File 4
“One in six Americans” is a CDC estimate from data mostly of 2000 to 2008
- Data, mostly
- 2000 to 2008
- Population
- 2006
- In the Register
- 2011 to 2026
The number is CDC’s estimate of the total of domestically acquired foodborne illness: 47.8 million a year, rounded to 48 million and put as 1 in 6 Americans.
It comes from two papers in Emerging Infectious Diseases, posted on December 15, 2010 and dated January 2011. About 80 percent of it, 38.4 million, is unspecified agents, and CDC’s press release said so on the first day. CDC says its 1999 and 2011 estimates “were not wrong; they were the best estimates available when the data were collected.” The same answer says new data and methods “have contributed to more accurate estimates for 2019.”
The rest of the summary
The papers say their data were mostly “from 2000 through 2007” and “from 2000–2008”, and every estimate uses the 2006 population of 299 million. On the day of release CDC’s division lead said 2011 is the journal’s date and the data spanned years “centering around 2006.” The newest input found ends in 2008, which is at least 17.8 years before this file’s date (SOVRINT arithmetic). CDC’s later estimates, on a CDC page by December 2024 and in a paper of March 2025, cover six pathogens for illnesses and seven for hospitalizations and deaths. They give 9.9 million, do not estimate unspecified agents and give no new total of that scope. CDC says the two sets cannot be meaningfully compared. CDC says it is working on estimates for another set of known pathogens, and that the estimates rely on certain data sources that are not updated regularly. On the first day it said such estimates cannot be made year by year.
A person drafting a federal or state text has no later CDC total to cite that includes unspecified agents, and no CDC text was found saying which figure to use. The written search found 42 Federal Register documents from May 5, 2011 to June 2, 2026. All use the present tense, none gives a data year or the 2006 population, and 4 print the year 2011. The next notice in the last collection to carry the sentence, on October 14, 2026, has no burden sentence, and no reason is stated. CDC’s FY 2027 justification and FDA’s 2026 Food Code print the number, and CDC’s Facts page printed it in the latest capture, of October 16, 2026. This file puts the age of the inputs beside the sentence.
- Status
- Draft
- Dated
- November 6, 2026
- Retrieved
- November 6, 2026
- Responsible
- SOVRINT Inc.
- Corrections
- None
- The claim
- Unverified
AI agents gathered the sources, read them, counted them and drafted this file, using scripts. No person has yet checked it or read a source. No SOVRINT software produced it.
Full title: “One in six Americans” is a CDC estimate from data mostly of 2000 to 2008 and the 2006 population, no later CDC total of that scope was found, and federal texts printed it in 2026
What this file does not say
The claim
One in six Americans, or about 48 million people, gets sick from foodborne disease each year.
Unverified for any year after the data window. Observed as CDC’s estimate on a 2006 base.
- This file does not say the number is false, wrong or superseded. True and false are left to fact-checkers. CDC’s own page says: “CDC's 1999 and 2011 estimates were not wrong; they were the best estimates available when the data were collected.” The same answer says new data and methods “have contributed to more accurate estimates for 2019”, and that those estimates cover seven major pathogens where 2011 covered 31.
- It does not say foodborne illness has gone up or down. CDC says the 1999 and 2011 sets cannot be compared to assess trends, and that the 2019 and 2011 sets cannot be meaningfully compared.
- It does not say the 2019 figure replaced the 2011 figure. The 9.9 million covers six named pathogens and has no unspecified agents, so it is a total for those pathogens only. Set side by side, 48 million and 9.9 million show no fall.
- It says nothing against CDC’s estimates or methods.
- It does not say CDC is unreliable. CDC stated the data years in the 2011 papers, said at its first briefing that the data centred around 2006, and says the 2011 estimates were not wrong. This file records which pages print which figure, and when.
- It does not say anyone misled a reader. CDC’s press release disclosed the 38 million on the first day, the papers published that day state the data years and the 2006 population, and CDC’s briefing said the data centred around 2006.
- It does not say any office should change a sentence. The wordings under stakes are quoted as they stand, and none is recommended.
- It does not say food safety rules rest on this number. FDA states that its preventive controls regulation is required by statute and that its burden figures are not the rule’s expected benefits, and no rule outcome was shown to turn on the number.
- It argues no policy. It takes no side on FoodNet funding, FDA rules, any budget or any current dispute over CDC.
- It makes no prediction, including about what CDC may publish next.
- It does not claim to be first. GAO set out most of this on February 3, 2025. See prior work.
Observed for the rows. The age note is SOVRINT arithmetic.
The years
The newest input found ends in 2008, which is at least 17.8 years before this file’s date (SOVRINT arithmetic).
Before the first rowOldest input found: California cases behind the Brucella foodborne share, by way of the 1999 paper.
1973 to 1992
“Mostly” in both papers covers exceptions like this one.
Major Technical Appendix 1 (archived copy)
Two more rows reach back past 2000: T. gondii serology starts in 1999, and outbreak counts for Streptococcus group A start in 1996, and in 1981 for hospitalizations and deaths.
Left of the axis: each rule spans an input’s data years
- T. gondii serology 1999–2004
- Norovirus outbreaks 2000–2005
- Hospitalizations 2000–2006
- Deaths 2000–2006
- Population at risk 2006
- Rotavirus group 2006
- FoodNet Population Surveys 2000–2001, 2002–2003, 2006–2007
- FoodNet sites’ share 2000 to 2007
- Case-patient reports 2000–2007
- Vibrio reports 2000–2007
- Outbreak illnesses 2000–2007
- Tuberculosis reports 2004–2007
- Outbreak multiplier 2004–2008
- FoodNet case counts 2005–2008
Right of the axis: one mark per Federal Register document
42 documents from the written search
| Inputs with data from this year | Year | Federal Register documents | Dated events |
|---|---|---|---|
| T. gondii serology, Norovirus outbreaks, Hospitalizations, Deaths, FoodNet Population Surveys, FoodNet sites’ share, Case-patient reports, Vibrio reports, Outbreak illnesses | 2000 | ||
| T. gondii serology, Norovirus outbreaks, Hospitalizations, Deaths, FoodNet Population Surveys, FoodNet sites’ share, Case-patient reports, Vibrio reports, Outbreak illnesses | 2001 | ||
| T. gondii serology, Norovirus outbreaks, Hospitalizations, Deaths, FoodNet Population Surveys, FoodNet sites’ share, Case-patient reports, Vibrio reports, Outbreak illnesses | 2002 | ||
| T. gondii serology, Norovirus outbreaks, Hospitalizations, Deaths, FoodNet Population Surveys, FoodNet sites’ share, Case-patient reports, Vibrio reports, Outbreak illnesses | 2003 | ||
| T. gondii serology, Norovirus outbreaks, Hospitalizations, Deaths, FoodNet sites’ share, Case-patient reports, Vibrio reports, Outbreak illnesses, Tuberculosis reports, Outbreak multiplier | 2004 | ||
| Norovirus outbreaks, Hospitalizations, Deaths, FoodNet sites’ share, Case-patient reports, Vibrio reports, Outbreak illnesses, Tuberculosis reports, Outbreak multiplier, FoodNet case counts | 2005 | ||
| Hospitalizations, Deaths, Population at risk, Rotavirus group, FoodNet Population Surveys, FoodNet sites’ share, Case-patient reports, Vibrio reports, Outbreak illnesses, Tuberculosis reports, Outbreak multiplier, FoodNet case counts | 2006 | ||
| FoodNet Population Surveys, FoodNet sites’ share, Case-patient reports, Vibrio reports, Outbreak illnesses, Tuberculosis reports, Outbreak multiplier, FoodNet case counts | 2007 | ||
| Outbreak multiplier, FoodNet case counts | 2008 | ||
| No data year after 2008 turned up in any of the five appendices outside their reference lists. | 2009 | ||
| 2010 | 2010-12-15 CDC press release and briefing print 48 million and 1 in 6. Both papers go online. CDC press release | ||
| 2011 | 3 Federal Register documents | 2011-05-05 First Federal Register documents found: two FDA rules the same day, 2011-10953 and 2011-10955. FDA, 2011-10953 | |
| At least 17.8 years | 2012 | 5 Federal Register documents | |
| 2013 | 7 Federal Register documents | ||
| 2014 | 3 Federal Register documents | ||
| 2015 | 4 Federal Register documents | ||
| 2016 | 4 Federal Register documents | ||
| 2017 | 1 Federal Register document | ||
| 2018 | 2 Federal Register documents | ||
| 2019 | 1 Federal Register document | 2019 The year CDC’s later estimates describe, circa 2019: six pathogens for illnesses and seven for hospitalizations and deaths. Major Pathogens, 2019 paper | |
| 2020 | 3 Federal Register documents | ||
| 2021 | 2 Federal Register documents | ||
| 2022 | 0 Federal Register documents | ||
| 2023 | 4 Federal Register documents | ||
| 2024 | 2 Federal Register documents | ||
| 2025 | 0 Federal Register documents | 2025-03-19 CDC’s later paper. CDC says: “We cannot meaningfully compare the estimates from 2019 with the ones published in 2011.” CDC Q&A (archived copy) | |
| 2026 | 1 Federal Register document | 2026-06-02 Last printing found in the Federal Register. CDC, 2026-11011 |
SOVRINT arithmetic
The newest input found is FoodNet case counts ending in 2008, and the surveys behind 91.8 percent of the total end in 2007. The papers give years with no end month, so these ages count from the last day of the year and are minimums, rounded down: at least 17.8 years and at least 18.8 years to November 6, 2026. The population base is 2006, which is 20 years by calendar year. CDC’s press release is 15.9 years old. These are ages of inputs. They say nothing about whether the estimate is right.
CDC’s stated plan ran through three wordings: 5 to 10 years, about 10 years, and no interval. The file does not call any date missed. “Will likely” is an expectation, and CDC gives its reasons for the time taken on the Q&A page. GAO describes the three studies as “about 10 years apart”.
CDC’s Q&A of 2025-03-19 answers “Why did it take so long to publish these 2019 estimates?” with a multi-step process, data that were not immediately available, and limits during the pandemic.
Not drawn, because no single range of years is printed for them: Foodborne share for unspecified agents, the 25 percent; Domestic share for unspecified agents, the 98 percent; Norovirus share of acute gastroenteritis.
All 27 dated rows
1973 to 1992
Oldest input found: California cases behind the Brucella foodborne share, by way of the 1999 paper. “Mostly” in both papers covers exceptions like this one.
1999–2004
NHANES serology used for T. gondii.
2000–2001, 2002–2003, 2006–2007
FoodNet Population Surveys supply the illness rate behind unspecified agents and norovirus, 91.8 percent of the total (SOVRINT arithmetic).
2000–2005
Norovirus outbreaks behind the 26 percent foodborne share.
2000–2006
National hospital and death data for acute gastroenteritis and some pathogens.
2000–2007
National reports and outbreak counts.
2005–2008
FoodNet case counts, the newest input found.
2006
Population base: 299 million persons, used for every estimate.
2010-12-15
CDC press release and briefing print 48 million and 1 in 6. Both papers go online. Asked why 2011, CDC’s division lead says it is the journal’s date and the data “spanned a number of years centering around 2006.”
2010-12-15
At the same briefing CDC’s division lead says of such estimates: “this is not something that we can just generate year by year.”
2010-12-27 capture
CDC’s Q&A says “The burden of foodborne illness will likely be estimated again in 5 to 10 years.”
2011-01
The journal issue date, and the “2011” in “2011 estimates”.
2011-05-05
First Federal Register documents found: two FDA rules the same day, 2011-10953 and 2011-10955.
Between 2016-07-27 and 2016-11-08 captures
CDC’s Q&A changes to “The burden of foodborne illness will likely be estimated again in about 10 years.” The new text carries a page date of July 12, 2016, but captures of late July 2016 still show the old text and the old date.
2024-12-10
CDC page shows 9.9 million and says CDC is planning to publish updated estimates for seven pathogens. Earliest sighting of 9.9 million found.
