Is the current regulatory framework for direct-to-consumer microbiome-based tests sufficient to protect consumers from medical, economic, and dignitary harms?

Researchers wanting to start direct-to-consumer (DTC) microbiome-based testing 1 companies have attracted ‘hundreds of millions in venture capital funding’ and established ‘dozens of new startups.’ 2 This proliferation in commercial microbiome testing follows on the heels of DTC genetic testing 3 and reflects growing consumer interest in personalized medicine. Gut microbiome–based testing particularly appeals to this interest because the gut microbiome has been associated with numerous medical conditions that may be improved through dietary and lifestyle changes. 4 Similarly, consumers may pursue vaginal microbiome–based testing in order to detect ‘disruptive and protective bacteria/fungi that research has shown are related to’ cases of chronic vaginal infections. 5 Relatively new to the marketplace are companies marketing DTC oral 6 and skin microbiome tests. 7 Recognizing a commercial opportunity, private companies are actively entering the market with offerings of these tests, which are moderately priced, provide laboratory results without the inconvenience of scheduling a medical appointment, and empower consumers with information about their microbiome and purportedly their health. 8 The growth of the DTC microbiome-based testing market, however, raises some of the same concerns encountered with the emergence of DTC genetic tests, such as confidentiality of consumers’ health data, limited evidence that the tests can accurately inform clinical decision-making or dietary recommendations, dubious claims, 9 and self-misdiagnosis when consumers rely on inaccurate results. 10 Legislators and regulators wrestled for over a decade with how to address these concerns for genetic tests. Based on evidence of deceptive marketing and misleading test results 11 and concerns about how third-party use of test results would affect genetic research, legislative and regulatory responses included new statutes, such as the Genetic Information Nondiscrimination Act (GINA) 12 and similar state laws, as well as changes to the Food and Drug Administration’s (FDA) regulation of laboratory developed tests (LDTs). With this new product class, ie DTC at-home sample collection kits and laboratory assessment of one’s microbiota, come new questions about whether the regulatory pathway for DTC genetic tests is a good fit for DTC microbiome-based tests or whether modifications to this pathway are necessary to adequately address the risks these tests may pose to consumers. The tests also share characteristics with at-home health tests that have skirted the line of medical device regulation because they fall into the category of ‘general wellness products,’ raising concerns about their relatively unregulated status. 13 Additionally, the recent federal district court case, American Clinical Laboratory Association v. FDA , 14 vacating FDA’s attempt to regulate LDTs, raises questions about the overall adequacy of the regulation of both DTC genetic and microbiome tests. A response to these concerns by federal legislators and regulators will need to consider whether consumers are adequately protected from harms to their health, safety, and privacy, as well as from financial exploitation, and, if additional regulation is necessary, how to implement it without stifling innovation in this nascent industry. In this article, we review the rapidly growing DTC microbiome testing industry with a focus on gut and vaginal testing companies. We look at how these companies operate and the types of claims they make. 15 Next, we introduce the relevant federal regulatory framework and describe how it has evolved to govern the related DTC genetic testing industry. 16 Finally, we apply the current federal regulatory framework, including the impact of the Am. Clin. Lab. Ass’n v. FDA decision, 17 to DTC microbiome-based tests and raise questions about whether they fit into the light-handed regulatory scheme that developed around DTC genetic tests, and, if not, whether additional or different regulation is required, including regulation in the areas of test validity and reliability, data privacy, human subjects protections, data discrimination, and forensic use of microbiome test results. 18

Regulation of the DTC microbiome testing industry, like the DTC genetic testing industry, requires a regulatory framework that will address the health and safety risks associated with these products. Such risks may include consumers improperly self-medicating, avoiding care from a provider, or taking supplements instead of their physician-prescribed drugs. Furthermore, many of the companies make claims that mislead consumers into believing that there is a proven connection between their microbiome composition and their health or specific diseases or health conditions, that changing their diet or taking recommended supplements can change their microbiome, and that changing their microbiome can improve their health. Many individuals who use these tests are ill and suffer from chronic conditions that significantly affect their quality of life. When ‘traditional’ therapies for their conditions are not effective or do not lead to full resolution of their symptoms, they seek out other information and hold out hope for a treatment that will provide relief from their pain and other debilitating symptoms. Physicians have reported seeing patients who have used DTC microbiome tests, coming to them after following harmful dietary restrictions or undergoing fecal microbiota transplants without first consulting a medical expert. Consumers may also not be aware that their test results are being sold by DTC microbiome testing companies, and that there is absolutely no privacy protection of their data. As a result, they could be required to disclose test results to insurers or employers who may use the information to deny them health insurance or employment. Finally, to the extent that one’s microbiome test results can be used to identify them, law enforcement may be able to obtain such results from companies and use them for forensic purposes. Over a decade ago, the U.S. Government Accountability Office observed that DTC genetic tests were unreliable because different companies were reporting different results using the same samples. 268 Similarly, a study conducted by NIST found that DTC microbiome testing companies that received the same samples reported different results. 269 Like the early DTC genetic tests, the DTC microbiome testing companies are ‘nowhere near… being able to interpret’ 270 their test results. Both the analytical and clinical validity of the tests are highly suspect. The manufacturers of the early DTC genetic tests led consumers to believe that the test result was linked to health and disease when in many cases it was not. The DTC microbiome testing companies are using the same playbook—leading consumers to believe the tests can inform them about their health and medical conditions when there is no substantiation of the relationship between test results indicating dysbiosis and risk of disease. While the DTC microbiome testing companies may not always name a specific disease, they want the consumer to believe that the test can diagnose a disease or health condition and the related dietary change or supplement they recommend can treat, mitigate symptoms, cure, or reduce the risk of a disease or condition. To be diagnostic or predictive, DTC microbiome-based tests would have to overcome their lack of analytical and clinical validation by developing consistent and curated reference databases and software algorithms to ‘ensure accurate, replicable, and comparable results.’ 271 Some may argue that the harms associated with false-positive or false-negative GHR tests are greater than those associated with DTC microbiome tests. In the case of the GHR tests, using the example of tests for BRCA1, a false-positive test could lead a consumer to experience anxiety or depression, or to have a preventive mastectomy or oophorectomy, which are invasive and serious procedures. A false-negative result could lead to a delay in seeing a physician or scheduling a mammogram that would risk not identifying breast cancer at an early stage. Risks associated with inaccurate microbiome test results can have similar harmful effects. A false positive could lead someone with gut symptoms to experience anxiety or to change their diet or take supplements in ways that are unnecessary or harmful. And, like the case of breast cancer, a false negative could lead to a delay in seeking treatment from a physician. These harms and risks are largely a result of gaps in the regulation of the manufacture, marketing, and use of these tests. Under CLIA, the laboratories that analyze these tests are inadequately regulated because, even if they are CLIA certified, the certification process does not ensure analytical validity. A major shortcoming of the CLIA regulatory process for DTC microbiome tests is that there are no established proficiency tests for them. Moreover, although FDA attempted to promulgate regulations that would roll back its enforcement discretion policy for LDTs in 2024, those regulations were recently vacated by a federal district court. 272 This court decision undermines FDA’s ability to regulate this entire category of tests, ie LDTs, at least in the short term. However, assuming FDA remains able to regulate DTC tests, DTC microbiome testing companies appear to have assumed that these tests are low-risk devices for general wellness purposes and thus exempt from medical device regulation. Many of these tests do not satisfy the requirements for exemption from medical device regulation as specified by the American Clinical Laboratory Association v. FDA case or under the 21st Century Cures Act and the FDA’s own guidance documents—they do not meet the definition of an LDT as described by the Court and are not low risk; nor do they meet the exemptions for software designed to improve health or encourage a healthy lifestyle; and they do not meet the exemptions for transferring and displaying test results or for clinical decision-making support. To remedy these shortcomings, we advocate several regulatory reforms. First, we recommend that CMS, under CLIA, require DTC microbiome testing companies to report their testing methods and the algorithms they are using to measure the analytical validity of their tests. NIST is in the process of developing a reference standard for stool that can be used as a touchstone for the DTC microbiome testing industry. 273 The reference material could significantly advance the field of microbiome-based testing. 274 Once NIST makes available the reference standard, we advocate that as part of CLIA certification, CMS require DTC microbiome testing companies to use the standard material to evaluate the accuracy of their own procedures and as a source for comparison of consumer samples. 