research involving pregnant women, human fetuses, and neonates, and it requires both the mother's informed consent and the father's where applicable, with a separate consent process for the fetus where research holds out no prospect of direct benefit. A pregnant employee recruited through her employer lands here, and the consent must make clear that participation or refusal has no bearing on her job.

Subpart C governs prisoners. Its language on "prisoners" extends to anyone involuntarily confined or detained in a penal institution, and the subpart imposes a strict condition: the research must fall into one of four permissible categories, and the IRB must find that risks are commensurate with risks accepted by non-prisoner volunteers. Subpart D covers children, including wards of the state, and limits research to categories tied to risk level and prospect of direct benefit.

The practical upshot: an employee enrolled in a workplace study is protected as a research subject under Subpart A and, if they belong to one of these populations, under the relevant subpart. State and local employment law, plus contract and tort doctrine, supply the rest.

The Pregnant Worker Scenario: Subpart B Consent and the Employer's Access Problem

Subpart B is where workplace research touches the most employees. A study recruiting pregnant workers for a workplace exposure assessment must secure informed consent under 46.116 and, where the research involves the fetus, address the father's consent under the subpart's terms. The mother's consent is never waived.

The consent form has to disclose the reasonably foreseeable risks, the procedures, alternatives, and a statement that participation is voluntary and refusal involves no penalty. At a workplace, "no penalty" carries unusual weight. The IRB will want to see how the study team separates the person who recruits from the person who supervises.

Employers often propose recruiting through HR or a plant manager. That arrangement places a supervisor in the position of asking a subordinate to sign a consent document. Even a well-meaning supervisor creates pressure. The fix is structural, not rhetorical: recruit through an independent third party, collect consent off-site, and never route signed forms through the management chain.

There's an edge case worth naming. A pregnant worker in a research study who is also subject to an employer's own occupational health surveillance program may receive two sets of forms with different rights attached. The research consent covers the study; the surveillance program may be a condition of employment. If the two are combined, the IRB should reject the combined form — a research consent cannot be bundled with an employment condition.

Prison Labor and Incarcerated Workers: Subpart C's Four Categories

Incarcerated people who work in prison industries, facility maintenance, or agricultural programs are workers inside an institution that is also their confinement. Subpart C applies to them as prisoners first. Under 45 CFR 46.306, research may involve prisoners only if it falls into one of four categories: studies of the possible causes, effects, and processes of incarceration and of criminal behavior; studies of prisons as institutional structures or of prisoners as incarcerated persons; research on conditions particularly affecting prisoners as a class; and research on practices with the intent and reasonable probability of improving the health or well-being of the subject.

The IRB must also find that any possible advantages from participation are not of a magnitude that would impair the prisoner's ability to weigh risks, against the limited alternatives available in a prison setting. That last finding is where "additional protections" becomes concrete. A small payment that would be trivial to a free worker can be coercive to someone earning cents per hour under a prison pay scale.

Consent is also affected. The subpart requires assurances that parole boards will not consider participation when making decisions, and that participants understand the limits of confidentiality in a correctional setting. A worker who fears that declining a study will be noted in a file is not consenting freely, and the IRB is required to catch it.

Check your own institution's policy: many IRBs apply Subpart C to any research involving people in custody, not just research conducted inside a prison. If your workforce includes people on work release or in halfway houses, confirm the status with counsel before recruitment begins.

Children and Minors in Workplace Studies: Subpart D's Risk Tiers

Subpart D covers children, and it applies to workplace research whenever the participants are under 18 — agricultural work, family businesses, apprenticeships, and retail. The subpart sets risk tiers that track whether the research offers direct benefit.

Under 45 CFR 46.404, research not involving greater than minimal risk requires the permission of one parent or guardian and the child's assent. Under 46.405, research involving greater than minimal risk but presenting the prospect of direct benefit requires both parents' permission when the IRB finds the risk is justified. Under 46.406, research involving greater than minimal risk with no prospect of direct benefit may proceed only if the risk is a minor increase over minimal, the intervention is commensurate with the child's experience, and both parents consent. Under 46.407, anything else requires review by a panel convened by the federal department head.