2025-02-03
GAO sets out the 2000-2008 data range, says the coming study has no unspecified agents, and gives CDC’s reason.
2025-03-19
CDC’s estimates for 2019: 9.9 million, a total for six named pathogens only. CDC’s Q&A gives no interval for updates and says estimates “are typically published when new or updated data sources or improved methods become available.”
2025-03-19
CDC’s Q&A says “Yes, CDC is working on estimates for another set of known pathogens.” It also says the estimates “rely on certain data sources that are not updated regularly.”
2025-04-25 to 2025-04-30
CDC’s old main page, with 48 million and 38.4 million, is live on the first date and redirects to the new page by the second.
Wayback captures, archive.cdc.gov copy (archived copy) (second capture)
2026-04
CDC’s FY 2027 justification prints the ratio and count with no “estimates”.
CDC FY 2027 Congressional Justification, p. 48 (archived copy)
Between 2026-05-11 and 2026-05-30 captures
CDC’s Facts page reorders its top-5 lists to match CDC’s 2019 burden page, gains a Sources line naming the 2019 paper, and keeps 48 million.
2026-06-02
Last printing found in the Federal Register.
Between 2026-08-08 and 2026-08-31 captures
CDC’s restaurant page swaps the ratio and count for 9.9 million.
2026-09-17 capture, post undated
A post on foodsafety.gov, a site managed by HHS, says CDC released on March 19 “the latest estimates providing the most accurate picture yet of foodborne illnesses in the United States” and gives 9.9 million for six germs. It prints no year and no 48 million in the text read. The same site’s Food Poisoning page carries the 48 million sentence in a capture of 2026-10-24. Neither page is a CDC text saying which figure to use.
foodsafety.gov, Wayback captures (archived copy) (second capture)
2026-09-17
FDA’s 2026 Food Code repeats the 48 million sentence.
2026-10-14
The next notice in the collection that printed the ratio on 2026-06-02 has no burden sentence. The filing changes from Extension to Revision and says CDC received four public comments, which were not read here. No reason for leaving the sentence out is stated.
2026-10-16
CDC’s Facts page shows 48 million in the latest capture read. A summary tool showed the same sentence on 2026-11-06.
Observed
Eighteen lines
- About 48 million people (1 in 6 Americans) get sick, 128,000 are hospitalized, and 3,000 die each year from foodborne diseases, according new estimates from the Centers for Disease Control and Prevention.CDC press release
- Combining the estimates for unspecified agents and major known pathogens provides an estimate of the total effect of contaminated food consumed in the United States: 47.8 million episodes of illness, 127,839 hospitalizations, and 3,037 deaths (Table).Unspecified Agents paper, Emerging Infectious Diseases 17(1):16-22, CDC authors
- Each year about 48 million people (1 in 6 Americans) are sickened, 128,000 are hospitalized, and 3,000 die from food borne diseases, according to recent data from the Centers for Disease Control and Prevention.FDA interim final rule, Federal Register 2011-10953
- CDC estimates that each year approximately 48 million illnesses, 128,000 hospitalizations, and 3,000 deaths are food related (Ref. 45) (Ref. 46). These numbers include all illnesses that CDC estimates are attributable to food, including those illnesses caused by unspecified agents.FDA proposed preventive controls rule, Federal Register 2013-00125
- Each year about 48 million people (1 in 6 Americans) get sick; 128,000 are hospitalized; and 3,000 die from food borne diseases, according to 2011 data from the Centers for Disease Control and PreventionFDA final rule, Federal Register 2013-12833
- About 48 million people in the U.S. (1 in 6) get sick, 128,000 are hospitalized, and 3,000 die each year from foodborne diseases, according to recent data from the Centers for Disease Control and Prevention.FDA, Food Safety Modernization Act page
- An estimated 47.8 million foodborne illnesses occur annually in the United States, resulting in 127,839 hospitalizations, and 3,037 deaths annually.CDC 30-day notice, Federal Register 2024-27469, OMB 0920-0792
- Each year, foodborne illnesses sicken one in six Americans, and thousands die, according to CDC’s most recent estimates.GAO-25-107613, report to congressional addressees, Highlights, “Why GAO Did This Study”
- Although the U.S. food supply is generally considered safe, tens of millionschanged of Americans get sick from foodborne illness every year, and some die, according to estimates from the Department of Health and Human Services’ (HHS) Centers for Disease Control and Prevention (CDC).GAO-25-107606, Food Safety: Status of Foodborne Illness in the U.S., p. 1
- The major pathogens, excluding Toxoplasma gondii, caused an estimated 9.9 millionchanged domestically acquired foodborne illnesses.CDC, Estimates: Burden of Foodborne Illness in the United States
- The Federal government estimates that there are about 48 million cases of foodborne illness annually – that’s about 1 in 6 Americans each year.FDA, Food Safety Education Month page
- CDC estimates that 48 million people get sick from a foodborne illness every year.CDC, Facts About Food Poisoning
- Each year, 48 million people – one in six Americans – experience foodborne illness, resulting in approximately 128,000 hospitalizations and 3,000 deaths.CDC FY 2027 Congressional Justification, p. 48
- The Centers for Disease Control and Prevention (CDC) estimates that each year roughly 1 in 6 Americans (or 48 million people) get sick, 128,000 are hospitalized, and 3,000 die of foodborne diseases.Virginia Department of Health, Foodborne Diseases and Outbreaks
- An estimated one in six Americans each year becomes ill with a foodborne disease.CDC 60-day notice, Federal Register 2026-11011, OMB 0920-1198
- CDC estimates that each year in the Untied States, the most common foodborne pathogens cause an estimated 9.9 millionchanged illnesses, 53,000 hospitalizations, and 931 deaths.CDC, About Restaurant Food Safety, National Center for Environmental Health
- Scallan et al. (2011a,b) estimated that foodborne diseases cause approximately 48 million illnesses, 128,000 hospitalizations, and 3,000 deaths in the United States each year.FDA Food Code 2026, Preface, p. v
- The CDC estimates that each year roughly 1 out of 6 Americans (or 48 million people) get sick, 128,000 are hospitalized, and 3,000 die of foodborne diseases.Northeast Colorado Health Department, Consumer Food Safety
The 18 lines are chosen examples with no sampling frame. They show no rate and no order of adoption. The Register years between 2013 and 2024 are in the series below. No news line is chosen, because the file follows the sentence as drafters of public texts meet it. Each quoted line is Observed: the wording was on the page when read for this file on or before November 6, 2026, whether or not any claim inside it is correct. All 18 are government or journal text and are exact. Lines marked changed make a different claim and are shown for contrast. A carries mark is true only when the text applies that thing to the number itself, in the quoted sentence, a sentence beside it or a footnote on it. A numbered reference counts for source, and not for publication year. For the paper itself the marks describe the paper as a whole.
The ratio and the count are one claim in CDC’s text, so both sit on the axis. Where a line prints both, the ratio is the keyword. Register spellings also include “1 in six”. Tens of millions and 9.9 million are not the claim. They mark lines that use other wording, and those lines are labelled changed. “Tens of millions” is not new wording: CDC’s briefing used it beside the ratio on 2010-12-15.
- Lines above
- 18, chosen
- They carry
- unit 7, data years 2
- And
- publication year 2, source 15
What the marks mean
- Unit
- means the number is counted in the paper’s unit, episodes or illnesses or cases, so people and Americans do not count.
- Data years
- means a named year or range the data come from.
- Publication year
- means 2011 or a citation year printed beside the number.
- Source
- means the line says where the number comes from.
Observed for what is printed. SOVRINT arithmetic is labelled.
The inputs
Each row is an input to the total, with the years printed for it. “Mostly” in both papers covers the exceptions in the last rows. No data year after 2008 turned up in any of the five appendices outside their reference lists. Major Technical Appendices 1 and 3 were searched and read in part, not line by line.
| Input | Used for | Data source | Data years | Value | Printed in |
|---|---|---|---|---|---|
| Population at risk | Everything | US Census population estimate | 2006 | 299,000,000, held constant | Both papers. Unspecified appendix table |
| Rate of acute gastroenteritis | The 178.8 million, and so the 38.4 million and norovirus | FoodNet Population Surveys, random-digit-dial telephone surveys in FoodNet sites | 2000–2001, 2002–2003, 2006–2007 | 0.49, 0.54 and 0.73 episodes per person per year. Average 0.6 | Unspecified paper. Table A1 |
| Share of the US population in FoodNet sites during the surveys | Reach of the rate | FoodNet | 2000 to 2007 | 11% (in 2000) to 15% (in 2007) | Unspecified paper |
| Foodborne share for unspecified agents, the 25 percent | The 38.4 million | Ratio borrowed from the 24 known gastroenteritis pathogens, then widened | As for those pathogens | Low 0.173, modal 0.251, high 0.455. Variance parameter changed from 19 to 2 | Unspecified appendix table |
| Domestic share for unspecified agents, the 98 percent | The 38.4 million | Ratio borrowed from the 24 known gastroenteritis pathogens | As for those pathogens | Low 0.950, modal 0.979, high 0.993. Variance parameter changed from 12 to 2 | Unspecified appendix table |
| Foodborne share of norovirus, main driver of the 25 percent | The 25 percent and the 5.46 million norovirus illnesses | 179 norovirus outbreaks examined by CDC, CDC unpublished data | 2000–2005 | 26%, from 3,628 of 13,944 persons ill | Major Technical Appendix 1 (archived copy) |
| Norovirus share of acute gastroenteritis | Norovirus illnesses | Studies in other industrialized countries | Not printed as one range | 11% | Major paper |
| Laboratory-confirmed illnesses, 10 pathogens | Known pathogens | FoodNet active surveillance | 2005–2008 | Applied to 2006 population estimates | Table A1 |
| Case-patient reports, 5 pathogens | Known pathogens | National Notifiable Disease Surveillance System | 2000–2007 (2002–2007 for G. intestinalis) | Table A1 | |
| Vibrio reports | Known pathogens | COVIS | 2000–2007 | Table A1 | |
| Outbreak illnesses, 5 bacteria | Known pathogens | FDOSS | 2000–2007 (Streptococcus group A 1996–2007, and 1981–2007 for hospitalizations and deaths) | Table A1 | |
| Tuberculosis reports, M. bovis | Known pathogens | National Tuberculosis Surveillance System | 2004–2007 | Table A1 | |
| Seroprevalence, T. gondii | Known pathogens | National Health and Nutrition Examination Survey | 1999–2004 | Table A1 | |
| Rotavirus, astrovirus, sapovirus | Known pathogens | US Census resident population estimates | 2006 | 75% of children ill by 5 years of age | Table A1 |
| Outbreak underreporting multiplier | Five outbreak-only bacteria | FoodNet pathogens also reported as outbreak associated | 2004–2008 per Technical Appendix 4 | 25.5 | Major paper. Technical Appendix 4 |
| Hospitalizations for acute gastroenteritis | Unspecified agents and norovirus hospitalizations | NHDS, NIS, NAMCS and NHAMCS | 2000–2006 | Table A1. Unspecified paper | |
| Deaths from acute gastroenteritis | Unspecified agents and norovirus deaths | National Vital Statistics System | 2000–2006 | 1.5 per 100,000 (1.2, 2.4) | Table A1. Unspecified appendix table |
| Older exception: Brucella foodborne share | Known pathogens | California cases, by way of the 1999 paper | 1973 to 1992 | 50% assumed | Major Technical Appendix 1 (archived copy) |
Unspecified Agents paper (archived copy)
Emerging Infectious Diseases 17(1):16-22, January 2011, online 2010-12-15
“Most of our data were from 2000 through 2007, and all estimates were based on the US population in 2006 (299 million persons)”
The total of 47.8 million and its interval. The online date comes from CDC’s press release (“published Wednesday”) and briefing (“being published today”). The journal page prints no online date.
Major Pathogens paper (archived copy)
Emerging Infectious Diseases 17(1):7-15, January 2011, online 2010-12-15
“Data were mostly from 2000–2008, and all estimates were based on the US population in 2006 (299 million persons).”