275 In the meantime, DTC microbiome testing companies should be required to use a ‘lockdown method’ that helps ensure internally consistent test results by utilizing an analytical algorithm and reference database that does not change until the accumulation of new data necessitates the development of a ‘version 2.0’ of these tools. Second, although we maintain that DTC tests do not meet the definition of LDTs put forth in the recent federal district court decision, we recommend that Congress pass legislation giving FDA clear authority to regulate LDTs either as medical devices or as a separate category. Third, we think FDA should treat DTC microbiome tests in the way it treated the GHR tests marketed by 23andMe. The two types of tests are similar. The risks identified by the FDA with respect to 23andMe’s personal Genome Services—risk of incorrect understanding of the device and test system, incorrect test results, and incorrect interpretation of the test results—apply equally well to DTC microbiome-based tests. As such, FDA should require DTC microbiome tests to submit an application similar to a de novo Class II medical device application for approval and to adopt general as well as several special controls. Like the requirements for 23andMe, the special controls should include information to consumers (in labeling and on websites) indicating the limits of the accuracy of the test results; that test results may be affected by external variables like low sample quantity or issues with sample collection, shipping, storage, or collection buffers; there is insufficient evidence to link the gut or vaginal microbiome to any disease or harmful condition or a predisposition to any disease or condition; the test may not enumerate all the microorganisms in the consumer’s gut or vagina and the microbiome may change over time; the detection of a pathogen does not necessarily indicate disease, and failure to detect a pathogen does not mean that the pathogen is not present; other companies offering DTC microbiome tests may follow different procedures and, thus, report different results; the test is not intended to diagnose diseases or health conditions or to be used in medical decision-making; and the test is not a substitute for a visit to a doctor or other health care professional. Additionally, the FDA required 23andMe to meet labeling requirements ‘designed to help consumers understand and interpret the test results along with a “hyperlink” to the manufacturer’s… website where the manufacturer was to make publicly available all the special control information requirements.’ This should also be required for DTC microbiome test manufacturers. Furthermore, as it did with 23and Me, FDA should require the companies to use an ‘FDA-cleared, -approved or -classified as 510(k) exempt sample collection device.’ 276 Fourth, FDA and FTC should collaborate to ensure that companies making false or unsubstantiated claims receive warning letters and, if they persist in their actions, are charged with violation of federal law. FTC may be in the better position to bring such legal action, and its prior action against GeneLink may provide a template for regulatory enforcement against several of the DTC microbiome testing companies. 277 Finally, in order to address regulatory gaps in HIPAA, GINA, and laws protecting human subjects, we advocate that the safeguards other scholars have recommended for DTC genetic testing companies, such as 23andMe, also be applied to DTC microbiome testing companies. These include requiring DTC microbiome testing companies ‘to obtain express consent to disclose [microbiome test results] with strict civil and criminal penalties if they do not,… limiting permissible disclosures [of microbiome test results] to research companies so that insurers, employers, and other third parties cannot exploit genetic information’; allowing consumers to have their genetic information ‘destroyed or deleted’; and providing consumers with ‘explicit information from DTC testing companies about how their DNA is being used,’ whether it be the specifics of the research or for marketing purposes. 278

The DTC microbiome-based testing market has been predicted to expand at a compound annual rate of 11.8 per cent from 2024 to 2032. 19 The industry has its roots in the Human Microbiome Project (HMP), a National Institutes of Health research initiative that was funded from 2007 to 2016. 20 The first phase of the HMP ‘characterized the microbial communities from 300 healthy individuals, across [five] different sites on the human body: nasal passages, oral cavity, skin, gastrointestinal tract, and urogenital tract.’ 21 The second phase addressed the function of microbial communities and their interaction with their human host. 22 Both phases yielded improved laboratory tools that helped to shed light on the relationship of the microbiome to health and disease. Leveraging associations between composition and structure of the microbiome and disease outcomes, private companies soon commercialized the new tools in the form of DTC microbiome tests. For this article, we conducted a thorough online search for DTC gut and vaginal microbiome testing companies. As of November 2024, our search yielded 30 companies operating in several countries/continents 23 : 23 in North America, 12 in Europe, three in Australia, two in India, and one in South Korea. 24 All of the DTC microbiome testing companies we reviewed use a similar business model. Most companies analyze the gut microbiome, but an increasing number are targeting the vaginal microbiome, and a few provide testing for the skin or oral microbiome. For the most part, DTC microbiome-based tests are ordered online by consumers without the involvement of a physician. 25 The tests typically consist of a kit containing a sterile swab and a plastic tube that the consumer uses to collect a small specimen (stool for gut test; vaginal secretions for vaginal test) to mail to the company for analysis. Most companies ask consumers to complete a questionnaire that includes demographics, dietary habits, health history, and, for vaginal tests, sexual practices. This information is referred to as ‘metadata.’ Tests cost between $100 and $300, 26 and results are received in 2–4 weeks. The report often includes a graphical scorecard that, at a minimum, identifies the microbes present, compares these microbes to a consumer database, and, in the case of gut microbiome analysis, makes dietary and/or nutritional supplement recommendations with the intent of improving the consumer’s gut health. Consumers pay more for tests that measure an extended panel of analytes and for consultation with a company-affiliated health care professional, usually a dietitian or nutritionist. However, several companies offer consultation with either a personal health, microbiome, or ‘vaginal coach.’ 27 Some companies also offer membership or subscriptions. 28 DTC microbiome testing companies use genetic techniques to analyze the microbiome, specifically, next-generation sequencing such as 16S rRNA gene amplicon sequencing or metagenomic sequencing. These sequencing techniques describe taxonomic composition of the microbiome by identifying which microorganisms are present in a specimen and their relative abundance. 29 This may include pathogens as well as commensal bacteria. Companies may develop their own in-house tests based on these genetic techniques. Alternatively, they may rely on external laboratories to perform the analysis. The companies/laboratories use software (often combinations of open-source and custom) to analyze the resulting data to make health and lifestyle recommendations. For the most part, DTC microbiome testing companies are not explicitly claiming that their tests can diagnose illness or inform treatment decisions, but some companies are using the data they collect from consumers to conduct research in-house, or they scour the literature and analyze the published data to learn how the gut microbiome relates to different health conditions. 30 Some companies provide this literature to customers to buttress their conclusions. Additionally, a few companies are partnering with academic research laboratories and using the academic data to clinically validate their tests. Through this research, some of these companies may aim to obtain FDA approval or clearance to ultimately use their microbiome tests to diagnose disease and/or make treatment recommendations. Most DTC microbiome testing companies, however, enter the market with semi-proven data analytics that they attempt to validate by building a robust dataset over time that relies heavily on the clinical or research information collected from their early users. 31 For example, the company might compare a consumer’s data with those in the database of patients with certain diseases 32 or to those who are deemed ‘healthy’ or ‘normal.’ 33 Companies then make recommendations based on this comparison. For example, one company provides a report to consumers with scores in six metabolic categories, an interpretation of the scores, and a recommendation based on the interpretation for personalized dietary modifications and supplements. 34 Another company provides test results and recommendations for a hypothetical consumer listing 63 vegetables the consumer should eat and five vegetables to avoid. 35 The report explains that the recommended vegetables are unlikely to cause an increase in blood sugar and may improve digestive efficiency and intestinal barrier health. Avoiding the other vegetables may improve the consumer’s ‘sulfide gas production pathways.’ 36 These companies state that they analyze the composition of an individual’s microbiota to look for dysbiosis that may suggest an ‘unhealthy’ gut or vaginal microbiome. However, their claims of having the ability to detect unhealthy microbiomes are not fully substantiated by the literature and may lead to consumer exploitation by marketing expensive products and recommending repeat testing and probiotic use, which may not have any value. 37 Until these tests can establish a verifiable link between one’s microbiome composition and a condition or disease, they cannot be used to diagnose disease and, more importantly, cannot inform clinical decision-making. 38 Additionally, DTC microbiome testing companies that use their tests to make unsubstantiated claims regarding health and wellness may put consumers at risk of harm when their products do not deliver on their claims. 39 While the commercial market is developing products that allegedly provide consumers with health and wellness insights, the academic community is focused on advancing microbiome-based tests for the actual diagnosis of disease. 40 One area of intensive research is the development of microbiome-based diagnostic tools for cancer. A 2022 review article concluded that ‘there is vast evidence indicating specific bacterial taxa and their associated genetic loci in cancer development and progression’ and that ‘many studies have suggested an association between variance in human bacterial composition and cancer.’ 41 An example is variations in the microbiome of the female genital tract, which have been linked to ‘cervical cancer, but also sexually transmitted disease and bacterial vaginosis recurrence.’ 42 Additionally, ‘imbalances in the lung microbiota’ have been associated with lung cancer. 43 Further academic research studies may pave the way for a future in which microbiome-based tests become a routine part of medical practice to (1) detect biomarkers for diagnosis, (2) analyze patient microbiota to predict treatment outcomes, (3) develop personalized treatment strategies, and (4) monitor treatment success using microbiome time-series analysis. 44 Despite ongoing academic research, most microbiome-based diagnostic tests have limited analytical 45 and clinical validity due to a lack of supporting data. 46 Determining the analytical validity of these tests is challenging because there is no standardization across the field. 47 Companies employ different pre-analytic and analytic laboratory techniques, leading to variability in test results. To establish clinical validity, each testing company must independently conduct well-powered clinical trials that demonstrate the ability of a company-specific process to diagnose a specific condition or disease.