A minor employee working in agriculture is covered by child labor rules separately. The Fair Labor Standards Act's agricultural provisions under 29 CFR Part 570 set minimum ages and hazardous occupation restrictions, and those hours can overlap with research observation. The IRB's job is the research; the employer's job is the wage and hour compliance. When the two intersect, the consent process should say plainly that declining the study has no effect on the minor's employment or schedule.

Assent is not a formality. A 16-year-old working a summer job can decline research even with a parent's permission, and the IRB will expect the study team to record the child's own willingness in language the child can read.

Coercion and Undue Influence in Workplace Recruitment: Applying 46.116

The general consent requirements at 45 CFR 46.116 are where most workplace research fails on paper. Consent must be sought under circumstances that provide the prospective subject sufficient opportunity to discuss and consider whether to participate, and that minimize the possibility of coercion or undue influence.

At a workplace, three recruitment designs reliably draw IRB scrutiny. First, distributing a sign-up sheet that passes through supervisors. Second, scheduling research appointments during paid work hours without a private alternative. Third, offering incentives that flow through payroll, so the payment becomes indistinguishable from a bonus.

An IRB reviewer will typically ask: who asks, where, when, and with what consequence for refusal? If the answer to "who asks" is a direct supervisor, the protocol usually comes back. If the incentive is large relative to the worker's usual pay, the reviewer will ask for the rationale. A $25 gift card for a 30-minute survey is ordinary; a $500 payment for the same survey, offered only to warehouse staff, is the kind of fact pattern that gets flagged as undue influence.

There's a subtler point. Voluntary participation under 46.116 includes the right to withdraw at any time without penalty. If a worker's shift assignment changes because they dropped out, the study team has a problem that consent language cannot fix. Design the study so withdrawal is invisible to management — a study ID, no names on the roster, and results aggregated.

None of this requires the employer to be adversarial. Most of it requires only that the company hand recruitment to someone who is not in the worker's chain of command.

The pattern is straightforward to describe and slow to fix. A study already underway that uses supervisor recruitment may need protocol amendment, not just a memo.

Practical Compliance: The Six Questions an IRB Will Ask About a Workplace Study

Before a protocol reaches the board, the study team can predict the review. These six questions cover nearly every workplace concern the regulations touch.

1: Who recruits and who consents? Name the individual. If it is anyone in the participant's supervisory chain, redesign. 2: Where does consent take place? A private location off the work floor, with no manager present. 3: What happens to a worker who declines or withdraws? Document in the protocol that no schedule, pay, or evaluation consequence follows. 4: Are participants members of a Subpart B, C, or D population? Check pregnancy status where relevant to the research question and never ask it in a way that forces disclosure to the employer; check age; check custody status. 5: How is payment structured? Payment must be a separate transaction from wages. 6: How is confidentiality protected against employer access? Data should be coded before it reaches anyone with a management role.

Failure handling, if any of these answers is unsatisfactory: pause recruitment and submit an amendment with the redesigned process. The IRB can also require re-consent for participants already enrolled, and many boards will. If the study is already funded, the amended protocol usually proceeds faster than a new submission.

A useful benchmark: many IRBs will accept a protocol that addresses all six questions in the first two pages. A protocol that skips them tends to cost two or three review cycles.

Making Consent Meaningful When the Employer Holds the Data

Even a well-designed consent has a confidentiality weakness in workplace research. The employer often controls the site, the schedule, and sometimes the records. If participants reasonably fear that their data will reach management, the consent is not meaningful regardless of the signature on the form.

The practical fix is a data firewall. Study staff should be employees of the research institution, not the company. Any identifiers should be held by the institution's research office, not by company HR. Reports back to the employer should be aggregated at a level that prevents re-identification, and the protocol should state that level in advance.

Many companies find this arrangement uncomfortable. They funded the study and want the data. The IRB's role is to say, in effect, that the funding relationship does not confer access to identifiable subject data. Some institutions will allow de-identified summaries only. Check your institution's policy, because practice varies.

There's a related point on record retention. The Common Rule sets requirements for how long records are kept, and an employer who stores research records may not follow the same schedule. If the records will be held by the company, the protocol should state which schedule applies and who is accountable.