The 9.4 million from 31 known pathogens and Table A1 of data sources. The online date comes from CDC’s press release (“published Wednesday”) and briefing (“being published today”). The journal page prints no online date.
The total, as the table prints it
| Row | Illnesses (interval) | Percent | Hospitalizations (interval) | Deaths (interval) |
|---|---|---|---|---|
| Major known pathogens (31) | 9,388,075 (6,641,440–12,745,709) | 20 | 55,961 (39,534–75,741) | 1,351 (712–2,268) |
| Unspecified agents | 38,392,704 (19,829,069–61,196,274) | 80 | 71,878 (9,924–157,340) | 1,686 (369–3,338) |
| Total | 47,780,779 (28,658,973–71,133,833) | 100 | 127,839 (62,529–215,562) | 3,037 (1,492–4,983) |
All estimates were based on US population in 2006.
How the paper reaches it, the unit and its conditions
How the paper reaches it
At 0.6 episodes per person per year on 299 million people, the paper reports 178.8 million episodes of acute gastroenteritis. Taking away 37.0 million from 24 known gastroenteritis pathogens leaves 141.8 million. The paper then applies 98 percent domestic and 25 percent foodborne, borrowed from those 24 pathogens, and prints 38.4 million. The paper’s figures come from simulation over probability distributions, so multiplying the rounded values shown here does not reproduce them.
The unit
- “(Hereafter, episodes of illness are referred to as illnesses.)”
- The rate is episodes per person per year, built from past-month recall.
- “we lack sufficient data on duration of illness, onset date, and multiple episodes in the past month necessary to estimate incidence”
- The papers also write “estimated number of ill persons” when describing how they scale up. So the paper is not strict either.
- The move to “people” and “1 in 6 Americans” is in CDC’s own release and briefing on the first day. A count of episodes divided by a population is a loose ratio of people, and the file does not press that point further.
What is not printed
- “48 million”, “1 in 6”, “one in six” and “1 out of 6” appear in neither 2011 paper, in no table page and in none of the five appendices (exact search). “47.8 million” appears once, in the Unspecified paper.
- No sentence in either paper says what share of Americans gets sick. The ratio first appears in CDC’s press release and briefing of 2010-12-15.
- The papers print no current count for any year after their data. They describe an estimate on a 2006 base.
- CDC’s press release of 2010-12-15 prints no data years, no population base, no interval and the word episodes nowhere. It does print 9.4 million known, 38 million unspecified and “roughly 80 percent”.
Conditions the papers state
The papers state these conditions themselves. They show the authors’ care and are listed so the sentence’s source can be read in full.
“If the proportions acquired by domestic foodborne transmission were similar to those for known gastroenteritis pathogens, then an estimated 38.4 million”
Unspecified paper, abstract. Attaches to: The 38.4 million, and so the total
“The proportion of these unspecified agents acquired through domestic foodborne transmission is unknown”
Unspecified paper, Results. Attaches to: The 38.4 million, and so the total
“The largest and most tenuous assumption in the estimates for the unspecified agents is the assumption that features of the modeled disease process for the remaining acute gastroenteritis agents are identical to those for the aggregate 24 known gastroenteritis pathogens. Although we have used a fairly straightforward method, the specific choices are subjective, and we introduced nonstatistical uncertainty to reflect this.”
Unspecified Technical Appendix, pp. 8-9. Attaches to: The 38.4 million, and so the total
“the foodborne proportion was driven largely by norovirus”
Unspecified paper, Discussion. Attaches to: The 25 percent
“the accuracy of our estimate of the number of acute episodes of gastroenteritis from the FoodNet Population Surveys has not been validated”
Unspecified paper, Discussion. Attaches to: The 178.8 million
“Combining different approaches is not optimal because the methods themselves may affect the estimates derived.”
Unspecified paper, Discussion. Attaches to: The total
“our estimates do not include unspecified foodborne illnesses that do not typically cause signs of acute gastroenteritis”
Unspecified paper, Discussion. Attaches to: The total
“the claim that 80% of foodborne illnesses are unspecified must be treated with caution”
Unspecified paper, Discussion. Attaches to: The split between known and unspecified agents, and not the total. The paper’s reason is that higher estimates for known pathogens would lower the unspecified estimate
Later work, the earlier estimate and the earliest sighting
Later work
Major Pathogens, 2019 paper, Emerging Infectious Diseases 31(4):669-677, April 2025, original publication 2025-03-19
- Covers
- Six pathogens for illnesses: Campylobacter, C. perfringens, invasive Listeria, norovirus, nontyphoidal Salmonella and STEC. T. gondii is added for hospitalizations and deaths.
- Result
- 9.9 million illnesses (90% CrI 5.9–15.4 million), about 53,300 hospitalizations and 931 deaths, circa 2019.
- Data years
- FoodNet 2017–2019, Listeria Initiative 2016–2019, FDOSS 2010–2019, Kaiser Permanente Northwest 2014–2016, MarketScan 2001–2015, FoodNet Population Survey 2018–2019, NIS 2016–2019.
- Leaves out
- Unspecified agents, any total that includes them, and the other pathogens on the 2011 list of 31. “Unspecified” appears only in two reference titles. 48 million, 47.8 million, 9.4 million and 1 in 6 have no hits.
- A trap
- The 37.6 million in the paper and CDC’s Q&A is six pathogens by all routes. It is not a foodborne total.
What CDC says about the two sets
- “We cannot meaningfully compare the estimates from 2019 with the ones published in 2011.”
- “CDC's 1999 and 2011 estimates were not wrong; they were the best estimates available when the data were collected.”
- “New surveillance data, methods, and other factors have contributed to more accurate estimates for 2019.”
- This is the next sentence of the answer that says the estimates were not wrong. The answer also says “which we believe have resulted in more accurate estimates than their predecessors”. It does not mention the 48 million or unspecified agents. This file gives no verdict on the 48 million.
- “Similar to estimates circa 2006, norovirus was the leading cause of domestically acquired foodborne illness (1).”
- “These estimates were the product of a multi-step process, with each step requiring careful and thorough analyses”
- “Yes, CDC is working on estimates for another set of known pathogens.”
- “the foodborne illness burden estimates also rely on certain data sources that are not updated regularly.”
CDC Q&A (archived copy), page dated 2025-03-19. The quotations from CDC, other than the one from the paper, are from CDC’s Q&A page, not the paper. They were read from an existing capture of 2026-10-18, because the live page returned 403 to curl. The page’s printed date is March 19, 2025.
GAO-25-107606, p. 10, attributing CDC officials
“Officials said that for the 2025 study, CDC focused on updating the disease burden estimates for priority pathogens and considers these estimates to be more valuable to food safety regulators than estimates for unspecified agents.”
The earlier estimate
The 1999 estimate also worked down from a gastroenteritis total, with no interval, using other data and methods. CDC says the 1999 and 2011 sets cannot be compared to measure trends.
0.79 episodes per person on 267.7 million people (1997) gives 211 million. Taking away 38.3 million known leaves 173 million. 36 percent of that is 62 million, and 13.8 million known brings it to 76 million.
The earliest sighting found
December 15, 2010. CDC’s press release and its 11:00am ET briefing. Earliest found for both the rounded count and the ratio. A pre-release text expects new numbers and gives none: Food Safety News dated 2010-12-13. Three searches found no embargoed report, abstract or talk. This is a sighting and does not establish first authorship.
2010-12-13: Food Safety News
Observed, except the sentences marked Assessed
What CDC prints now
In the latest texts read CDC prints three forms at once. No CDC text was found saying which to use.
As of November 6, 2026
| Form | Where it stands |
|---|---|
| 48 millionThe count | The count is on its Facts page and Food Safety Basics page.
|
| 9.9 millionThe named-pathogen figure | The named-pathogen figure is on its burden page and restaurant page.
|
| one in sixThe ratio | The ratio is in its FY 2027 justification and was in its Register notice of 2026-06-02.
|
| Collection | Last | After it | Status |
|---|---|---|---|
| OMB 0920-1112, FoodNet Population Survey | 2018-12-17 | No later notice under that number or title in the written searches. | No later notice found, so nothing can be said about dropping |
| OMB 0920-0997, Standardized National Hypothesis Generating Questionnaire | 2023-06-27 | A new CDC module (2024-07287, 2025-10724) takes over the form and carries no sentence. | Later notice exists without the sentence |
| OMB 0920-0792, Environmental Health Specialists Network | 2024-11-25 | First carried the figure on 2021-04-05. No later notice found. | No later notice found, so nothing can be said about dropping |
| OMB 0920-1198, Cyclosporiasis National Hypothesis Generating Questionnaire | 2026-06-02 | The next notice, 2026-20906 of 2026-10-14, has no burden sentence. The collection continues. | Later notice exists without the sentence |
| FDA rules | 2015-09-17 | None later in the audited set. | No later document in the audited set |
| Unified Agenda, HHS-FDA entries | 2015-12-15 | None later in the audited set. | No later document in the audited set |
These are dates only. No stated reason was found for any change, and none is offered.
Observed
How the wording moved
These rows order dated sightings. Each date is the earliest found for that wording in the texts the row names, and a row says so where the wording is older than its first date. They do not show who copied whom, and they are not a chain of transmission. For most of its life the sentence did not change, and the list below says where.
2010-12-15
Episodes become people
From “47.8 million episodes of illness” in the paper
To “48 million people (1 in 6 Americans)”
“About 48 million people (1 in 6 Americans) get sick”
The note
The same day as the papers, in CDC’s own text. Later writers did not make this change. The papers also write “estimated number of ill persons” when describing scaling, so this point should not be pressed.
2010-12-15
A rounded count and a ratio appear
From 47,780,779 in the table, “47.8 million” in the text
To “48 million” and “one in six”
“these estimates indicate that one in six Americans or 48 million people get sick each year”
CDC press briefing transcript, 11:00am ET (archived copy)
The note, and 1 more
Neither form is in the papers or their appendices. Earliest sighting, which does not establish authorship.
2010-12-27 capture
CDC estimates that each year roughly 1 out of 6 Americans (or 48 million people) gets sick, 128,000 are hospitalized, and 3,000 die from foodborne diseases.
CDC, Estimates of Foodborne Illness main page (archived copy)
A third form in the same month. “1 out of 6” survives on the local line of 2026.
2010-12-21
The word estimate drops out
From “CDC estimates” and “it is estimated”
To “48 million Americans are sickened”, “according to recent data” and “data showing”
“Each year, 48 million Americans are sickened from consuming contaminated food”
A House floor statement, Congressional Record
The note, and 4 more
Earliest text found without the word, six days after the release. A statement printed in the same day’s debate calls the figures CDC estimates. Most lines keep the word: 31 of 42 key sentences have “estimat” outside a page title or address, and 11 do not.
2011-05-05
according to recent data from the Centers for Disease Control and Prevention
FDA interim final rule, Federal Register 2011-10953 (archived copy)
First in the Register, with 2011-10955 the same day. See the row on recent data. Nine of the 11 key sentences without the word are dated before 2014-12-22.
2014-12-22
data showing that such illnesses strike one in six Americans each year
Unified Agenda, HHS-FDA entry, Federal Register 2014-28927
The phrase “data showing”. Earliest for this phrase only.
Fiscal 2024 justification
CDC estimates the burden of foodborne disease in the U.S. to be approximately 48 million cases per year (one out of every six Americans), 128,000 hospitalizations, and 3,000 deaths per year.
CDC FY 2024 Congressional Justification (archived copy)
Before the change in CDC’s budget text. FY 2025 says the same, FY 2026 has no such sentence.
File created 2026-04-02
Each year, 48 million people – one in six Americans – experience foodborne illness
CDC FY 2027 Congressional Justification, p. 48 (archived copy)
Cases become people and “estimates” is gone. The file records wording only and takes no view on any budget request.