In determining the adequacy of the current regulatory framework for DTC microbiome-based tests, we consider several parameters: (1) effectiveness/accuracy of the test, (2) safety, (3) accuracy and sufficiency of consumer information, and (4) data privacy and potential harmful use of information by third parties. The efficacy or accuracy of these tests, along with their safety, is arguably the feature of most importance for regulation. If not accurate, consumers are paying for tests that have no value, and they may take action that is harmful and based on unreliable results. For non-invasive diagnostic tests, safety is linked to accuracy as it relates to the tests’ false-positive and false-negative rates and how that information is used by the consumer or patient. Consumer information is the information provided by the company, including claims made on its website and other marketing materials about the benefits of the test and how the information will be used by the company. Finally, data privacy relates to whether the company will provide the consumer’s personal information to third parties and the potential harms that could come to the consumer from outside parties having access to their data. In this section, we address how each of these factors are currently regulated for DTC health tests and how well those regulations apply to DTC microbiome tests. The Centers for Medicare and Medicaid Services (CMS) and FDA both potentially have a role in regulating the accuracy and safety of DTC microbiome tests. Since DTC microbiome testing companies handle human specimens and run laboratories that analyze the specimens, these microbiome testing companies, like other DTC health tests, are subject to the Clinical Laboratory Improvement Amendments of 1988 (CLIA) overseen by CMS. 87 To the extent that DTC health tests are considered diagnostic, they would be subject to FDA regulations governing ‘medical devices’ 88 and/or to FDA guidance on ‘software as a medical device’ 89 unless exempt. Under the software guidance they may be exempt from medical device regulation as low-risk tests for general wellness and/or as tests for clinical decision-making support. Additionally, the 2025 Am. Clin. Lab. Ass’n v. FDA case, discussed infra , has interjected some uncertainty into FDA’s future regulation of DTC health and microbiome tests under medical device regulations. 90 CMS enforces CLIA to ensure quality testing in US laboratories. 91 CLIA certification is mandatory for all laboratories that test ‘materials derived from the human body for the purpose of providing information for the diagnosis, prevention, or treatment of any disease or impairment of, or the assessment of the health [emphasis added] of, human beings.’ 92 Under CLIA, LDTs 93 that have not been cleared or approved by FDA must be analytically validated in the laboratory’s own environment before results using the test can be released. 94 Furthermore, CLIA regulations do not distinguish between facilities performing DTC testing and facilities providing testing that is ordered by a clinician. 95 CLIA requires that a covered laboratory verify the analytical validity, 96 but not the clinical validity, of its tests. 97 CMS issues different types of certification to laboratories based on the level of complexity of their tests, ie high, moderate, or low. 98 Any test that is moderately or highly complex must meet CLIA standards for quality control and assessment, and personnel, and the laboratory performing the test likely needs to enroll in a proficiency testing program. 99 The methodological requirements are to ensure consistency across laboratories, while proficiency testing ensures that an individual laboratory’s results are consistent with an external reference standard. The CLIA certification process can be time consuming and highly particularized. To achieve certification, laboratories must document and implement auditable tracking for all aspects of their operation and submit to an inspection that occurs at random and is repeated every 2 years. 100 Furthermore, high-complexity tests have more stringent personnel requirements under CLIA, while simple tests with a low risk of producing an incorrect result may obtain a waiver from CLIA requirements. 101 While CLIA requirements appear robust, there are shortcomings in the oversight process. For example, the analytical validation may not be verified by CMS until after the laboratory has already started marketing its test. 102 Although some of the DTC microbiome testing companies we identified for this article provide disease information, many provide only health and general wellness advice. However, this still brings them within the umbrella of CLIA, as CLIA certification is required for all laboratories that provide information used to assess the ‘health of, human beings.’ 103 Furthermore, DTC microbiome-based tests would likely be considered high complexity by FDA because they employ genetic analysis (considered high-complexity tests), 104 and a CLIA certificate of waiver does not apply to such tests. 105 Eleven of the 23 DTC microbiome testing companies we reviewed that sell their product in the USA noted on their websites that they are CLIA certified or contract with CLIA-certified laboratories. It is unclear from our research whether the remaining US companies attempted or achieved CLIA certification. If a DTC microbiome testing company has the resources, CLIA certification can confer a competitive advantage by indicating to consumers that the company’s tests meet the federal standards for commonly used laboratory procedures and techniques. 106 However, we have reason to doubt that CLIA certification of DTC microbiome-based tests demonstrates their analytical validity. The CLIA standards for analytical validity do not explicitly address microbiome tests, 107 and there are currently no CLIA-approved proficiency testing programs for these tests, likely due to the difficulty of confirming their analytical validity. Unlike other diagnostic tests that target certain biomarkers, as discussed supra , 108 it is unclear whether microbiome-based tests can accurately assess what they purport to measure. Fundamentally, microbiome tests represent a paradigm shift in the way laboratory ‘testing’ has been performed. This is the case for a few reasons. First, the tests focus on the relative abundance of microorganisms in a sample, rather than detecting a specific gene, bacteria, fungi, or mutation. Second, in traditional diagnostic testing, the target is clearly known in advance, eg Staphylococcus aureus , and standards can be developed to validate the test. In the case of microbiome-based tests, the test must ‘discover’ the composition of the sample, making the tests much more difficult to validate as we do not know ex ante what the sample contains. While CMS jurisdiction under CLIA does not extend beyond assessing analytical validity, FDA evaluates both the analytical and clinical validity of products but only if they are subject to the Agency’s premarket submission process. Of potential relevance to DTC microbiome-based tests, FDA oversees the regulation of (1) medical devices, (2) software as a medical device (SaMD), and (3) medical device marketing claims and misbranding. Diagnostic tests potentially fall into the regulated category of medical devices. The Food, Drug, and Cosmetic Act (FD&C Act), section 201(h) defines a medical device in part as ‘an instrument, apparatus, implement, [or] machine,… which is… intended for use in [among other things] the diagnosis of disease or other conditions.’ 109 FDA regulates medical devices using a three-tiered classification system based on the intended use of the product and its level of risk to consumers. The higher the risk, the higher the classification and the more regulatory control FDA exerts over the product. For Class I devices, the lowest risk classification, 110 only general regulatory controls apply. These are provisions that control for adulteration, misbranding, labeling, and good manufacturing processes. 111 For Class II devices, general and special regulatory controls apply. 112 For Class III, the highest risk classification, general controls apply and a premarket approval (PMA) application for the product containing data from non-clinical laboratory studies and clinical investigations must be submitted to FDA. 113 Of the three medical device classes, only Class II and III products require a submission to FDA prior to marketing. 114 Diagnostic tests in which specimens are taken from the body for external testing are referred to as in vitro diagnostic tests (IVDs). 115 Like other diagnostic tests, IVD test safety relates to the likelihood of a false-positive or false-negative test result and the impact on the consumer/patient of such inaccurate results. 116 A test that produces many false negatives may cause a consumer’s serious or treatable disease to be missed, whereas a test that produces many false positives may cause a consumer needless worry and costly follow-up testing or treatment for a disease that they do not have. IVD tests can include those based on next-generation sequencing 117 and are classified using the same three-tiered risk-based system as other medical devices. Many IVDs fall into the Class II category because they may produce incorrect test results, incorrect interpretation of test results, or incorrect understanding of the test system. 118 Class II devices require a premarket notification to FDA under section 510(k) of the FD&C Act. 119 In a 510(k) submission, a company must demonstrate that its product is substantially equivalent to a previously FDA-approved product already on the market, called the ‘predicate.’ Historically, the predicate device was to relate back to a device marketed prior to 1976 when the medical device amendments were passed. Today, as the quality and designs of medical products have changed over time, the predicate device is the most recent 510(k) medical device approved by FDA for that particular product. Alternatively, a company may request that a device with no substantial equivalent be classified as a Class I or II device using a de novo application. 120 The de novo application includes an explanation of why the device would be safe and effective if general and special controls are applied, as well as clinical and non-clinical data demonstrating safety and effectiveness. 121 If a company can successfully demonstrate either equivalency or safety and effectiveness with general/special controls, the device is ‘cleared’ or approved for marketing by FDA. 122 Class III devices, in contrast, require a more comprehensive PMA submission to FDA wherein a company must prove that its product is both safe and effective based on clinical trials. 123 LDTs, a type of IVD, have traditionally not been reviewed by FDA prior to marketing as a matter of policy. 124 Although FDA consistently claimed to have the authority to regulate LDTs as medical devices if they were prescribed by a health care provider, it generally did not do so, instead applying its policy of enforcement discretion. This policy, however, was fraught with controversy in light of changes in the type of LDTs that became available over time and the possibility of harms based on false-negative or false-positive test results. 125 As a result, in April 2024 FDA announced in a rulemaking that it would regulate LDTs as medical devices. 126 This decision was subject to widespread criticism within the laboratory testing community and, not long afterwards, the decision was challenged by the American Clinical Laboratory Association (ACLA) and the Association for Molecular Pathology. On March 31, 2025, the U.S. District Court for the Eastern District of Texas vacated the Agency rule in its entirety, concluding that it exceeded the Agency’s statutory authority. The Court emphasized that ‘“devices” under the FDCA “are articles of commerce, not the kinds of services performed by doctors and laboratories,”’ and that Congress specifically addressed the regulation of clinical laboratories and their services through CLIA, an independent statutory framework. 127 The FDA did not appeal the decision and on September 19, 2025, the FDA released a final rule rescinding its 2024 regulation that had classified LDTs as medical devices subject to FDA oversight. This leaves the regulation of LDTs by FDA unlikely in the near term unless the Agency interprets the decision in a way that allows it to continue to regulate some LDTs, such as DTC tests, or Congress steps in and provides explicit authorization for FDA to regulate them. As to the former, although many believe FDA pushback to the Court’s ruling is unlikely during the current Trump administration, 128 there is at least one argument FDA could make in issuing revised regulations or guidance or taking enforcement action. The Agency could argue that, at a minimum, it continues to have the authority to regulate DTC tests as the definition of LDTs in the Court’s opinion is quite narrow. The Court defines the tests as ‘in-house diagnostic tests developed, validated, and performed by trained professionals within a single clinical laboratory. They are performed on specimens at the request of an individual physician, in the context of a specific doctor-patient relationship.’ 