If the employer cannot agree to a firewall, the honest answer is that the study should not proceed with employee participants. Recruiting the same population outside the company is slower but cleaner.

State Law and Local Policy: What Federal Rules Leave Open

The Common Rule is a floor, not a ceiling. State laws on genetic testing, HIV status, and worker health databases can impose additional requirements on workplace research, and some states restrict employer access to employee medical information outright. A study that is compliant federally can still be unlawful locally.

Several states require separate written consent for genetic testing and prohibit employer retaliation tied to genetic information. Others regulate how workplace health records are stored and shared. If your study involves biological samples, review the state statute before drafting the consent, not after.

Employers subject to collective bargaining agreements may also face contractual consent requirements. A union-represented workforce often has a negotiated process for studies that touch working conditions, and an IRB protocol that ignores it creates an industrial relations problem even where the research consent is sound.

The advice in this paragraph is the article's scope boundary: this is general educational information, not legal advice. For a specific study, consult counsel licensed in the state where the research will take place, and consult your IRB's own policy on Subpart C custody determinations. Research involving prisoners or pregnant workers deserves a second opinion from the institution's human research protections office before recruitment opens.

The narrow point that federal regulation actually fixes is this: 45 CFR 46 protects workers through the subparts and the consent requirements, not through a dedicated employee provision, and the employer's role is the variable the IRB scrutinizes most closely.

FAQ

Does 45 CFR 46 contain a specific section on employee protections?

No. The regulation has no standalone employee protections section, and the protection reaches workers only through Subpart A's consent rules and Subparts B, C, and D. A search for "employee" in the regulatory text will surface the general requirements, not a dedicated provision. Study teams that write protocols as if such a section existed tend to misdescribe the IRB's actual concerns.

Can my employer require me to participate in a research study?

Under 45 CFR 46.116, consent must be voluntary and free of coercion or undue influence, and a condition of employment cannot serve as consent to research. An employer may require attendance at a workplace program that is not research, but a study that produces generalizable knowledge requires consent that the worker can refuse. Refusal should carry no employment consequence, and the IRB will ask how that is ensured.

What happens if a participant is both an employee and a prisoner?

Subpart C applies first, and the research must fit one of the four permissible categories under 45 CFR 46.306. The IRB must also find that any advantages of participation do not impair the person's ability to weigh risks, which is a harder test in custody because pay and conditions are constrained. Consent must include assurances that parole decisions and disciplinary records will not reflect participation.

Is the father's consent required for research involving a pregnant employee?

Under Subpart B, the mother's consent is always required for research involving her and the fetus, and the father's consent is required where the research involves the fetus and the father is not unknown or unavailable. The subpart also allows the IRB to waive the father's consent in specified circumstances. The protocol should state which category applies rather than leaving the question to the consent form.

How large can a participation incentive be before it becomes undue influence?

There is no dollar threshold in the regulation, and the IRB makes a study-specific judgment. A useful test is whether the payment is large enough relative to the participant's usual income that a reasonable person would feel unable to refuse. For a worker earning an hourly wage, a payment equal to a day's pay for a short survey is the kind of figure that draws scrutiny.

Does the revised Common Rule change any of this?

The 2018 revisions took effect January 21, 2019, with most compliance required by January 20, 2020. The revisions changed the consent form requirements, added a new category of exempt research, and adjusted continuing review rules, but they did not add a worker-specific provision. The subparts remain the primary source of population-specific protections.

What if the employer wants to see individual research results?

The protocol should specify a data firewall, with identifiers held by the research institution and only aggregated results returned to the employer. If the employer insists on individual-level access, the study should be redesigned or relocated outside the company, because the consent cannot honestly promise confidentiality the employer can override.

This article is general educational information about federal research regulations, not a legal opinion or a compliance determination for any specific study. Requirements vary by institution, state, and funding agency, so confirm the governing rules with your IRB and a qualified local attorney before enrolling participants. If you are a worker facing pressure to participate in research, contact your institution's human research protections office or a local employment attorney; for immediate danger, use local emergency services.