2011-03-15
Two sets are read as a fall
From CDC: the 1999 and 2011 sets cannot be compared to assess trends
To “from 76 million to 48 million”
A member at a House appropriations subcommittee hearing
The note
From the GovInfo count, read at the passage only. The Major Pathogens abstract says the estimates “cannot be compared with prior (1999) estimates to assess trends”.
2011-05-05
New estimates become recent data
From “new estimates” in the press release
To “recent data”
“according to recent data from the Centers for Disease Control and Prevention”
FDA interim final rule, Federal Register 2011-10953
The note, and 1 more
Seven key sentences in the Register say “recent data”, the last on 2015-12-15. Three say “recent estimates”. None says “most recent”.
Content current as of 2024-02-05
according to recent data from the Centers for Disease Control and Prevention
FDA, Food Safety Modernization Act page
On the page as read on 2026-11-06.
2013-02-05
“2011 estimates” appear as “2011 data”
From 2011 as the journal date, per CDC’s briefing
To “2011 data”
“according to 2011 data from the Centers for Disease Control and Prevention”
FDA rule, Federal Register 2013-02497
The note
Also 2013-12833 and 2014-07550. Each gives a CDC page address. The file does not say anyone meant to mislead. CDC itself calls them the 2011 estimates.
2015-09-29 capture, page dated 2015-08-31
CDC’s web pages split into ratio pages and count pages
From Ratio and count together on CDC pages
To “one in six” alone on one page, “48 million” alone on others
“Each year, one in six Americans get sick from contaminated foods or beverages; 3,000 die.”
CDC, CDC and Food Safety page (archived copy)
The note, and 1 more
The count was still there in a capture of 2015-03-16. On the CDC web pages checked, the two forms parted by 2015. CDC’s Register text printed the ratio alone from 2012-07-24.
2016-08-27 capture, page updated 2016-08-19
CDC estimates 48 million people get sick, 128,000 are hospitalized, and 3,000 die from foodborne diseases each year in the United States.
CDC, main burden page (archived copy)
The ratio left between captures of 2016-08-08 and 2016-08-27.
2015-09-29
The total is credited to named pathogens
From A total that is 80 percent unspecified agents
To “one of 31 major pathogens”
“one in six Americans suffers from foodborne illness attributed to one of 31 major pathogens”
Outside witness statement, House hearing
The note, and 1 more
From the GovInfo count, read at the passage only. The same statement then gives 9.4 million for those pathogens.
2022-07-13
nearly 48 million people get sick from one of the 31 pathogens
A House member’s statement, Congressional Record
From the GovInfo count, read at the passage only.
2021-04-05
The unrounded figures enter the Register
From “48 million”, “128,000”, “3,000”
To “47.8 million”, “127,839”, “3,037”
“An estimated 47.8 million foodborne illnesses occur annually in the United States”
CDC 60-day notice, Federal Register 2021-06883, OMB 0920-0792
The note, and 2 more
First in the Register: four notices in one collection, through 2024-11-25. They keep the paper’s precision with no interval, and “illnesses” is faithful to the paper. The figures had not left CDC’s own table page or the record, as the entries below show.
Page last reviewed 2018-11-05, capture of 2025-04-26
47.8 million (28.7–71.1 million)
CDC, archived findings page (archived copy)
The table row for the total, with its interval.
2012-03-07
the roughly 47.8 million illnesses yearly
Senate hearing record, CHRG-112shrg72333
From the GovInfo count, read at the passage only.
2025-02-03
GAO moves off the ratio
From “one in six Americans”, 2025-01-22
To “tens of millions of Americans”
First GAO use found “tens of millions of Americans get sick from foodborne illness every year”
The note, and 2 more
First GAO use found, twelve days after “one in six” and before CDC’s paper appeared. No reason is stated, and the phrase is not new. GAO’s move is to a form CDC had already spoken. The same report’s Key Takeaways print CDC’s forthcoming figure and call the 2011 work “Prior CDC estimates”, and its Figure 5 keeps “9.4 million from known pathogens, 38.4 million from unspecified agents” for 2011. GAO-26-107394 of 2026-01-07 keeps the new form.
2010-12-15
It affects tens of millions of Americans each year. And I repeat, that translates into one in six Americans ill from foodborne illnesses each year.
CDC press briefing transcript (archived copy)
Earliest sighting of the phrase. The two forms sat side by side on the first day.
2026-01-07
Each year, foodborne illnesses sicken tens of millions of Americans and cause thousands of deaths.
Latest GAO form found.
Between 2026-05-11 and 2026-05-30 captures
A source line points to a paper without the number
From No source title in the text of the capture of 2026-05-11
To “Foodborne illness acquired in the United States—Major pathogens, 2019”
“48 million”
CDC, Facts About Food Poisoning (archived copy)
The note
The page keeps 48 million, 128,000 and 3,000 while its top-5 lists are reordered to match CDC’s 2019 burden page and its only listed source becomes the 2019 paper. The printed date stays Nov. 24, 2025. The file records this and offers no reason.
Page dated 2026-08-18
One CDC page swaps one figure for the other
The only swap found. The file does not call it a correction. This page’s content source is the National Center for Environmental Health, and the burden page’s is NCEZID.
CDC, About Restaurant Food Safety (archived copy)
The sentences on that page
“roughly 1 in 6 Americans (or 48 million people)”
“the most common foodborne pathogens cause an estimated 9.9 million illnesses, 53,000 hospitalizations, and 931 deaths”
2026-10-14
A later notice in one collection has no burden sentence
From “An estimated one in six Americans each year becomes ill with a foodborne disease.” on 2026-06-02
To No burden sentence
“Foodborne outbreaks of cyclosporiasis have been reported in the United States since the mid-1990s and have been linked to various types of imported fresh produce”
CDC 30-day notice, Federal Register 2026-20906, OMB 0920-1198 (archived copy)
The note
This now stands where the burden sentence stood. The text has no ratio, no count and no burden sentence, and it states no reason. The notice changes the request from an Extension to a Revision and says CDC received four public comments. The comments were not read here, and the notice links none of this to the background text. The collection continues.
Observed
What did not move
The number, the tense and the absence of data years held from 2010 to 2026 in the texts read. CDC’s and FDA’s rounded sentence prints 48 million, 1 in 6, 128,000 and 3,000 throughout. The exceptions are the four CDC notices with the unrounded 47.8 million, 127,839 and 3,037, and looser wording in some floor statements and hearing records.
- All 42 Register documents use the present tense.
- The press release printed no data years on the first day, and 0 of 42 Register documents print a data year or the 2006 population.
- The interval was absent from the release on the first day, and it stays in the table. 0 of 42 Register documents print an interval for the total. Background sentences rarely carry intervals, so this is a minor point.
- The unspecified share was disclosed on the first day: “The remaining 38 million illnesses result from unspecified agents”. In the Register, 1 of 42 documents says so.
- Within a collection the wording mostly holds. The Register’s sentence families track the paperwork collection, not the year. 0920-0997 has been ratio-only since 2012.
- FDA’s Food Code sentence is the same in the 2022 and 2026 editions. CDC’s Food Safety Basics sentence is the same under page dates of 2024-04-29, 2025-11-24 and 2026-03-09. That shows the sentence survived those dates, and it does not show anyone reviewed it.
- Healthy People 2030 has printed “About 1 in 6 people in the United States get foodborne illnesses every year.” from its first capture, 2020-09-30, to a fetch on 2026-11-06.
- State and local pages found keep CDC’s or FDA’s sentence. No state or local page found uses 9.9 million, in one search.
CDC’s Food Safety Basics sentence (archived copy). FDA’s Food Code sentence, 2022 edition. FDA’s Food Code sentence, 2026 edition (archived copy). Healthy People 2030
Observed for the counts. Assessed for what they can show.
The series
The written search
("1 in 6" | "one in six" | "1 in six" | "48 million" | "1 out of 6" | "47.8 million") (foodborne | "food borne" | "food-borne")
The series meets rule 1.2 as a thin audited series. It is one written search of one named source, counted in items per calendar week, with copies folded and every claim document read against the original under stage 2.2. It holds 32 items in 29 of 810 weeks.
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
A week with an item
A week with a document only, folded into an item in an earlier week
All dates the API holds to 2026-11-06. Weeks run from the Monday of the first claim document, 2011-05-02, to 2026-11-02, which is 810 calendar weeks by Monday date
How it was counted
Federal Register (federalregister.gov), public API, with GPO official copies at govinfo.gov for text after rate limits, November 6, 2026. Distinct items per calendar week. A 60-day notice and the 30-day notice after it for the same collection and cycle are one item when the sentence family matches. A later renewal is a new item. Separate rules are separate items, even on the same day.
Written before counting. Each document keeps its own row and date. A later document folds into an earlier one only when it is the same publisher’s next step in the same paperwork cycle, or the same document reprinted, and repeats the key sentence. As applied, repeats means the same sentence family. A family groups key sentences that match after normalising case, whitespace, punctuation, digit and word numerals and reference marks.
Folding every step of a cycle whatever the wording gives 29 items. The gap is the 2012 and 2016 cycles of OMB 0920-0997, where the wording differs between steps.
Too few to read
Rule 1.5 on folded items: the last 12 full weeks against the 12 before give 0 and 1, so Too few to read (floor 24). The last 52 full weeks against the 52 before give 1 and 0, under the floor of 104, so no percentage is given. By document the same windows give 0 and 1, and 1 and 1.
It is not a trend under rule 1.3, and no rise or fall is claimed. Rule 1.5 says a Federal Register series cannot be a macro trend, and the count has not risen.
- Raw records
- 55
- Documents
- 42
- Folded items
- 32, or 29 folded loosely
- Sentence families
- 18
- Near misses
- 1
- Weeks with an item
- 29 of 810
- Weeks with a document
- 39 of 810
- Most in one week
- 2
- Since 2021
- 5 weeks with an item, 9 with a document
What it can show
- When the sentence was printed in the Register, by which office, and in what words.
- The last printing found in each collection.
- That since 2015-12-15 it has appeared in the audited set only in CDC paperwork notices.
- That no document in the audited set prints a data year or the 2006 population beside the number.
What it cannot show
- Belief, reach, or whether the number drove any decision.
- A rise, a fall or a peak. The weekly rhythm mostly reflects the renewal clock of paperwork clearances and rulemaking calendars, not fresh decisions to restate the claim.
- Why any collection carries or lacks the sentence.
- Use outside the Register. CDC web pages, budget text and GAO reports stand outside this count.
Trends is not changed, and this file adds no topic to it.
What sits beside the number in the 42 documents
- Data year
- 0 of 42 print a data year beside the number.
- Population
- 0 of 42 print the 2006 population or 299 million.
- Publication year
- 4 of 42 print the publication year: 2012-4165 (“CDC 2011 Estimates”), and 2013-02497, 2013-12833 and 2014-07550 (“2011 data”).
- Recency
- 7 key sentences say “recent data” and 3 say “recent estimates”, all from 2011 to 2015. None says “most recent”.
- Frequency only
- The other key sentences carry only a frequency word: each year, per year or annually.
- Tense
- All 42 are in the present tense.
- The word estimate
- 31 key sentences have “estimat” outside a page title or address, and 11 have “data”.
- Form
- By key sentence: both ratio and count 16, ratio 19, count 7. By whole document: both 17, ratio 18, count 7.
- Noun
- 38 of 42 key sentences count Americans or people. 4 count illnesses, all in one collection.
- Source
- 7 of 42 give a checkable source: 2011-19333, 2012-4165, 2013-00125, 2013-02497, 2013-12833, 2014-07550 and 2015-21920. Only 2011-19333 and 2013-00125 cite the Unspecified Agents paper, which holds the total. 2015-21920 cites the Major Pathogens paper. That is a citation fact and is not a fault.
By agency, and known misses
By agency
CDC: 25, FDA: 13, Unified Agenda, HHS-FDA entries: 4.
Under rule 1.5 a bureau folds into its department, so CDC and FDA are HHS and this is close to a one-department series. All 20 documents from 2016 onward are CDC Paperwork Reduction Act notices. FDA rules and Agenda entries run from 2011-05-05 to 2015-12-15. 26 of 42 documents fall in 2011 to 2016 and 16 in 2017 to 2026.