129 This definition clearly does not include DTC tests ordered by consumers over the internet without the involvement of a physician. Some critics have argued that FDA could also attempt to use the 21st Century Cures Act to argue that Congress assumed when it passed the Act that FDA had the authority to regulate tests that rely on software as medical devices, even if they were LDTs. That assumption is why Congress needed to pass the Cures Act, carving out certain exceptions to the existing law at the time. These exceptions included tests using software for general wellness and for some other types of software-based tests. Using this rationale, FDA could regulate software-based LDTs as medical devices under revised regulations; however, such action would be highly unlikely at this time. 130 As to congressional action, a number of commentators think this is unlikely any time soon. Congress considered such legislation in the recent past. In 2018, members of Congress introduced the VALID Act, which would have explicitly given FDA the authority to regulate LDTs using a risk-based framework similar to the one it applies for medical devices. 131 The legislation did not pass, however, and was reintroduced in subsequent years, without success, despite the support of several medical device advocacy organizations. Of potential relevance, in its Final Rule for the regulation of LDTs, consistent with the District Court’s definition of LDTs, FDA stated that its prior general enforcement discretion policy for LDTs did not apply to DTC tests, 132 which it believes require a heightened level of oversight especially when they are ‘used by consumers to make potentially significant healthcare decisions without the involvement of a learned intermediary in a legitimate health care practitioner-patient relationship.’ 133 As a result, FDA categorized DTC tests into multiple subcategories with different regulatory pathways, with varying degrees of regulation. 134 The subcategories include genetic carrier screening tests, genetic health risk tests, pharmacogenetic tests, cancer predisposition tests, low-risk general wellness tests, and ancestry tests. 135 Although the issue of how LDTs will be regulated remains uncertain in the longer term, if FDA is able to successfully issue regulations for DTC tests or software-based tests, of relevance for the regulatory pathway of DTC microbiome tests would be whether the test is used for general wellness purposes and is low risk. If so, assuming FDA is able to regulate DTC tests, under its current framework it would likely not require premarket review of the DTC microbiome tests as it views them as low-risk general wellness tests. 136 This policy stems from the Agency’s Industry Guidance initially drafted in 2015, reissued in 2016 137 and finalized in 2019, 138 as well as the 21st Century Cures Act. 139 FDA does not define ‘low risk tests’ but provides some guidance as to what does not constitute a ‘low risk’ general wellness product. For example, products that are invasive or implanted would not be low risk; nor would products that involve an ‘intervention or technology that may pose a risk to the safety of consumers and other persons if specific regulatory controls are not applied.’ 140 The Agency also provides a number of examples of products that would and would not be considered low risk. 141 None of the examples in FDA’s Industry Guidance documents for ‘low risk’ tests include ‘health tests’ that might have false positives or false negatives that could lead to consumer harm. For DTC tests intended for moderate- or high-risk medical purposes, FDA has, at least up until now, reviewed the tests for analytical validity, clinical validity, and the company’s claims about its product. 142 Additionally, FDA has considered whether the instructions for collecting the specimen and interpreting the test results can be understood by the consumer without involvement of a physician. 143 If FDA determined that a DTC test met each of these criteria, it would ‘clear’ the test for marketing. The earliest DTC genetic tests cleared by FDA were 23andMe’s carrier screening test for Bloom Syndrome in February 2015 144 and its genetic health risk (GHR) tests 145 in April 2017. 146 FDA granted 23andMe’s requests for marketing the carrier test based on a 510(k)submission and, for its GHR tests, based on a de novo Class II medical device submission. The approvals came after a contentious history during which FDA issued warning letters in 2010 to 23andMe and other DTC genetic testing companies stating that their tests were medical devices and would be regulated as such. Despite the warning letters, 23andMe continued to market its tests. In 2013, FDA ordered 23andMe to stop marketing its DTC tests because the company had not established the tests’ analytical or clinical validity. 147 For the GHR tests, the company was required to adopt general as well as numerous special controls. These special controls included specific information to consumers in labeling and test reports about the limits of the test results, specifically that other companies offering a genetic health risk test for the same disease may use different genetic variants and thus get different results; and that other factors, such as the environment or lifestyle risks, may play a role in whether the consumer develops the disease. FDA also required that certain information be provided to health care providers, such as a statement that the test is not intended to diagnose a disease and that any diagnostic decisions made should be based on other diagnostic testing and/or other information that the provider deems appropriate. Additionally, the company was required to ‘use a sample collection device that was FDA-cleared, -approved, or -classified as 510(k) exempt, with an indication for in vitro diagnostic use in over-the-counter DNA testing.’ 148 Finally, the FDA mandated extensive labeling requirements designed to help consumers understand and interpret the test results along with a hyperlink to the manufacturer’s website where the manufacturer was to make available all the special control information requirements. 149 Subsequent to its approval of 23andMe’s GHR tests, FDA cleared several pharmaco-genetic tests and a cancer risk test developed by 23andMe. 150 As of October 2024, 23andMe was the only company to have had its DTC tests cleared for marketing by FDA. 151 Despite FDA asserting its authority over certain health-related genetic tests, many DTC genetic and microbiome testing companies likely assumed they were low-risk general wellness tests and, thus, not regulated by FDA as medical devices. Although FDA had issued its draft Industry Guidance for low-risk wellness products in January 2015 and a subsequent draft in July 2016, Congress passed the 21st Century Cures Act in December 2016, 152 exempting certain DTC tests using software from medical device regulation. 153 The legislation listed five bases for exemption. Among the exemptions was that the software be ‘for maintaining or encouraging a healthy lifestyle and unrelated to the diagnosis, cure, mitigation, prevention, or treatment of a disease or condition.’ 154 The subsequent Industry Guidance by FDA 155 clarified that the Agency did not intend to regulate software-based tests that are intended only for ‘general wellness’ as medical devices. 156 The Agency defined a general wellness product as one that is low risk and has an intended use that relates to maintaining or encouraging a general state of health or a healthy activity, or… that relates the role of healthy lifestyle with helping to reduce the risk or impact of certain chronic diseases or conditions and where it is well understood and accepted that healthy lifestyle choices may play an important role in health outcomes for the disease or condition. 157 This definition breaks down products that meet the general wellness category into those that do not make any reference to a disease or condition and those that do. 158 Included in the prior group are ‘claims to promote or maintain a healthy weight, encourage healthy eating, or assist with weight loss goals.’ 159 Not included in this group are claims ‘to restore a structure or function impaired due to a disease or condition.’ 160 The second group, which makes reference to diseases or conditions, is further divided into those ‘intended… to promote, track, and/or encourage choice(s), which, as part of a healthy lifestyle, may help to reduce the risk of certain chronic diseases or conditions… [or] help living well with certain chronic diseases or conditions.’ 161 Claims made by this second group must be based on references ‘where it is well understood that healthy lifestyle choices may reduce the risk or impact of a chronic disease or medical condition.’ 162 By ‘well understood,’ the association between the healthy lifestyle choices and health outcomes must be generally accepted in the scientific community. 163 While the 21st Century Cures Act likely bolstered the view of software-based DTC microbiome testing companies that they were exempt from medical device regulation, this assumption may have been wrong. DTC microbiome-based tests may not meet the requirements for ‘general wellness.’ Claims made by many of the companies go beyond general wellness claims. As previously stated, claims that a product may ‘restore a structure or function impaired due to a disease or condition’ are not exempt general wellness claims. 164 Many of the claims made by DTC microbiome testing companies link the tests to the ability to restore a structure or function of the gut or vaginal microbiome, eg the structure of the microbiome is impliedly and in some cases explicitly impaired (in ‘dysbiosis’) due to a disease or condition. While some companies reference specific diseases, such as IBS or bacterial vaginosis, others imply that such dysbiosis is due to an abnormal condition. These claims are being made even though there is no substantial evidence that microbiome test results can indicate these conditions or disorders and, even if they could, such associations are not ‘well understood’ or accepted in the scientific community. Moreover, there is no acceptance in the scientific community that choices to take certain supplements or eat certain foods will resolve dysbiosis, improve health, or prevent disease. A final argument that these tests do not meet the exemption for general wellness tests is that they appear to be marketed to individuals with a health problem or condition, not to healthy individuals. If DTC microbiome-based tests were determined to be medical devices (ie not considered low risk or general wellness tests and not exempt as LDTs), the question would remain whether they would be exempt from device regulation under the other 21st Century Cures Act software exemptions. 165 In addition to exemptions for software products that encourage or promote a healthy lifestyle, the 21st Century Cures Act exempts from medical device regulation four software functions. 166 Of these, the most relevant for DTC microbiome-based tests is likely software that is intended for transferring, storing, or displaying clinical laboratory test or other device data and results, findings by a health care professional with respect to such data and results, general information about such findings, and general background information about such laboratory test or other device unless such function is intended to interpret or analyze clinical laboratory test or other device data, results, and findings. 167 How this provision applies to DTC microbiome-based testing requires an understanding of the different software functions associated with these tests. Most testing companies likely employ several software functions in the testing process. These may include (1) analysis of the composition of the sample, (2) a comparison of the sample composition with the ‘average’ sample composition from company data and/or with ‘healthy’ or ‘unhealthy’ populations, (3) flagging areas of concern (those that fall outside the ‘norm’ or are characteristic of unhealthy individuals), (4) displaying results, and (5) making recommendations based on test outcomes. As such, the tests would likely be considered multiple function device products. 168 Arguably, functions 1 (analyzing the composition of the sample) and 5 (making recommendations) would involve the interpretation or analysis of device data and would be ‘device functions’ subject to premarket review. The other three functions more likely would be considered ‘other functions.’ 169 Under the Cures Act 170 and FDA Industry Guidance, ‘FDA may assess the impact of “other functions” when assessing the safety and effectiveness of the device functions-under-review of a multiple function device product.’ 171 FDA Guidance further provides that the Agency will ask two questions when assessing a product with multiple functions: ‘(A) Is there an impact on the safety or effectiveness of the device function-under-review as a result of the “other function”?; and, if there is an impact, (B) Could the impact result in increased risk or have an adverse effect on performance of the device function-under-review, i.e., negative impact?’ 172 Based on the Guidance document, FDA would likely evaluate these functions holistically rather than evaluate each one separately, as the recommendation function is dependent for its accuracy on both the device functions (numbers 1 and 5) and the ‘other functions,’ 173 ie if the analysis of the sample is incorrect or the data regarding the ‘normal’ or ‘healthy’ person is inaccurate, that will affect the recommendations. Thus, under the provisions for transferring, storing, converting, or displaying data, these functions would likely not be exempt because at least the first and fifth functions fall under the transfer provision exception, ie the software ‘interpret[s] or analyze[s] clinical laboratory test or other device data, results, and findings.’ 174 Furthermore, even if the software is not considered a ‘clinical’ laboratory test, as required under the transferring provisions, because it is not aimed at physicians, if it is considered a device, it would be interpreting or analyzing ‘other device data, results and findings,’ 175 and would still not be exempt from regulation. The other exemption under the 21st Century Cures Act that is of possible relevance is referred to as the clinical decision support (CDS) software exemption. 176 It exempts software that is for (i) displaying, analyzing, or printing medical information about a patient or other medical information…; (ii) supporting or providing recommendations to a health care professional about prevention, diagnosis, or treatment of a disease or condition; and (iii) enabling such health care professional to independently review the basis for such recommendations that such software presents so that it is not the intent that such health care professional relies primarily on any of such recommendations to make a clinical diagnosis or treatment decision regarding an individual patient. 177 The software must meet all three tests to be exempt. These provisions limit FDA’s regulatory authority because clinical decision-making constitutes the practice of medicine and is governed largely by state medical practice regulations and tort law. 178 It is likely that these provisions also reflect the assumption that, with a health care provider acting as an intermediary, consumers will be less likely to misinterpret a test result. It is questionable whether these provisions would apply to DTC microbiome-based tests as they are marketed as tests for consumers and, for the most part, bypass the need for physician involvement. While this is generally the case, some physicians (eg functional medicine physicians) will order these tests for their patients and use them as part of their diagnostic process. However, they use them along with an array of other tests to build a diagnostic picture when the presenting symptoms are not easy to interpret. 179 Also, as indicated supra , a few microbiome testing companies market directly to physicians. 180 In these cases, the tests might meet elements (i), (ii), and (iii) of the exemption requirement. Furthermore, the CDS exemption could apply if it covers dietitians or nutritionists as the exemption applies to ‘health care professionals,’ not just physicians. In its Industry Guidance, FDA states that it uses the term health care professional to mean an individual who is licensed, registered, or certified by a State, territory, or other governing body, to administer health care, including but not limited to, nurse practitioner, registered nurse, licensed practical nurse, clinical social worker, dentist, occupational therapist, pharmacist, physical therapist, physician, physician assistant, psychologist, respiratory therapist, speech-language pathologist, technologist, or any other practitioner or allied health professional. 181 The broad inclusionary language of ‘any other practitioner or allied health professional’ would seem to encompass these two professionals. Approximately half the states require licensure of dietitians and almost a dozen require licensure of nutritionists. 182 Many DTC microbiome testing companies employ nutritionists or dietitians to consult with consumers about their test results and make recommendations regarding diet or dietary supplements allegedly based on those test results. Each state that licenses these practitioners has its own definition of the practice of dietetics and nutrition. An example from one state provides that the practice of dietetics includes, among other things, ‘applying scientific research to the role of food in the maintenance of health and treatment of disease.’ 183 Some DTC microbiome-based test software could meet the requirement that it is ‘supporting or providing recommendations to a health care professional about prevention, diagnosis, or treatment of a disease or condition,’ a requirement for CDS software. 184 However, in the context of a physician, dietitian, or nutritionist, it would also have to ‘enable such health care professional to independently review the basis for such recommendations that such software presents so that it is not the intent that such health care professional rely primarily on any of such recommendations to make a clinical diagnosis or treatment decision regarding an individual patient.’ 185 Thus, it would depend on how much the provider is relying on the software recommendation versus providing their own assessment of the test results. 186 To the extent that the tests do not meet the requirements for the other exemptions, ie are not considered general wellness tests, etc. and are being used by health care professionals, the CDS software exemption provisions would come into play. However, there is also an exception to the software exemption that may be relevant, ie if ‘the software function is intended to acquire, process or analyze a medical image or a signal from an in vitro diagnostic device or a pattern or signal from a signal acquisition system’ then it will be regulated as a medical device 187 if it meets the medical device definition. 188 The term pattern ‘refers to multiple, sequential or repeated measurements of a signal or from a signal acquisition system,’ such as next-generation sequencing (NGS). 189 Specifically, ‘software functions that process or analyze the genetic sequence or patterns from an NGS analyzer to identify genetic variants or mutations or their clinical implications or relevance do not meet’ 190 the exemption criteria and would be regulated as a medical device. While clear this provision applies to DTC genetic testing software, there does not appear to be a reason why it would not also apply to DTC microbiome tests in which NGS is used to identify the microorganisms in a sample. Thus, even if DTC microbiome-based testing software met requirements (i)–(iii) for the CDS exemption, it likely falls under the exception to the exemption because it is intended to ‘acquire, process, or analyze a medical image or a signal from an in vitro diagnostic device or a pattern or signal from a signal acquisition system’ and would not be exempt from the medical device regulations. 191 For tests that market directly to consumers, the most likely basis for exemption from FDA medical device regulation would be that the tests are low-risk general wellness tests. 192 However, this is not a certainty and would depend on the claims companies are making and the potential harms that result from their recommendations. If consulted, FDA would consider whether such tests are marketed as non-medical, general wellness products before determining whether they would be exempt from premarket review. 193 If the Agency were to determine that the tests are for general wellness, in addition to not exercising oversight of their analytical or clinical validity, FDA would not regulate the company’s marketing claims if those claims fit within the parameters described above, ie (1) they are ‘health claims’ and do not make any reference to diseases or conditions, or (2) they focus on a healthy lifestyle and make reference to diseases or conditions only if it is ‘well understood and accepted’ that a healthy lifestyle choice can reduce the risk of the chronic disease or condition that the product references. 194 If FDA were to determine that the manufacturer’s intent is to market a product that diagnoses, treats, or reduces the risk of a disease or condition, despite the manufacturer’s claim that it is a general wellness product, FDA could decide that the product is misbranded and should be treated as a medical device. 195 Under the FD&C Act, all medical devices, regardless of risk classification, are subject to general controls that relate to misbranding. 196 Misbranding includes labeling that is false or misleading; does not state the name of the manufacturer, quantity, or adequate directions for use; or describes an intended use that is not approved by FDA. 197 FDA also is tasked with ensuring that products are not misbranded by using false or misleading advertising. 198 Although FDA jurisdiction over advertising is limited to restricted medical devices, 199 it has jurisdiction over promotional labeling of both restricted and unrestricted medical devices. 200 Some examples of promotional labeling are product brochures, print ads, and internet websites. 201 Although, by regulation, promotional labeling refers to drug products, FDA has applied the concept to medical devices. 202 FDA also assesses the risk and benefit claims for medical devices that are not low-risk general wellness products to ensure that the risk–benefit profile of a given product is balanced; if not, the overall promotional message (ie not the individual claim) is considered to be false or misleading. 203 FDA has actively regulated some DTC genetic test marketing claims. For example, on April 4, 2019, FDA sent Inova Genomics Laboratory a warning letter for not obtaining marketing clearance or approval for its MediMap genetic tests. 204 The Agency flagged Inova for its website claim that ‘the MediMap tests [are] genetic tests for predicting medication response’ and determined that the product is a diagnostic medical device for which clinical validity had not been established. 205 FDA was concerned that providing test results directly to patients ‘could lead to patients inappropriately increasing, decreasing, or stopping their medication without their physician’s involvement, which poses significant risks to patient safety.’ 206 FDA also clarified that the MediMap LDT was not exempt from premarket approval because the agency retains its discretion to regulate LDTs that pose a risk to public health. 207 In addition to the FD&C Act, the Federal Trade Commission (FTC) Act protects consumers from unfair, deceptive, and fraudulent business practices. 208 Under Section 5 of the Act, advertising claims are deceptive if they mislead consumers in their conduct or decisions with respect to a product. Product advertising includes print and broadcast materials and internet advertisements. 209 For unrestricted medical devices, FTC considers whether claims are ‘truthful, non-misleading and substantiated by “competent and reliable scientific evidence.”’ 210 In practice, FTC is less concerned about the truth or falsity of a claim than whether the claim is substantiated by reasonable supporting evidence. 211 Furthermore, the substantiation must be material in that a consumer would be less likely to rely on the company’s claim if the company ‘did not have a reasonable basis for believing [the claim] to be true.’ 212 FTC, like FDA, has enforced its prohibition of false and misleading claims against genetic test manufacturers. In 2014, FTC brought an enforcement action against GeneLink, Inc. and a former subsidiary company because they claimed that their customized nutritional supplements based on the DNA analysis of a consumer-supplied cheek swab could treat diabetes, heart disease, arthritis, insomnia, and other ailments. 213 GeneLink settled with FTC by agreeing not to make disease or health claims related to ‘genetic disadvantages’ unless ‘the claim is true and based on competent and reliable scientific evidence,’ ie supported by ‘at least two adequate and well-controlled studies.’ 