Where a number drives a calculation the Register can keep a range. FSIS rules 2013-13669 and 2015-11916 carry CDC’s credible interval for E. coli O157:H7 into their figures. That is a separate number, and it is one point for a minor finding.
Known misses
The first two strings missed one document. The corrected search returned all 42 known claim documents. That is a statement about the audited set, because a further unusual wording could still be missing.
- 2017-06869, “1 in six”, found only through a collection-title search.
Near miss: 2020-18589, 2020-08-25. “The CDC estimates that millions of Americans become ill from foodborne illness every year, resulting in roughly 128,000 hospitalizations and 3,000 deaths.” USDA FSIS notice. No ratio or count, so not counted.
An item count is only as good as its search. So this file writes “found by the written search”, never “all”.
The 18 sentence families
F18 was split from F07 at review. 2013-17993 ends “(CDC)” and 2013-00123 does not, and the folding rule does not normalise that. Item and week counts are unaffected, because the two were always separate items.
F01, 2011-05-05 to 2011-08-01
Each year about 48 million people (1 in 6 Americans) are sickened, 128,000 are hospitalized, and 3,000 die from food borne diseases, according to recent data from the Centers for Disease Control and Prevention.
F02, 2012-02-13 to 2015-12-15
According to recent data from the Centers for Disease Control and Prevention, each year, about 48 million Americans get sick, 128,000 are hospitalized, and 3,000 die from foodborne diseases.
F03, 2012-02-23 to 2012-02-23
Each year about 48 million people (1 in 6 Americans) get sick, 128,000 are hospitalized, and 3,000 die from food borne diseases, according to recent data from the Centers for Disease Control and Prevention (CDC) (Estimates of Foodborne Illness in the United States--CDC 2011 Estimates, available at [address: CDC page] ).
F04, 2012-07-24 to 2012-07-24
Each year, it is estimated that roughly 1 in 6 Americans become ill with a foodborne disease.
F05, 2012-08-30 to 2016-08-25
It is estimated that each year roughly 1 in 6 Americans gets sick, 128,000 are hospitalized, and 3,000 die of foodborne diseases.
F06, 2012-10-09 to 2023-06-27
It is estimated that each year roughly 1 in 6 Americans get sick, 128,000 are hospitalized, and 3,000 die of foodborne diseases.
2012-24757, 2016-24668, 2019-22735, 2020-03342, 2023-04971, 2023-13568
F07, 2013-01-16 to 2013-01-16
Each year, about 48 million Americans (1 in 6) get sick, 128,000 are hospitalized, and 3,000 die from foodborne diseases, according to estimates from the Centers for Disease Control and Prevention.
F08, 2013-01-16 to 2015-09-17
Each year, about 48 million Americans (1 in 6) get sick, 128,000 are hospitalized, and 3,000 die from food-borne diseases, according to recent estimates from the Centers for Disease Control and Prevention (CDC).
F09, 2013-02-05 to 2013-05-30
Each year about 48 million people (1 in 6 Americans) get sick, 128,000 are hospitalized, and 3,000 die from food borne diseases, according to 2011 data from the Centers for Disease Control and Prevention ( [address: CDC page] ).
F10, 2014-04-04 to 2014-04-04
Each year about 48 million people (1 in 6 Americans) get sick from foodborne diseases, 128,000 are hospitalized, and 3,000 die, according to 2011 data from the Centers for Disease Control and Prevention ( [address: CDC page] ).
F11, 2014-09-24 to 2014-09-24
While the American food supply is among the safest in the world, the Centers for Disease Control and Prevention (CDC) estimates that 1 in 6 Americans is sickened by foodborne illness annually, resulting in about 3,000 deaths each year.
F12, 2014-12-22 to 2014-12-22
High-profile outbreaks of foodborne illness over the last decade and data showing that such illnesses strike one in six Americans each year have caused a widespread recognition that we need a new modern food safety system that prevents food safety problems in the first place not a system that just reacts once they happen.
F13, 2015-06-16 to 2018-12-17
It is estimated that each year, 48 million Americans (1 in 6) become ill, 128,000 are hospitalized, and 3,000 die as the result of a foodborne illness.
F14, 2016-12-22 to 2017-04-06
An estimated 1 in 6 Americans per year become ill with a foodborne disease.
F15, 2020-02-25 to 2023-09-13
An estimated one in six Americans per year becomes ill with a foodborne disease.
F16, 2021-04-05 to 2024-11-25
An estimated 47.8 million foodborne illnesses occur annually in the United States, resulting in 127,839 hospitalizations, and 3,037 deaths annually.
F17, 2026-06-02 to 2026-06-02
An estimated one in six Americans each year becomes ill with a foodborne disease.
F18, 2013-07-29 to 2013-07-29
Each year, about 48 million Americans (1 in 6) get sick, 128,000 are hospitalized, and 3,000 die from foodborne diseases, according to estimates from the Centers for Disease Control and Prevention (CDC).
How the search met the Method
- Method 1.2: distinct items per calendar week, from a written search of one named public source, with copies folded first.
- Search strings and definitions were written before any count. The final string is a corrected search, and its date and reason are stated above. The API list came from federalregister.gov. Its full-text pages returned repeated 429s, so requests stopped there. 33 of the 42 claim documents were read from federalregister.gov and 9 from the GPO official copy.
- Under rule 1.1 all 42 carry one claim. Count-only sentences are the same claim, because CDC’s text prints the count and the ratio as equals. 2020-18589 stays out because it states neither.
- Folding a 30-day notice into its 60-day notice extends rule 1.2’s folding of copies, so the document count is printed beside it.
The other 6 written searches, and the markers
Final (the series). 55 records, 42 true documents, 32 items. Corrected search, written after S2 missed 2017-06869. Reason: a printed form, “1 in six”, matched no earlier string. 13 records are false hits. The API list holds 55 records and 54 distinct numbers, because 01-1 is used by two documents of 2001.
S1
("1 in 6" | "one in six" | "48 million" | "1 out of 6") (foodborne | "food borne")
47 records, 37 true documents. The scouting string. Written before counting.
S2
("1 in 6" | "one in six" | "48 million" | "1 out of 6" | "47.8 million") (foodborne | "food borne")
54 records, 41 true documents. Written before counting. Misses only 2017-06869.
S12a
"1 in six" (foodborne | "food borne" | "food-borne")
1 record, 1 true document. Returns 2017-06869 alone. That document first surfaced through a collection-title search.
S12b
"one in 6" (foodborne | "food borne" | "food-borne")
0 records, 0 true documents.
S10
Subject-free phrases such as "one in six Americans" and "48 million people", then read for food
71 records. Added no claim document.
S12c
"50 million" (foodborne | "food borne" | "food-borne")
45 records. Added no claim document.
Marker searches
"47.8 million" with the food words
7 records, 4 true. The four EHS-Net notices.
"38.4 million"
0 records, 0 true.
"71.1 million"
0 records, 0 true.
"28.7 million"
1 record, 0 true. An FSIS poultry proposal, which is not a claim document.
"9.4 million"
7 records, 1 true. 2013-00125.
"9.9 million"
5 records, 0 true. None is about food burden. The Register carries no 9.9 million for this subject in the searches run.
"credible interval"
4 records, 0 true. Two are FSIS uses on E. coli O157:H7, not the total.
"unspecified agents"
3 records, 1 true. 2013-00125 in claim text.
"299 million" and "2006 population"
3 records, 0 true. No claim document prints either, by search and by reading.
The 39 weeks with a document
| Week of | Items | Documents | Numbers |
|---|---|---|---|
| 2011-05-02 | 2 | 2 | 2011-10953, 2011-10955 |
| 2011-08-01 | 1 | 1 | 2011-19333 |
| 2012-02-13 | 1 | 1 | 2012-1620 |
| 2012-02-20 | 1 | 1 | 2012-4165 |
| 2012-07-23 | 1 | 1 | 2012-17982 |
| 2012-08-27 | 1 | 1 | 2012-21312 |
| 2012-10-08 | 1 | 1 | 2012-24757 |
| 2013-01-07 | 1 | 1 | 2012-31480 |
| 2013-01-14 | 2 | 2 | 2013-00123, 2013-00125 |
| 2013-02-04 | 1 | 1 | 2013-02497 |
| 2013-05-27 | 1 | 1 | 2013-12833 |
| 2013-07-29 | 2 | 2 | 2013-17993, 2013-17994 |
| 2014-03-31 | 1 | 1 | 2014-07550 |
| 2014-09-22 | 1 | 1 | 2014-22682 |
| 2014-12-22 | 1 | 1 | 2014-28927 |
| 2015-06-15 | 1 | 1 | 2015-14709 |
| 2015-08-24 | 0 | 1 | 2015-20811 |
| 2015-09-14 | 1 | 1 | 2015-21920 |
| 2015-12-14 | 1 | 1 | 2015-30690 |
| 2016-06-20 | 1 | 1 | 2016-14727 |
| 2016-08-22 | 0 | 1 | 2016-20333 |
| 2016-10-10 | 1 | 1 | 2016-24668 |
| 2016-12-19 | 1 | 1 | 2016-30778 |
| 2017-04-03 | 0 | 1 | 2017-06869 |
| 2018-08-06 | 1 | 1 | 2018-17176 |
| 2018-12-17 | 0 | 1 | 2018-27223 |
| 2019-10-14 | 1 | 1 | 2019-22735 |
| 2020-02-17 | 0 | 1 | 2020-03342 |
| 2020-02-24 | 1 | 1 | 2020-03656 |
| 2020-06-01 | 0 | 1 | 2020-11798 |
| 2021-04-05 | 1 | 1 | 2021-06883 |
| 2021-08-09 | 0 | 1 | 2021-17345 |
| 2023-03-06 | 1 | 1 | 2023-04971 |
| 2023-06-26 | 0 | 1 | 2023-13568 |
| 2023-07-03 | 1 | 1 | 2023-14414 |
| 2023-09-11 | 0 | 1 | 2023-19708 |
| 2024-08-05 | 1 | 1 | 2024-17765 |
| 2024-11-25 | 0 | 1 | 2024-27469 |
| 2026-06-01 | 1 | 1 | 2026-11011 |
Every week not listed is 0. An item sits in the week of its first document.
Documents and items by year
- 2011
- 3 documents, 3 items
- 2012
- 5 documents, 5 items
- 2013
- 7 documents, 7 items
- 2014
- 3 documents, 3 items
- 2015
- 4 documents, 3 items
- 2016
- 4 documents, 3 items
- 2017
- 1 document, 0 items
- 2018
- 2 documents, 1 item
- 2019
- 1 document, 1 item
- 2020
- 3 documents, 1 item
- 2021
- 2 documents, 1 item
- 2022
- 0 documents, 0 items
- 2023
- 4 documents, 2 items
- 2024
- 2 documents, 1 item
- 2025
- 0 documents, 0 items
- 2026
- 1 document, 1 item
Observed for the counts. A separate count that is never added to the Register count.
GovInfo, counted apart
("1 in 6" OR "one in six" OR "48 million" OR "1 out of 6" OR "47.8 million") AND (foodborne OR "food borne")
GovInfo (govinfo.gov), public search page, after the API demo key returned 429 with retry-after 9100 seconds, November 6, 2026. 136 non-Register records dated from 2010-12-15 were logged and 81 were read: 46 hearings, 27 daily Record, 5 reports and 3 bills.
Treat this as a floor. The same search returned only 11 of the 42 known Register claim documents, so its recall is weak, and hearings post late, so recent years are under-counted. GovInfo’s GAO collection stops in 2008, so it gives no GAO count. No CRS record was among the 479, so CRS use is unknown. Speaker roles were not all checked, and no finding rests on them.
- Records
- 479
- Claim documents
- 43: hearings 21, daily Record 19, committee reports 3, bills 0
- Folded items
- 40, in 36 of 829 weeks
- First and latest
- 2010-12-21 and 2023-06-13
Congressional Record, House floor: “Each year, 48 million Americans are sickened from consuming contaminated food”
Observed
The lines in full
Earliest observed source does not establish original authorship.