214 Many of the DTC microbiome testing companies we reviewed make general wellness claims, including claims about how the tests’ results can help consumers improve their nutrition, weight, mood, energy, and/or fitness. 215 These claims would likely exempt the tests from medical device regulation. However, numerous companies make healthy lifestyle claims about (1) identifying and reducing disease risks for intestinal complaints or vaginal issues, or (2) improving living with a chronic disease. 216 These latter claims arguably do not meet the criteria for exemption because claims that the tests can identify disease risks or the supplements the companies sell can improve the microbiome and ultimately reduce risks of disease are claims that are not generally accepted by the scientific community. For example, the following statements made on the promotional materials or websites of DTC microbiome testing companies may fall short of meeting the requirements for healthy lifestyle claims: [Our] personalized probiotics support GI motility, SIBO, leaky gut, and overall digestive wellness. [Our test results] can reveal important information about the root cause of many common gastrointestinal symptoms and non-GI conditions including: Indigestion/reflux Abdominal pain/cramps Diarrhea Constipation Inflammatory bowel disease (IBD) Irritable bowel syndrome (IBS) Atopic dermatitis/eczema Allergies Autoimmune diseases Mood disorders (depression) Joint aches Diabetes Weight issues This test is useful for women who have or are at risk of pregnancy and fertility issues, bacterial vaginosis, candida, pelvic inflammatory disorder and other vaginal problems. [Our test can help] Improve fertility and prenatal health to support full-term pregnancy Reduce the risk of cervical cancer and endometriosis Prevent recurring urinary tract infections (UTIs), bacterial vaginosis (BV), and aerobic vaginitis Decrease risks of sexually transmitted infections (STIs) and human papillomavirus infection (HPV) 217 Additionally, even if the company does not mention a specific disease or condition, if a consumer understands the claim to be referring to a disease or condition, FDA may consider it a more than low-risk product. 218 Also of possible relevance, under regulations for dietary supplements, a claim that a product ‘has an effect on an abnormal condition associated with a natural state or process, if the abnormal condition is uncommon or can cause significant or permanent harm,’ would be a disease claim. 219 While a claim regarding a probiotic may be permissible under the Dietary Supplement Health and Education Act of 1994 (DSHEA) 220 as a structure/function claim, 221 the Act does not apply to claims about medical devices. However, in its Guidance about low-risk general health and wellness devices, FDA has stated that structure function claims that do not refer to a disease or condition would fall under the low-risk general health and wellness exemption, whereas claims ‘to restore a structure or function impaired due to a disease or condition’ would fall outside the general wellness exemption. 222 If the claims are not considered to fall under the exemption requirements, FDA could, among other things, issue a warning letter to the manufacturer that it is operating in violation of the FD&C Act. As stated above, the Agency will also consider whether a consumer would think the claim refers to a disease and, if so, will categorize the claim as a disease claim. 223 While safety, accuracy, and truthfulness are perhaps the main concerns of regulators, consumers have also become concerned about privacy, especially as it relates to their health information, and how their data may be used by companies with whom they do business. In the context of genetic data collection, consumers have alleged that companies are using the data they collect for research without adequate consent, and that law enforcement may have access to genetic databases for forensic purposes. When FDA approved 23andMe’s tests for predicting the genetic risk of disease, the Agency did not address the potential for invasion of privacy in the case of a data breach or company use of the genetic data for purposes unapproved by the consumer. A significant part of 23andMe’s business model, however, was the collection of genetic data from consumers to sell to third parties including researchers, insurers, and pharmaceutical companies. 224 Anticipating a lucrative return on investment from these sales, 23andMe subsidized the cost of its genetic services to consumers and, thereby, collected even more data to sell to third parties. 225 We assume DTC microbiome testing companies are also selling consumer data to third parties. 226 In the USA, there is limited law protecting collection of patient data by commercial entities. In the European Union (EU), in contrast, protection of consumer data is governed by the General Data Protection Regulation (GDPR), which establishes how personal data (including genetic data) may be processed and stored by companies doing business in the EU. 227 In the USA, however, there is no overarching federal data protection law. 228 Rather, the US digital marketplace is addressed by bifurcated law 229 that covers governmental entities or private entities 230 and sectoral laws that apply to entities based on the type of information they collect and use. 231 An example of a sectoral law is the Health Insurance Portability and Accountability Act (HIPAA), 232 which applies only to health data collected by covered health care entities. HIPAA protects the privacy and security of a patient’s individually identifiable information (eg name, address, birth date, and Social Security number) and applies to health care industry entities and certain of their business associates. 233 HIPAA also applies to sensitive health information such as genetic data, but it does not apply to deidentified genetic data that is collected by DTC genetic testing companies not affiliated with the health care industry. 234 Since DTC microbiome testing companies are not health care providers, HIPAA does not apply to them. 235 This creates a regulatory gap that allows companies to avoid federal oversight of genetic data privacy. In fact, if they perform metagenomic sequencing, they generate human genetic information (especially in vaginal or buccal samples). That information could be useful to the company, and its use is not regulated. DTC genetic testing companies exploit this regulatory gap through the use of privacy policies and terms of service (ToS) agreements that enumerate how a company collects and uses consumers’ genetic data. 236 As a condition of using a company’s services, consumers must consent to the contents of the privacy policy and ToS, but the complexity (ie unintelligible legalese), ambiguity, and lack of standardization make this form of consent largely ineffective. 237 Furthermore, many companies ‘reserve the right to unilaterally change their ToS, often without users’ consent.’ 238 After a consumer consents to the terms, a company may post changes directly to the ToS and advise the consumer to check the terms periodically to become aware of them. If a consumer continues to use the company’s services after the ToS have changed, the consumer has impliedly accepted the change(s). Although such ‘bait and switch’ tactics that unilaterally make material changes to the ToS without notice to consumers may constitute a cause for legal action, most consumers do not have the resources to assert a legal challenge and are faced with either accepting the new terms or discontinuing their use of the company’s services. 239 In practice, most consumers consent without reading the privacy policy or ToS and, therefore, do not understand the potential uses of their data to which they are asked to consent. 240 Furthermore, notwithstanding FDA’s approval of GHR tests as a type of medical device, DTC genetic tests are considered commercial transactions and not medical tests for the purposes of patient informed consent. 241 And, although most states have some form of statute protecting the privacy of genetic data, 242 the scope of these laws vary widely, leaving the USA with no comprehensive privacy protection for genetic data. Other laws, however, may provide consumers with some protection. In the case of GeneLink, in addition to bringing false claims charges regarding its customized nutritional supplements, FTC charged the company with ‘deceptively and unfairly’ stating that it had taken ‘reasonable and appropriate security measures to safeguard and maintain personal information collected from nearly 30,000 consumers.’ 243 FTC asserted that the company failed to protect consumers’ personal information including their genetic information. 244 Thus, under the FTC Act, not only can companies not make false claims about their products, they also may not make false claims about their privacy policies. DTC microbiome testing companies, like their genetic testing counterparts, use privacy policies and ToS to obtain consent for the collection and use of consumer data. Consumers impliedly consent to the collection and use of their data when they click through the privacy policy and ToS to register to use the company’s services. Each of the US DTC microbiome testing companies we reviewed clearly state that failure to consent to their privacy policy or ToS results in reduced (or zero) use of the company’s services. The 23 privacy policies we reviewed were generally easy to understand and most explained in detail how they collect, use, and share consumer data. Several policies also claimed to adhere to privacy laws such as GDPR and state laws that apply to consumers in those jurisdictions. Just like genetic testing companies, DTC microbiome testing companies that do not analyze samples supplied by a physician’s office are not covered entities under HIPAA and, even if they were, they are not bound by HIPAA requirements if they anonymize consumers’ sensitive (ie health) data before it is used or shared. When DTC microbiome testing companies use privacy policies to obtain consent to collect consumer data, they also rely on these policies to inform consumers about how their data might be used. For example, their data might be used in secondary research, eg attempting to find a link between microbiome composition and a specific disease. The Federal Policy for the Protection of Human Subjects (ie Common Rule) establishes an overarching framework for the regulation of federally funded human subjects research or research conducted at an academic research institution that receives federal funds. 245 Although the Common Rule does not cover privately funded research conducted by non-academic entities, clinical trials conducted as part of premarket approval for medical devices are subject to FDA regulations governing the protection of human subjects. These provisions, which are similar to the Common Rule guidelines, have the effect of extending the Common Rule to private entities seeking FDA premarket approval for a drug or device. 246 Since, however, the DTC microbiome-based tests described herein are not being regulated as medical devices, such rules would not apply to secondary research conducted with the data obtained from these tests. Although not legally required for these tests, informed consent ‘is essential to the conduct of ethical research.’ 247 The requirement for informed consent was crafted at a time when human subjects research was conducted within the confines of randomized controlled trials that utilized individually identifiable information. 248 In the case of DTC genetic tests and microbiome tests, however, data is collected when the consumer consents to use the service and can be anonymized and aggregated for ‘undefined future research purposes.’ 249 Most of the 23 US DTC microbiome testing companies we reviewed for this article obtain consent through their privacy policies to use consumer data for research or other unspecified commercial and non-commercial purposes, which could include secondary research. Half of the reviewed companies using consumer data to conduct secondary research indicate on their websites that they aggregate or otherwise deidentify the data before it is used. With the absence of federal privacy protections for genetic data comes the interrelated risk of discriminatory use of this data. Unfettered access to consumers’ genetic data gives employers and health insurers an unfair advantage in making employment and insurance coverage decisions. As a result, Congress passed the Genetic Information Nondiscrimination Act (GINA) in 2008. 250 Proponents of the bill argued that federal protections against genetic discrimination were needed to advance personalized medicine by making consumers willing to participate in genetic research and comfortable with having a genetic test for clinical purposes. 251 Critics of the bill argued that the legislation would not resolve state law inconsistencies and would increase frivolous litigation, but, to date, GINA has not been heavily litigated. 252 GINA prohibits health insurers from requesting or requiring a prospective or currently insured person from undergoing a genetic test or using their genetic information to make eligibility determinations or decisions regarding coverage, underwriting, or premium-setting. 253 Additionally, GINA prevents employers from using a person’s genetic information in hiring, firing, promotion, pay, or job assignment decisions, and employers cannot request, require, or purchase genetic information about an employee. 254 Forty-nine states also have statutory restrictions on the use of genetic information or genetic testing by health insurers, and 35 states extend these restrictions to employers. 255 Twenty-four states also limit the use of genetic information by life, disability, and long-term care insurers. 256 The application of GINA to DTC microbiome-based tests requires that these tests fit within the definition of genetic information or genetic tests. Under GINA, genetic information includes genetic test results. A genetic test result is defined as ‘an analysis of human DNA, RNA, chromosomes, proteins, or metabolites, if the analysis detects genotypes, mutations, or chromosomal changes.’ 257 If an employer requests an employee’s microbiome test results, the application of GINA would be based, first, on whether the test results are considered ‘human’ DNA. Whether microbiome DNA is human DNA has not been definitively decided by any court, government agency, or scientific body. In fact, there is some debate within the regulatory and scientific community about whether microbiome material obtained from inside or on the human body is human since it is made up of non-human microorganisms. 258 However, while microbiome samples are primarily made up of microorganisms, they may also include some human tissue. 259 The second issue is whether, as a practical matter, microbiome tests could be used by an employer or insurer to discriminate against the employee or insured. For example, failure to regulate these tests as SaMD could result in the industry developing algorithms with inherent biases or software used to identify individuals exposed to certain drugs or toxins or living in certain neighborhoods. 260 Given these regulatory gaps, consumers may be at risk of insurance or employment discrimination. Consumers of DTC microbiome test results may also be at risk of other kinds of discrimination or stigma. Many testing companies collect personal information about a consumer’s lifestyle that may include diet, drug use, and sexual practices. Women with specific vaginal microbiome profiles may be labeled at risk for bacterial vaginosis, fertility problems, or premature births. Such information, if inappropriately shared, could target them for drug promotions or advertisements that are embarrassing or stigmatizing. While laws against discrimination may not address this type of stigma, laws aimed at prohibitions on companies selling this information to other companies without consumer consent may be a way to curtail such practices. The use of genealogy databases such as 23andMe, Ancestry.com , and GEDmatch by law enforcement for criminal investigations 261 raises the question of whether a law enforcement agency could obtain a consumer’s microbiome test results for forensic purposes. Recent research suggests that each individual has a personal microbiome that can be used to identify them. 262 This so-called ‘microbial cloud’ 263 may one day be used to reveal aspects of, and predict, a person’s habits and lifestyle. 264 For example, ‘neighborhood built environment characteristics, such as increased interaction among home occupants and indoor ventilation, may help determine the presence of microbiome species’ 265 that could identify an individual from their microbiota and possibly place them at a specific location. What is deeply troubling to many in the genealogy community is that this area is wholly unregulated. Health law scholars who have written about this issue in the context of genetic information argue that it implicates Fourth Amendment privacy rights. 266 The Fourth Amendment of the U.S. Constitution guarantees ‘the right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures….’ 267 Under this constitutional right, the U.S. Supreme Court identified several individual privacy rights that the government is obligated to protect. If microbiome-based information could provide law enforcement with meaningful evidence of identity in the commission of a crime that could not otherwise be obtained, it would raise Fourth Amendment privacy concerns. To the extent that an individual can be identified by their microbiome, use of that information by law enforcement could put an individual at the scene of a crime. Additionally, an individual’s gut microbiome analysis could indicate that they used illegal drugs or that they visited a foreign country where the USA prohibits its citizens from travel.

In sum, the DTC microbiome testing industry is growing both domestically and internationally. Consumers are paying hundreds of dollars to know if they have a healthy gut or vaginal microbiome. In some cases, these consumers are healthy and simply curious about their microbiome, but, in others, consumers have chronic illnesses, such as IBS or bacterial vaginosis, and are looking for information that may help them address their conditions. Consumers are relying on microbiome-based test results to purchase dietary supplements, including probiotics and prebiotics, and/or are changing their diet based on company recommendations. Yet, there are no standards for confirming the analytical or clinical validity of these tests, and company marketing claims border on disease claims. Yet, DTC microbiome testing companies are largely being ignored by the federal agencies that have the authority to regulate them, either because they lack the resources or they view them as exempt from medical device regulation under existing law or as a low priority. We reviewed the regulatory frameworks that may apply to the current DTC microbiome testing industry and identified a number of potential regulatory gaps. While we conclude that the current regulations for DTC genetic tests likely apply to DTC microbiome-based tests, more effort needs to go into regulating the tests’ analytical validity by CLIA and regulating the tests’ analytical and clinical validity by FDA. The latter will require reassertion by FDA that it has the authority to, at a minimum, regulate DTC tests or Congressional action to clarify the law. Regulators should also reassess the decision that these tests are truly ‘low risk.’ Actual risks of the tests include providing consumers with recommendations that may be harmful to their health, or, at a minimum, are ineffective and a waste of money. Furthermore, it is unlikely that a number of the tests meet the definition of general wellness tests that have been exempt from medical device regulation. Finally, there are significant regulatory gaps in data privacy, human subject protections, and the use of consumer data for discriminatory or forensic purposes. The DTC microbiome testing industry has the potential to improve consumers’ lives, but these companies are not doing the research necessary to analytically and clinically validate their tests. Thus, there is a need for regulation that considers consumers’ safety from physical harms, unsubstantiated claims and false information, as well as economic and dignitary harms.

DTC microbiome tests appear to attract two types of consumers: (1) healthy consumers who are curious about their microbiome and (2) chronically ill consumers who order these tests to address their medical conditions. 48 As part of an NIH-funded research grant, we conducted focus groups with representatives of four stakeholders: microbiome researchers, clinicians, patients who might purchase these tests, and consumers who had ordered them. The purposes of the focus groups were to 1) identify the … motivations and perceptions of patients currently engaged with microbiome-based diagnostic tests; 2) determine under what circumstances potential consumers would or would not order such tests; 3) identify the … practice variables and perceptions of providers ordering these tests for their patients or concerns they have about ordering such tests; 4) determine how well patients and providers understand the test results; and 5) elucidate potential legal or social implications of these [tests] that should inform their regulation. 49 Participating patients included individuals from a national medical center who had been diagnosed with various gastrointestinal conditions and who had not ordered the tests but might be interested in such tests. Consumers included individuals who had taken the tests offered by the American Gut Project (now called the Microsetta Initiative), an open platform for citizen science microbiome research. 50 When asked about the potential benefits of these tests, one healthy participant stated that she thought ‘the benefits would include a better understanding of which microorganisms you are lacking and which you may have an overabundance of.’ Another stated, ‘This particular type of test could be good for a baseline, [and subsequent tests could then reveal] how those numbers (high, low, and normal) have fluctuated over time.’ In describing why she got the test, a participant with a chronic illness stated, ‘I had been dealing with [an illness] for about 10 plus years, but I just reached my breaking point where I decided to go all in … and figure it out once and for all. Ultimately, it led to a fecal transplant…. So yeah, [this test] was like just a stop on the journey.’ Focus group responses such as these suggest that consumers may have unrealistic expectations of the tests and a misunderstanding (likely fueled by company claims) of what these tests can reliably tell them. Focus groups with specialty physicians who were most likely to be familiar with these tests confirmed the conclusion we drew from the consumer/patient focus group results that microbiome tests are likely to be misunderstood. 51 These specialists included adult and pediatric gastroenterologists and gynecologists (hereinafter traditional providers), and functional medicine physicians. 52 Functional medicine physicians were the most likely providers to have ordered these tests for their patients or suggested that their patients order the tests. 53 In contrast, traditional providers did not order these tests, viewing them as difficult to interpret and not providing meaningful results. Many traditional providers were concerned about these companies’ analytical procedures and their ability to detect variants in the microbiome, as well as whether the testing process was standardized across the DTC testing companies. Providers from both the traditional and functional medicine groups were concerned that these tests may provide recommendations to address a specific microorganism that may not have clinical relevance. Additionally, they asserted that addressing specific microorganisms by recommending supplements or elimination diets may be harmful and disrupt the complex, symbiotic nature of the microbiome, leading to worse health outcomes. They said this was of particular concern for DTC tests, where patients may implement these interventions without consulting a physician. In addition to potentially affecting health outcomes, some providers from both groups (ie traditional and functional medicine) said that focusing on specific microorganisms may delay the detection and treatment of the root cause of the health issue. 54 A few traditional providers viewed these tests as ‘baseless products [intended] to take advantage of patients.’ 55 They further expressed concerns about the costs of the tests and associated ‘treatments,’ including probiotics, herbs, and dietary changes. Several also said that their patients come to them with their results after ordering the tests online for themselves or their children and ask for help with interpretation. A few pediatric gastroenterologists stated that it was parents of children with various gastrointestinal (GI) symptoms and no definitive cause that often order these tests, seeking answers to their children’s complex health issues. These physicians also shared that parents of children who have not shown improvement through conventional treatments, or who are on the autism spectrum, have ordered these tests. One pediatric gastroenterologist said: It often is families that have had challenges in finding what they see as therapeutic relationships or diagnostic answers to their questions…. In addition to patients with irritable bowel, we see patients who have functional syndromes [including] functional abdominal pain, and also a number who have children who are on the autism spectrum…. Rates of GI diseases and GI symptoms, in particular, are higher in our autism population. Often these families really struggle with having children who have significant behavioral needs. And then on top of it [have] GI symptoms that range from belly pain to aggression to constipation to diarrhea. They'll often come in with these tests looking for answers. 