About 48 million people (1 in 6 Americans) get sick, 128,000 are hospitalized, and 3,000 die each year from foodborne diseases, according new estimates from the Centers for Disease Control and Prevention.
- Unit
- no
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: new estimates from the Centers for Disease Control and Prevention
2011-01 (online 2010-12-15)
Combining the estimates for unspecified agents and major known pathogens provides an estimate of the total effect of contaminated food consumed in the United States: 47.8 million episodes of illness, 127,839 hospitalizations, and 3,037 deaths (Table).
Unspecified Agents paper, Emerging Infectious Diseases 17(1):16-22, CDC authors (archived copy). See also Table page with the total row
- Unit
- yes
- Data years
- yes
- Publication year
- no
- Source
- yes
Source given: this paper
Each year about 48 million people (1 in 6 Americans) are sickened, 128,000 are hospitalized, and 3,000 die from food borne diseases, according to recent data from the Centers for Disease Control and Prevention.
FDA interim final rule, Federal Register 2011-10953 (archived copy)
- Unit
- no
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: recent data from the Centers for Disease Control and Prevention
CDC estimates that each year approximately 48 million illnesses, 128,000 hospitalizations, and 3,000 deaths are food related (Ref. 45) (Ref. 46). These numbers include all illnesses that CDC estimates are attributable to food, including those illnesses caused by unspecified agents.
FDA proposed preventive controls rule, Federal Register 2013-00125
- Unit
- yes
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: CDC estimates, (Ref. 45) (Ref. 46)
Each year about 48 million people (1 in 6 Americans) get sick; 128,000 are hospitalized; and 3,000 die from food borne diseases, according to 2011 data from the Centers for Disease Control and Prevention
FDA final rule, Federal Register 2013-12833
- Unit
- no
- Data years
- no
- Publication year
- yes
- Source
- yes
Source given: 2011 data from the Centers for Disease Control and Prevention
Content current as of 2024-02-05
About 48 million people in the U.S. (1 in 6) get sick, 128,000 are hospitalized, and 3,000 die each year from foodborne diseases, according to recent data from the Centers for Disease Control and Prevention.
FDA, Food Safety Modernization Act page
- Unit
- no
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: recent data from the Centers for Disease Control and Prevention
An estimated 47.8 million foodborne illnesses occur annually in the United States, resulting in 127,839 hospitalizations, and 3,037 deaths annually.
CDC 30-day notice, Federal Register 2024-27469, OMB 0920-0792
- Unit
- yes
- Data years
- no
- Publication year
- no
- Source
- no
Source given: none
Each year, foodborne illnesses sicken one in six Americans, and thousands die, according to CDC’s most recent estimates.
GAO-25-107613, report to congressional addressees, Highlights, “Why GAO Did This Study”
- Unit
- no
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: CDC’s most recent estimates
Although the U.S. food supply is generally considered safe, tens of millions of Americans get sick from foodborne illness every year, and some die, according to estimates from the Department of Health and Human Services’ (HHS) Centers for Disease Control and Prevention (CDC).
GAO-25-107606, Food Safety: Status of Foodborne Illness in the U.S., p. 1 (archived copy)
- Unit
- no
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: estimates from the Department of Health and Human Services’ (HHS) Centers for Disease Control and Prevention (CDC)
Page dated 2025-03-19
The major pathogens, excluding Toxoplasma gondii, caused an estimated 9.9 million domestically acquired foodborne illnesses.
CDC, Estimates: Burden of Foodborne Illness in the United States (archived copy)
- Unit
- yes
- Data years
- yes
- Publication year
- no
- Source
- yes
Source given: Major pathogens, 2019
Content current as of 2025-08-25
The Federal government estimates that there are about 48 million cases of foodborne illness annually – that’s about 1 in 6 Americans each year.
FDA, Food Safety Education Month page (archived copy). See also FDA, People at Risk of Foodborne Illness
- Unit
- yes
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: The Federal government
Page dated 2025-11-24
CDC estimates that 48 million people get sick from a foodborne illness every year.
CDC, Facts About Food Poisoning (archived copy). See also Capture before the edit. See also Capture after the edit
- Unit
- no
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: CDC estimates
FY 2027 justification. File created 2026-04-02, first capture 2026-04-07
Each year, 48 million people – one in six Americans – experience foodborne illness, resulting in approximately 128,000 hospitalizations and 3,000 deaths.
CDC FY 2027 Congressional Justification, p. 48 (archived copy)
- Unit
- no
- Data years
- no
- Publication year
- no
- Source
- no
Source given: none
Last Updated 2026-04-27
The Centers for Disease Control and Prevention (CDC) estimates that each year roughly 1 in 6 Americans (or 48 million people) get sick, 128,000 are hospitalized, and 3,000 die of foodborne diseases.
Virginia Department of Health, Foodborne Diseases and Outbreaks (archived copy)
- Unit
- no
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: The Centers for Disease Control and Prevention (CDC) estimates
An estimated one in six Americans each year becomes ill with a foodborne disease.
CDC 60-day notice, Federal Register 2026-11011, OMB 0920-1198 (archived copy)
- Unit
- no
- Data years
- no
- Publication year
- no
- Source
- no
Source given: none
Page dated 2026-08-18, first seen in a capture of 2026-08-31
CDC estimates that each year in the Untied States, the most common foodborne pathogens cause an estimated 9.9 million illnesses, 53,000 hospitalizations, and 931 deaths.
CDC, About Restaurant Food Safety, National Center for Environmental Health (archived copy). See also Capture before the swap. See also USDA ERS, Cost Estimates of Foodborne Illnesses
- Unit
- yes
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: CDC estimates
2026-09-17 release. File created 2026-08-27
Scallan et al. (2011a,b) estimated that foodborne diseases cause approximately 48 million illnesses, 128,000 hospitalizations, and 3,000 deaths in the United States each year.
FDA Food Code 2026, Preface, p. v (archived copy). See also FDA Food Code 2022
- Unit
- yes
- Data years
- no
- Publication year
- yes
- Source
- yes
Source given: Scallan et al. (2011a,b)
Page metadata modified 2026-10-26
The CDC estimates that each year roughly 1 out of 6 Americans (or 48 million people) get sick, 128,000 are hospitalized, and 3,000 die of foodborne diseases.
Northeast Colorado Health Department, Consumer Food Safety (archived copy). See also Capture of 2025-03-20
- Unit
- no
- Data years
- no
- Publication year
- no
- Source
- yes
Source given: The CDC estimates
Observed, except the sentences marked Assessed
Where the number sits
As of November 6, 2026
The texts read use the 2011 estimates in two ways. The sentence appears in FDA rule preambles, CDC paperwork notices, state health pages and CDC’s FY 2027 budget justification. Its parts are inputs: USDA’s cost figures use the case counts, and FDA’s preventive controls analyses used pathogen counts and the 80 percent split. Separately, a reader who wants one national figure that includes unspecified agents has no later CDC total to reach for, as far as was found.
GAO reports that CDC and its partners use such estimates “to help set public health goals, allocate resources, and measure the economic impacts of disease.” GAO also reports CDC saying these periodic estimates give the most accurate picture of the magnitude of foodborne illnesses “for the pathogens responsible for the most illnesses at the time of the study.” GAO is describing the periodic studies in general there, not the 48 million alone.
GAO-25-107606, p. 10 (archived copy)
Assessed: in the rule texts read the sentence gives scale and context. Parts of the estimate entered FDA’s burden figures, which FDA says are not the rule’s expected benefits. No rule outcome was shown to turn on the number.
What the texts read show about use
- FDA’s final analysis rests on a break-even, and it says of its burden figures “these figures are not the expected benefits associated with the provisions in this rule”. It uses pathogen counts and the 80 percent, and it does not use the 48 million as a multiplier. It states that the regulation is mandated by statute.
- FDA’s Response 213 opens with the 48 million sentence, calls the burden significant and largely preventable, then declines further exemptions. The text does not say how much weight the sentence carried. The same passage excludes illnesses from home and food service handling, “which would not be addressed by this rule”.
- Rule preambles and paperwork notices in the audited set use the sentence as background. Nothing in their text shows a rule or a collection turning on it.
- CDC’s 2019 figures give a total for six named pathogens only, and they are not derived from the 2011 total. CDC publishes surveillance data every year on its pathogen pages.
FDA final impact analysis. FDA 2015 rule, Federal Register 2015-21920, Response 213
USDA Economic Research Service cost estimates
Rest on the 2011 estimates: 47,780,837 mean cases and $74.7 billion in 2023 dollars. ERS says the 80 percent of cases in other foodborne gastroenteritis are 40 percent of cost. CDC’s restaurant page now sets this cost beside the 9.9 million. That page’s cost line changed from $17.6 billion to USDA’s $74.7 billion between captures of 2026-08-08 and 2026-08-31, the window of the swap.
FDA preliminary impact analysis, preventive controls, with 78 FR 3646
FDA’s proposed analysis applied a 2.1 percent share to about 39 million illnesses from unknown pathogens, giving 806,247 of about 917,118 illnesses, printed a lower alternative of 443,484 and asked for comment. It states that the regulation is required by the FDA Food Safety Modernization Act.
FDA final impact analysis, with 80 FR 55907
Multiplies identified illnesses by 4, citing the 80 percent. Table 4 gives 172,019 identified, 688,064 unidentified. The text has no “48 million” or “47.8”. It states that the regulation is mandated by statute.
The 2022 and 2026 Prefaces carry the count. The Preface is background to a model code. Colorado incorporates the 2022 Food Code by reference and a Connecticut town health department posts a copy. Whether any jurisdiction’s adoption takes in the Preface was not traced, so this file does not say the sentence is in any law.
Colorado Board of Health memo, 2024-01-17. Town of Wallingford, Connecticut, posted copy
A Colorado Board of Health memo of 2024-01-17 and a Washington analysis of July 2020 use the sentence as background to adopting the Food Code. Whether any calculation depends on it was not traced.
States the ratio on the foodborne illness objectives page, citing a CDC page that prints only the count.
foodsafety.gov, managed by HHS
A post says CDC released on March 19 “the latest estimates providing the most accurate picture yet of foodborne illnesses in the United States” and gives 9.9 million for six germs. It prints no year and no 48 million in the text read. The same site’s Food Poisoning page carries the 48 million sentence in a capture of 2026-10-24.
Observed for the quoted wordings.
Wordings already in federal text
No CDC instruction on which figure to cite was found. These sentences are quoted from CDC and FDA as they stand. They are not guidance, and this file does not recommend any of them.
Total including unspecified agents, with its condition
“These numbers include all illnesses that CDC estimates are attributable to food, including those illnesses caused by unspecified agents.”
Total including unspecified agents, attributed to the papers
“estimated that foodborne diseases cause approximately 48 million illnesses, 128,000 hospitalizations, and 3,000 deaths in the United States each year.”
Named pathogens, circa 2019
“The major pathogens, excluding Toxoplasma gondii, caused an estimated 9.9 million domestically acquired foodborne illnesses.”
On comparing the two sets
“We cannot meaningfully compare the estimates from 2019 with the ones published in 2011.”
Estimates elsewhere, for context
Observed. Context only, and no country here is offered as a model. Rates should not be compared across countries.
WHO, 2026 edition of its foodborne disease estimates
United States, 2021, all ages, all hazards: 9,919,553 illnesses (5,548,535 to 19,114,298). Modelled, with 42 named hazards and no unspecified category in its codebook. It is not a CDC figure and it does not use CDC’s method. It sits close to CDC’s 9.9 million with a different scope and method, and SOVRINT labels the two not comparable. So this file says CDC’s only total that includes unspecified agents.
Public Health Agency of Canada
About 4 million a year (1 in 8), called the most recent average yearly estimates, on a page modified 2016-07-05. A companion page says 2000 to 2010 data. Read only through a summary tool. A figure of similar age also sits in the present tense there.