56 Gastroenterologists also expressed concern about some of the dietary recommendations the test results included. Pediatric gastroenterologists were particularly worried that, because of the test recommendations, their patients may restrict their diet too severely, leading to greater health issues. One said she has seen patients develop avoidant restrictive food intake disorder after following microbiome or food intolerance testing recommendations to avoid certain foods. 57 Several gynecologists expressed concern about the accuracy of these tests, in particular results about microorganisms associated with sexually transmitted infections (STIs), given that very little information is provided about how these tests detect STIs. These providers also discussed the potential harms to patients of receiving false-negative results for STIs, as well as the difficulties interpreting the findings for their patients. Additionally, the gynecologists expressed concerns about false-negative test results for human papilloma virus (HPV) as patients who are actually HPV positive may think they are fine and not seek attention from their medical provider. 58 During the focus groups, providers were given a set of “mock” test results modeled after several DTC microbiome test results available on the internet. They were asked to answer a set of questions based on their own experience reviewing these tests or on the mock test results. Some noted that, while typical laboratory results cite to ranges for what is considered ‘average,’ the mock results did not provide this information and the reports that did include ranges, did not state how ranges are determined and on which populations they are normed. This lack of clarity regarding ranges and associated algorithms contributed to their lack of trust in the results. One pediatric gastroenterologist stated, ‘I worry about how these tests were performed.’ An adult gastroenterologist said, ‘I would suggest there be some kind of standardization in the process of doing these tests, so that… if different labs are doing it, they’re using the same process and… results are comparable.” 59 Several providers, particularly the gastroenterologists, suggested that the companies that manufacture these tests should provide data on how they determine analytical and clinical validity. The lack of this data also made it difficult for them to have confidence in the validity of the results. As one pediatric gastroenterologist stated, ‘I rarely find [the results] to be useful because… firstly, you don’t get much information about how they perform the test to see how valid they are.’ Another gastroenterologist explained the need to assess the predictive ability of these tests, stating, ‘they’d have to validate it, they’d have to do this in 1,000 patients and see if they can confirm that something was predictive.’ 60 Several participants across provider groups stated that our current understanding of the microbiome has not advanced to the point of understanding the complex criteria for a healthy microbiome, what is needed to prevent dysbiosis, or the relationship between specific bacteria and health conditions. Often, microbiome-based test results will focus on the presence of a few specific bacteria. With the current understanding of the microbiome, some providers said it is difficult to comprehend the impact of these bacteria in the context of the larger microbiome. Participants explained that, due to this limited understanding, it is not possible to use microbiome-based test results to guide treatment options. 61 Providers across all specialty groups believed that there should be greater regulation of DTC microbiome tests. Several suggested regulating the use of these test results as patient health information, regulating test recommendations, and requiring standardization and transparency of test validation processes. 62 The comments by providers in the focus groups cast a shadow on the legitimacy of DTC microbiome tests and some of the practices of these companies. For one early-to-market DTC microbiome testing company, uBiome, these concerns were warranted. uBiome, a market leader, was identified by law enforcement as engaging in insurance fraud and other illegal activities. The company, founded in 2012 by Jessica Richman, a computer scientist, and Zac Apte, a biophysicist, was the first to bring to market tests based on an alleged link between human health and the microbiome. 63 uBiome marketed two clinical tests (ie SmartGut and SmartJane (an analysis of the vaginal microbiome)) that were available through a consumer’s health care provider 64 and were covered by some health insurers. Additionally, the company marketed a microbiome test for the DTC market called Gut Explorer. Richman and Apte obtained over $100 million in venture capital funds for their start-up, and, within 6 years, uBiome grew in value to $600 million. In 2018, uBiome made Fast Company’s list of the 10 most innovative companies in the world 65 and, as a key player in the life sciences commercial market, ‘basically invented the category of the microbiome.’ 66 However, to facilitate the company’s rapid growth, uBiome engaged in several illegal activities including pressuring consumers through repetitive emails to submit multiple samples for testing and then billing the consumer’s private health insurer or Medicare ‘two or three times for the same set of tests.’ 67 Further, uBiome pressured its contract physicians to approve all test requests without establishing a doctor–patient relationship or determining that a given test was medically necessary. 68 Usually, insurance companies flag vendors that repeatedly bill more than $500 without a clear medical reason, 69 and health insurers began to deny uBiome’s billing claims. In April 2019, the FBI raided uBiome’s San Francisco offices and seized employee computers in an investigation of the company’s Medicare billing. 70 Soon thereafter, Richman and Apte, as well as uBiome’s general counsel and its interim CEO, stepped down and uBiome filed for bankruptcy. When the company failed to secure a new buyer and lost its laboratory certification, uBiome ceased all operations and liquidated its business. 71 By December 2019, uBiome had sold its testing patents and data, including personal health information that participants had given them, at auction to another company for less than 1 per cent of the company’s original value. 72 On March 18, 2021, the Securities and Exchange Commission (SEC) charged Richman and Apte with ‘defrauding investors out of $60 million by falsely portraying uBiome as a successful start-up with a proven business model and strong prospects for future growth.’ 73 According to an SEC press release, the co-founders of uBiome ‘portrayed’ the company ‘as having a strong track record of receiving health insurance reimbursement for its clinical tests, which purportedly could detect microorganisms and assist in diagnosing disease.’ 74 The SEC, in its complaint, argued that these were false and misleading claims as the company’s income was dependent on ‘duping doctors into ordering unnecessary tests’ and, if the insurers had realized this, they would not have reimbursed the company for the tests. In addition to these charges, the U.S. Attorney’s Office for the Northern District of California filed criminal charges against Richman and Apte for their activities. 75 While uBiome’s legal trouble stemmed from its fraudulent billing to Medicare and commercial insurers, the statement by the SEC that the tests were medically unnecessary speaks to the limited benefit of these tests. Although we believe very few, if any, DTC microbiome companies are now billing health insurers to cover the costs of their tests, we think there is reason to be concerned about the relative lack of regulation of these tests as compared to many other medical devices. Health and wellness tests can harm consumers when the tests are not valid, companies make inaccurate or unsubstantiated claims about what the test results can tell consumers, and companies make recommendations for changes in behavior or diet that are not proven. As to the first of these, DTC microbiome tests currently on the market lack both analytical and clinical validity. 76 Microbiome tests differ from genetic tests where the goal is to find a single biomarker, eg a gene mutation. 77 In the microbiome testing context, the objective is to identify (1) all the microorganisms in a sample and (2) the amount or relative abundance of each. The prior is like finding a needle in a haystack, whereas the latter is analogous to identifying each strand of straw in the stack, categorizing each by color or size, and indicating the percent of each category in the stack. At this point in time, such reporting is not possible as scientists have not yet identified all the microorganisms that exist in the gut or vaginal microbiome as ‘the custom bioinformatic tools used do not identify or quantify all bacteria in a sample.’ 78 Even if all microorganisms were known, the tests would still likely lack analytical validity, or at least a way to discern it, as there is no reference standard by which to compare the findings of each laboratory’s assay. Without such a standard, we cannot know if organisms are missing or over-represented. Currently, most companies compare their findings of the composition of a consumer’s sample to others in their consumer database, which is not likely to be representative of the general population. Such databases are limited. 79 To supplement their in-house data, companies may ‘purchase patient data from researchers who have characterized the microbiomes of individuals with certain chronic conditions. In either case, the company is relying on nonrepresentative data generated with unknown computational techniques.’ 80 To establish accuracy, companies must be able to replicate their results from the same sample. A recent study by researchers at the National Institute of Standards and Technology (NIST) found that, for a number of DTC microbiome testing companies, this was not the case. 81 The researchers sent three separate samples from the same fecal composite sample to each of seven different testing companies. The results indicated not only a lack of consistency across companies, but also within the same company. 82 As to clinical validity, while human diseases are increasingly ‘linked with an altered or “dysbiotic” gut microbiota,… whether such changes are causal, consequential, or bystanders to disease is, for the most part, unresolved.’ 83 In addition, to tell a consumer that their microbiome is healthy or unhealthy requires comparison to a standardized reference of a healthy microbiome. Currently, there is lack of agreement among scientific experts as to what constitutes a healthy or unhealthy microbiome. 84 Despite the shortcomings of the current state of the science, many of the DTC microbiome testing companies that we reviewed make claims on their website that state or imply that the tests can tell a consumer whether their microbiome is causing their current symptoms or is related to a disease, health condition, ‘healthy gut,’ or healthy microbiome. While the companies also state that their tests are not diagnostic, their claims may lead consumers to think otherwise. Consumers are likely to incorrectly believe that these tests are regulated by the FDA when, in fact, ‘they [have until recently] straddled the line between regulated medical devices (if making diagnostic claims) and virtually unregulated health and wellness products’. 85 As we noted in a prior publication, the lack of regulation of these tests may put consumers at risk of harm when they rely on inaccurate test results and clinically unproven nutritional or food supplement recommendations. These harms may include self-misdiagnosis, delay in seeking medical treatment, and substituting nonmedical supplements for prescription medications. Many individuals who seek out these tests have serious chronic illnesses and are desperate to try anything that may mitigate their pain and suffering. 86 Due to the lack of regulation of these tests, they are not subject to adverse event reporting. Thus, there is no centralized database of reports of cases where consumers may have been harmed by following the recommendations included with their test results.