A 2025 paper says the 2018 update gave about 2.4 million cases a year, and that data from September 2023 to August 2025 will be used to update those estimates.
Observed for the items. Assessed for what this file adds.
Who said what first
GAO set out most of this file’s centre on February 3, 2025, and CDC said “centering around 2006” on the first day. A commercial page stated it in one sentence on September 4, 2026. This file does not claim to be first.
2010-12-15
CDC’s division lead, press briefing (archived copy)
“the data that is used to estimate the illnesses actually spanned a number of years centering around 2006.”
2025-02-03
Figure 5 gives the 2011 data range as 2000-2008, says of 2025 “unspecified agents not estimated”, and warns the estimates cannot be compared. The text reports CDC officials saying the 2025 report omits unspecified agents and gives CDC’s reason. This is the main prior official text.
2025-04
2025-12-03
USDA Economic Research Service, documentation
States that all three versions of its cost estimates rest on the case, hospitalization and death estimates CDC published in 2011, because those were the most recent for all 31 pathogens and unspecified agents when the product was developed.
2026-09-04
A commercial inspection firm’s statistics page
Calls the figure a 2011 estimate built on circa-2006 data and CDC’s only all-cause number, and rejects reading 48 million to 9.9 million as a fall. It states the file’s centre in plain words.
2026-09-08
A food-safety litigation blog
Says four-fifths of the figure was unspecified agents the new paper did not re-estimate, and argues from that for the opposite gloss: the consumer-page figure is “not out of date”. An interested party.
Assessed
What this file adds
- A dated, folded audit of 42 Register documents showing what sits beside the number, including the last printing found in each collection.
- Table A1’s data years with the inputs for unspecified agents and the appendix inputs set beside them, with SOVRINT’s share arithmetic.
- Dated change windows on CDC’s own pages, including the May 2026 source line and the August 2026 swap.
- The wording of the sentence in CDC’s FY 2024 to FY 2027 justifications.
- A partial list, from the texts read, of which documents print the total and which used its parts.
Observed
Fact-checks, and an earlier remark
- About a dozen wordings (outdated, not updated, decade old, 2000-2008 data, 2006 population, fact check, unspecified agents not re-estimated) and a Europe PMC full-text query found no fact-check from PolitiFact, FactCheck.org or similar, no journal comment on the age and no news article on it. The tools are not exhaustive, so this depends on those searches, and 23 Europe PMC hits were judged by title only.
- A personal newsletter of 2022-09-21 is the earliest outside remark on the age found. It is not linked here, because the file names no private individuals.
SOVRINT arithmetic
Arithmetic
SOVRINT arithmetic. None of this is a CDC figure, and none of it is an estimate of illness.
About 1 in 6.3 (6.26)
The ratio on the paper’s own base
299,000,000 divided by 47,780,779
1 in 4.2 to 1 in 10.4
The interval on the same base
299,000,000 divided by 71,133,833 and by 28,658,973
They add exactly to 47,780,779, 127,839 and 3,037
Whether the table’s means add
9,388,075 + 38,392,704, with the same for hospitalizations and deaths
80.4 percent + 11.4 percent = 91.8 percent
Share of the total tied to the FoodNet survey rate
38,392,704 and 5,461,731 (norovirus), each over 47,780,779
Unspecified Agents paper, table. Major Pathogens paper, Table 2
At least 17.8 years, at least 18.8 years, and 20 years by calendar year
Age of the inputs on the retrieval date
2008-12-31, 2007-12-31 and calendar 2006, each to 2026-11-06. The year ends are assumptions, because the papers give years only
14.3 years to CDC’s later paper, 15.9 years to today
Time from CDC’s press release
2010-12-15 to 2025-03-19, and to 2026-11-06
December 2015 to December 2020
Up 9.3 percent and up 14.3 percent, about 42.8 million
Growth in the population base
299 million to 326,763,427 (2017–2019 average) and to 341,784,857 (July 1, 2025)
76 percent of illnesses, 13 percent of dollars
Unidentified illnesses in FDA’s Table 6
688,064 of 904,526 illnesses, and $295 million of $2,237 million
The papers state 299 million and cite a Census page accessed 2010 Mar 9. Which vintage they used is not stated. Census gives 298,593,212 for July 1, 2006 in the series then current, 299,398 thousand in the Vintage 2006 table, and 298,379,912 in the later intercensal series. So this file says the paper states 299 million.
Observed
Copies
14 groups of wording. Copies in a group add no source.
The 14 groups of copies
2015-06-16
It is estimated that each year, 48 million Americans (1 in 6) become ill, 128,000 are hospitalized, and 3,000 die as the result of a foodborne illness.
One Register item, dated by its 60-day notice (0920-1112)
2016-06-22
It is estimated that each year roughly 1 in 6 Americans gets sick, 128,000 are hospitalized, and 3,000 die of foodborne diseases.
One Register item, dated by its 60-day notice (0920-0997)
2016-12-22
An estimated 1 in 6 Americans per year become ill with a foodborne disease.
One Register item, dated by its 60-day notice (0920-1198)
2018-08-10
It is estimated that each year, 48 million Americans (1 in 6) become ill, 128,000 are hospitalized, and 3,000 die as the result of a foodborne illness.
One Register item, dated by its 60-day notice (0920-1112)
2019-10-18
It is estimated that each year roughly one in six Americans get sick, 128,000 are hospitalized, and 3,000 die of foodborne diseases.
One Register item, dated by its 60-day notice (0920-0997)
2020-02-25
An estimated one in six Americans per year becomes ill with a foodborne disease.
One Register item, dated by its 60-day notice (0920-1198)
2021-04-05
An estimated 47.8 million foodborne illnesses occur annually in the United States, resulting in 127,839 hospitalizations, and 3,037 deaths annually.
One Register item, dated by its 60-day notice (0920-0792)
2023-03-10
It is estimated that each year roughly one in six Americans get sick, 128,000 are hospitalized, and 3,000 die of foodborne diseases.
One Register item, dated by its 60-day notice (0920-0997)
2023-07-07
An estimated one in six Americans per year becomes ill with a foodborne disease.
One Register item, dated by its 60-day notice (0920-1198)
2024-08-09
An estimated 47.8 million foodborne illnesses occur annually in the United States, resulting in 127,839 hospitalizations, and 3,037 deaths annually.
One Register item, dated by its 60-day notice (0920-0792)
2022 edition
Scallan et al. (2011a,b) estimated that foodborne diseases cause approximately 48 million illnesses, 128,000 hospitalizations, and 3,000 deaths in the United States each year.
- FDA Food Code 2022
- FDA Food Code 2026
- Town of Wallingford, Connecticut, posted copy of the 2022 Food Code
One FDA sentence, outside the Register count
Last reviewed 2018-11-05 on the archived CDC page
CDC estimates that each year roughly 1 in 6 Americans (or 48 million people) gets sick, 128,000 are hospitalized, and 3,000 die of foodborne diseases.
Content current as of 2022-02-17 on the earliest FDA page read
The Federal government estimates that there are about 48 million cases of foodborne illness annually
- FDA, What You Need to Know about Foodborne Illnesses, the earliest FDA page read
- FDA, People at Risk of Foodborne Illness
- FDA, Food Safety Education Month
- foodsafety.gov, Food Poisoning (archived copy)
- Missouri Department of Health and Senior Services
- Central District Health Department
One FDA sentence, outside the Register count
2012 to 2014
Three GovInfo hearing packages are reprints of another package.
- CHRG-112shrg72333 into CHRG-112shrg29104493
- CHRG-113shrg69104727 into CHRG-113shrg24504
- CHRG-113shrg78042 into CHRG-113shrg39104568
Three folds in the separate GovInfo count, 43 documents to 40 items
Observed
Limits
- CDC’s current pages were read from existing Wayback captures, the latest from October 2026, because www.cdc.gov returned 403 to curl. A summary tool read the Facts, Food Safety Basics, restaurant and burden pages on 2026-11-06 and showed the same sentences. No person has opened the live pages.
- Every “none” and “only one” is scoped to the written searches and the audited set. Register documents in another unusual wording cannot be ruled out, as 2017-06869 showed. About 20 CDC pages were checked, not all of them, so other CDC pages may print the ratio or 9.9 million.
- No reason was found for any change, including 2026-20906, the Facts page edit, the restaurant page swap and GAO’s change of wording. CDC, FDA and GAO have not been asked. A page’s printed date is not a log of its edits, so change dates here are windows between captures.
The other 10 limits
- Major Technical Appendices 1 and 3 of the 2011 work were searched and read in part, not line by line. Each pathogen’s foodborne share was taken from Table 2 and spot reads. The 2025 appendices were read in part.
- The 2018-2019 FoodNet Population Survey paper, American Journal of Epidemiology 194(1):5-11, was read by its abstract only, through PubMed. The journal page returned 403, PubMed Central holds no copy and no archived copy was found. The abstract prints no gastroenteritis rate and no total, so whether the paper reports a rate is not known here.
- The briefing’s audio was not heard. Quotes come from CDC’s posted transcript, which has typos, through a capture of 2010-12-24. It is not on archive.cdc.gov at the two addresses tried.
- The earliest-sighting search was three web searches. Paid trade press was not read, CDC’s 2010 fact sheets were not opened as PDFs, and exact strings of the web searches were not logged.
- The state and local set is a sample from searches, most pages are undated, and metadata dates may be site-wide. No series was made for it.
- The GovInfo count is a floor and stands apart. The API key returned 429. 55 logged records were not read, speaker roles were not all checked, and the three passages used in the drift rows were read at the passage only.
- Refused with 403: fsis.usda.gov, foodsafety.gov, osha.gov, gao.gov assets, weld.gov, tpchd.org and www.cdc.gov. Public quotes from these sites rest on existing captures. A fetch tool was also used after a 403 on some of the same addresses: as a summary cross-check on CDC’s Facts, Food Safety Basics, restaurant and burden pages, and to retrieve GAO-25-107606 and CDC’s FY 2026 and FY 2027 justifications, whose text was then read with pdftotext. The fetch tool failed on the foodsafety.gov post, which was read from a capture. federalregister.gov full-text pages returned 429, so GPO’s official copy was used. The Canada pages came through a summary tool only.
- Cost work was read in part: the Foodborne Pathogens and Disease 2025 paper by abstract only, with a corrigendum of 2026-09-16 unread, so the 38 and 40 percent gap is not explained. FDA produce safety and FSIS impact analyses were not read.
- No CRS report was found in one search, and congress.gov was not searched directly, so CRS use is unknown. No later fact-check was found, which depends on the searches named under prior work.
- CDC’s notice 2026-20906 says it received four public comments on the notice of 2026-06-02, docket CDC-2026-0925. Those comments were not read, because the regulations.gov API returned a rate limit. Nothing here says what they contain.
Not established
Other explanations
Each of these fits the same pattern, and none is chosen. The order carries no weight, and no one’s intent is stated.
Not established
It remains the only CDC total that includes unspecified agents, so a writer who wants one national figure has nothing later from CDC to use.
CDC’s 2019 work gives no total that includes unspecified agents, and GAO reports the 2025 study has no unspecified agents. ERS says it built on 2011 because those were the most recent estimates for all 31 pathogens and unspecified agents.
Not established
Renewal clocks repeat old background text.
Since 2015-12-15 the audited set is all CDC paperwork notices, and sentence families follow the collection across cycles. One family runs from 2012 to 2023. The one notice found without the sentence, 2026-20906, is a filing that changed from Extension to Revision, and the notice links neither to the sentence.
Not established
Texts copy earlier texts.
The Food Code sentence is the same in 2022 and 2026. State and local pages keep CDC’s or FDA’s wording, including CDC’s “1 out of 6” form of December 2010.
Not established
CDC’s own pages print it, so citing it is citing CDC.
The Facts page and Food Safety Basics print the count, and Food Safety Basics was re-dated twice with the sentence unchanged.
Not established
CDC says the estimates were not wrong, and no retirement statement was found.
CDC’s 2019 work has a different scope. CDC told GAO it considers the priority pathogen estimates more valuable to regulators than estimates for unspecified agents.
Not established
Pages have different owners, and no one instruction reaches them all.
The Facts page’s content source is NCEZID and the restaurant page’s is the National Center for Environmental Health. They now print different figures. No CDC text saying which to use was found.
Not established
Writers use it for scale, so its exact year matters little to them.
In the texts read it is background, and no rule outcome was shown to turn on it. GAO’s “tens of millions” gives scale with no number.
Assessed
Judgments
These are assessments of what the sources show now. None concerns anything still to come.
Assessed
Very likely
Confidence: moderateNo data year after 2008 entered the 2011 total.
Both papers state their ranges, Table A1 ends at 2008, and a search of the five appendices found no later data year. Two appendices were not read line by line.
Assessed
Very likely
Confidence: moderateAs of 2026-11-06 CDC has published no total including unspecified agents later than the 2011 estimates.
CDC’s 2025 pages and paper give none, GAO reports none was planned for that study, and no later total turned up in the searches. The live pages were read through captures, and about 20 CDC pages were checked.
Assessed
Very likely
Confidence: moderate“2011 data” in three FDA documents refers to the publication and not to the data years.
Each cites CDC’s 2011 estimates page, and CDC said on the first day that 2011 is the journal’s date. FDA has not been asked.
Assessed
Likely
Confidence: moderateIn the 42 Register documents the sentence is background, and it feeds no calculation in their text.
Each was read at its key sentence, and the preventive controls analyses show the parts of the estimate at work in separate documents. Other impact analyses were not read. FDA’s Response 213 opens its reasoning with the sentence, and the text does not say how much weight it carried.
Assessed
Likely
Confidence: lowThe Register’s weekly count mostly reflects paperwork and rulemaking calendars.
Notices arrive in 60-day and 30-day pairs about every three years, and 32 items sit in 29 of 810 weeks.
Status of the claims
- One in six Americans, or about 48 million people, gets sick from foodborne disease each year.
- Unverified
- Unverified for any year after the data window. Observed as CDC’s estimate on a 2006 base.
- The claim appears in the texts in the present tense. No CDC total that includes unspecified agents was found for any year after the data window. CDC’s 2019 work and its 2018-2019 survey are later records. The 2019 work gives a total for six named pathogens only, and the survey paper was read by its abstract only, which gives none. That is a statement about the record, and it does not say the number is false. The survey paper
- CDC estimated about 47.8 million illnesses a year on the 2006 population
- Observed
- Printed in the Unspecified Agents paper and its table. Observed describes the text.
- The number is “2011 data”
- Observed
- Both wordings are printed. The papers give the data years, and CDC’s briefing says 2011 is the journal’s date. True and false stay out of the file.
- Who first wrote 48 million and 1 in 6
- Not established
- Earliest observed source does not establish original authorship.
- Whether foodborne illness has risen or fallen since the data years
- Not established
- CDC says the sets cannot be meaningfully compared, and this file makes no comparison.
- Why CDC left 48 million on two pages, changed its restaurant page, or left the sentence out of 2026-20906
- Not established
- No stated reason was found.
- Why GAO changed its wording on 2025-02-03
- Not established
- No stated reason was found.
- Whether any rule or budget outcome turned on the number
- Not established
- None was shown in the texts read, and not all impact analyses were read.
What the status words cover
- Observed
- The quoted lines, printed text from the sources, the dated rows of inputs and changes, the quotes on how the wording moved, the series counts, the prior work items, the stakes except the sentences marked Assessed, the copies and the limits
- Assessed
- What the series can show, what this file adds, and the judgments marked Assessed
- Unverified
- The central claim for any year after the data window. Printed in full as: Unverified for any year after the data window. Observed as CDC’s estimate on a 2006 base.
- Not established
- First author, any rise or fall in illness, why any office kept, changed or dropped the sentence, and every entry under other explanations
- Projected
- None. This file makes no prediction.
Made by agents, unchecked
AI agents gathered the sources, read them, counted them and drafted this file, using scripts. No person has yet checked it or read a source. No SOVRINT software produced it.
Sources, search, counting and tools
- Sources
- CDC’s two 2011 papers with tables and appendices, the 1999 and 2025 papers, CDC’s press release and briefing transcript, CDC pages through archive.cdc.gov and Wayback captures, CDC’s budget justifications for FY 2024 to FY 2027, the Federal Register, GovInfo, GAO reports, FDA pages, rules, impact analyses and the Food Code, USDA ERS and FSIS pages, Census tables, and state and local pages linked in this file.
- Search
- Every Federal Register and GovInfo search is listed below with its date and record count, from the search protocol. Search strings and definitions were written before any count, and later additions are marked. Web searches for the first report, prior work, fact-checks and state and local use ran through a web search tool on 2026-11-06. Their exact strings were not logged, so those searches are described by target, and any “none found” resting on them is weaker for that.
- Counting
- Distinct items per calendar week, with copies folded first. Documents and folded items are reported side by side. The GovInfo count stands apart and is never added.
- Tools
- curl, pdftotext, the Federal Register public API, the GovInfo public search page, the Wayback availability and capture pages, a web search tool, a web fetch tool that returns a summary, and PubMed for one abstract. Anything read only through a summary is marked as such.
- Refusals
- Pages that returned 403, 404 or 429 were recorded, and public quotes from those sites rest on existing archived copies. No new captures were triggered. Where curl got 403, a fetch tool was also used on some of the same addresses as a cross-check or to retrieve a file. See limits for the list.
- Snapshot
- All sources were read on or before 2026-11-06. For CDC pages the capture date is the as-of date.
- Left out of scope
- Closed groups, private messages, anything behind a login, paid trade press, and the truth of the claim.
The 42 logged searches
- 2026-11-06, Federal Register API, S1: ("1 in 6" | "one in six" | "48 million" | "1 out of 6") (foodborne | "food borne") (47 records)
- 2026-11-06, Federal Register API, S2: ("1 in 6" | "one in six" | "48 million" | "1 out of 6" | "47.8 million") (foodborne | "food borne") (54 records)
- 2026-11-06, Federal Register API, S3: ("1 in 6" | "one in six" | "48 million" | "1 out of 6" | "47.8 million") (foodborne | "food borne" | "food-borne") (54 records)
- 2026-11-06, Federal Register API, S4: ("one out of six" | "one in every six" | "1 in every 6" | "one out of every six" | "1 out of every 6" | "one of every six" | "one-in-six" | "1-in-6" | "one sixth" | "one-sixth" | "1/6") (foodborne | "food borne" | "food-borne") (103 records)
- 2026-11-06, Federal Register API, S5: ("48,000,000" | "47,800,000" | "47,780,779" | "47.8" | "48 million people" | "48 million Americans" | "48 million illnesses" | "48 million cases" | "forty-eight million" | "47.8 million") (foodborne | "food borne" | "food-borne") (34 records)
- 2026-11-06, Federal Register API, S6: ("1 in 6" | "one in six" | "48 million" | "1 out of 6" | "47.8 million") ("food poisoning" | "food-related" | "food related" | "foodborne" | "contaminated food" | "illness from food" | "food safety") (70 records)
- 2026-11-06, Federal Register API, S7a: "128,000" (foodborne | "food borne" | "food-borne") (31 records)
- 2026-11-06, Federal Register API, S7b: "127,839" (9 records)
- 2026-11-06, Federal Register API, S7c: "3,037" (foodborne | "food borne" | "food-borne") (6 records)
- 2026-11-06, Federal Register API, S7d: "128,000" "3,000" food (49 records)
- 2026-11-06, Federal Register API, S8a: ("millions of Americans" | "tens of millions") (foodborne | "food borne" | "food-borne") (24 records)
- 2026-11-06, Federal Register API, S8b: "million Americans" (foodborne | "food borne" | "food-borne") (27 records)
- 2026-11-06, Federal Register API, S9a: "47.8 million" (foodborne | "food borne" | "food-borne") (7 records)
- 2026-11-06, Federal Register API, S9b: "38.4 million" (foodborne | "food borne" | "food-borne") (0 records)
- 2026-11-06, Federal Register API, S9c: "38 million" (foodborne | "food borne" | "food-borne") (12 records)
- 2026-11-06, Federal Register API, S9d: "9.4 million" (foodborne | "food borne" | "food-borne") (7 records)
- 2026-11-06, Federal Register API, S9e: "9.9 million" (foodborne | "food borne" | "food-borne") (5 records)
- 2026-11-06, Federal Register API, S9f: "71.1 million" (foodborne | "food borne" | "food-borne") (0 records)
- 2026-11-06, Federal Register API, S9g: "28.7 million" (foodborne | "food borne" | "food-borne") (1 record)
- 2026-11-06, Federal Register API, S9h: "credible interval" (foodborne | "food borne" | "food-borne") (4 records)
- 2026-11-06, Federal Register API, S9i: "unspecified agents" (foodborne | "food borne" | "food-borne") (3 records)
- 2026-11-06, Federal Register API, S9j: Scallan (32 records)
- 2026-11-06, Federal Register API, S9k: "Foodborne illness acquired in the United States" (24 records)
- 2026-11-06, Federal Register API, S9l: "299 million" (foodborne | "food borne" | "food-borne") (3 records)
- 2026-11-06, Federal Register API, S9m: "2006 population" (foodborne | "food borne" | "food-borne") (0 records)
- 2026-11-06, Federal Register API, S9n: "76 million" (foodborne | "food borne" | "food-borne") (16 records)
- 2026-11-06, Federal Register API, S10: ("one in six Americans" | "1 in 6 Americans" | "one in six people" | "1 in 6 people" | "one out of six Americans" | "1 out of 6 Americans" | "48 million Americans" | "48 million people" | "48 million illnesses" | "47.8 million illnesses" | "48 million cases" | "48 million foodborne") (71 records)
- 2026-11-06, Federal Register API, S11a: "0920-0997" (9 records)
- 2026-11-06, Federal Register API, S11b: "0920-1112" (2 records)
- 2026-11-06, Federal Register API, S11c: "0920-1198" (6 records)
- 2026-11-06, Federal Register API, S11d: "0920-0792" (11 records)
- 2026-11-06, Federal Register API, S11e: "National Hypothesis Generating Questionnaire" (20 records)
- 2026-11-06, Federal Register API, S11f: "FoodNet Population Survey" (5 records)
- 2026-11-06, Federal Register API, S11g: "Environmental Health Specialists Network" (40 records)
- 2026-11-06, Federal Register API, S12: ("1 in six" | "one in 6" | "1 out of six" | "one out of 6" | "1 of 6" | "1 of six" | "1 in every six" | "one in every 6" | "48-million" | "nearly 48 million" | "about 48 million" | "roughly 48 million" | "over 48 million" | "more than 48 million" | "50 million") (foodborne | "food borne" | "food-borne") (64 records)
- 2026-11-06, Federal Register API, S12a: "1 in six" (foodborne | "food borne" | "food-borne") (1 record)
- 2026-11-06, Federal Register API, S12b: "one in 6" (foodborne | "food borne" | "food-borne") (0 records)
- 2026-11-06, Federal Register API, S12c: "50 million" (foodborne | "food borne" | "food-borne") (45 records)
- 2026-11-06, Federal Register API, S13: ("1 in six Americans" | "one in 6 Americans" | "1 in six people" | "one in 6 people") (1 record)
- 2026-11-06, Federal Register API, SFINAL: ("1 in 6" | "one in six" | "1 in six" | "48 million" | "1 out of 6" | "47.8 million") (foodborne | "food borne" | "food-borne") (55 records)
- 2026-11-06, api.govinfo.gov DEMO_KEY search: ("1 in 6" OR "one in six" OR "48 million" OR "1 out of 6" OR "47.8 million") AND (foodborne OR "food borne") (HTTP 429 OVER_RATE_LIMIT)
- 2026-11-06, govinfo.gov public search: the same words (479 records, all collections)
Corrections
None so far.