Prevalence and Correlates of Perceived Food Noise Among U.S. Adults in a Nationally Representative Sample

ABSTRACT

Background

Food noise is a term that has recently emerged in the media, social media, and public discourse to describe people's experiences around food‐related thoughts. Despite recent efforts to define and measure food noise and the abundance of online content on the topic, the proportion of US adults who believe they experience it remains unknown.

Objective

To estimate the prevalence and demographic and behavioral correlates of perceived food noise experiences in a nationally representative sample of U.S. adults.

Methods

Data were drawn from Verasight's SBM Omnibus Survey (May 2026; N = 1000). Perceived experiences of “food noise” were assessed using a single item and categorized as frequent (often/always) versus infrequent. Weighted analyses examined differences across demographic groups, GLP‐1 medication use, and eating‐ and weight‐related behaviors.

Results

50.89% of respondents reported experiencing food noise at least occasionally and 17.98% frequently. Frequent endorsement was higher among female‐identifying respondents (22.70% vs. 12.68% in males, p < 0.001), younger adults (26.50%, p = 0.002), and Asian respondents (32.22%, p = 0.007). It was also more common among individuals reporting loss of control over eating (56.77%, p < 0.001), dietary restriction (27.84%, p < 0.001), recent weight loss attempts (24.33%, p < 0.001), and current and former GLP‐1 users (25.76% and 26.78% vs. 16.30% among non‐users, p = 0.009). Those reporting frequent food noise also reported poorer diet quality and mental health. In multivariable analyses, loss of control over eating (OR = 5.95), current GLP‐1 use (OR = 1.60), and female sex (OR = 1.47) were the strongest independent predictors.

Conclusions

Self‐reported experiences of “food noise” are common and associated with demographic and disordered eating–related factors, highlighting widespread identification with this emerging construct and the need for further research to define and understand its clinical significance.

1. Introduction

“Food noise” is a rapidly emerging term with increasing use in the media, social media, marketing, and patient‐clinician discussions to describe experiences of persistent, intrusive thoughts about food [1, 2, 3, 4]—particularly in the context of GLP‐1 receptor agonists (GLP‐1RAs), in which patients often report experiencing a reduction in their food noise in response to such medications [5, 6]. Although potentially related constructs, such as food preoccupation and food‐related intrusive thoughts (FRITs), have been previously described in the scientific literature [7, 8], the rise in use of the term “food noise” itself is recent, with Google search data indicating meaningful initial interest in early 2023, followed by rapid growth in online searches [9].

Given the recent emergence of the term in public discourse, the study of food noise as a distinct construct is still in its infancy. Therefore, the scientific definition of food noise is still debated [10, 11], with different authors providing varying hypothetical definitions; the earliest being “heightened and/or persistent manifestations of food cue reactivity, often leading to food‐related intrusive thoughts and maladaptive eating behaviors,” [1] followed by more recent definitions that have expanded upon it by adding elements such as being disruptive to daily life and making healthy behaviors difficult [2], potentially causing social, mental, and physical problems [12], and shifting mental focus from long‐term goals to immediate food‐related reward‐seeking [13]. Regardless of which (if any) of these working definitions proves most accurate in the long run, since they all raise valuable hypotheses yet to be tested, a more fundamental question remains unanswered: despite the massive presence of food‐noise‐related content on social media and even the recent use of the term by advertisers to market anti‐obesity medications and behavioral treatments that promise to “silence” the food noise [5], the proportion of US adults who believe they experience food noise hasn't been determined yet.

Different authors have been working on developing and testing psychometric tools to potentially measure levels of food noise, such as the Food Noise Questionnaire (FNQ) [2] and the Ro‐Allison‐Indiana‐Dhurandhar Food Noise Inventory (RAID‐FN) [12], both of which have been critiqued for their significant conceptual and semantic overlap with existing scales that examine well‐described constructs (namely food preoccupation, food addiction, and food cravings) [9, 11]. The FNQ and RAID‐FN were designed to quantify the severity of food noise as a clinical construct in individuals presumed to experience it at varying levels; however, their application of assessing food noise severity within a continuum in which any respondent can receive a score representing their food noise levels serves a fundamentally different goal from estimating how many adults in the general population self‐identify with the experience of food noise as depicted in lay terms by current media and social media discussions [1, 5]. Previous research has discussed how employing measures that use a construct's lay label directly and measures that assess it indirectly through symptom‐representing items answers different questions because, while scales estimate the underlying construct without requiring self‐labeling, direct items capture respondents' own recognition of the term or endorsement of it as a label for their experiences (e.g., asking participants directly “How often do you feel lonely?” addresses a different question than using scales that assess isolation and lack of companionship as measures of levels of loneliness without explicitly naming it) [14]. In the case of food noise, while the RAID‐FN and the FNQ answer the question “how much food noise does this respondent experience?” a lay‐language single‐item question framed consistently with how food noise is described in media captures whether a given respondent would use the label “food noise” to describe their experiences. Applying it to a representative sample can help estimate population‐level prevalence of self‐identification with this experience, capturing a snapshot of how prevalent endorsement of “food noise” is as a descriptor for one's perception of their thoughts about food. It asks not whether their food noise is high or low, but whether they think they experience it.

Learning the prevalence of the endorsement of experiences of food noise in the adult US population and its correlates is an important initial step in investigating it because it provides a glimpse beyond how common the term is in social media discourse or Google searches [5, 9], giving researchers, clinicians, and other stakeholders an idea of how likely people of different demographic groups are to report experiencing food noise, and which clinical, behavioral, attitudinal, and social factors are associated with the likelihood of endorsing such experiences. Those pieces of information are valuable for gauging the magnitude of food noise at the population level and the need to further investigate it. Therefore, findings are reported from a secondary analysis of cross‐sectional data from a representative sample of 1000 US adults exploring the proportion of respondents who believe they experience food noise and how this proportion varies across demographic, behavioral, attitudinal, and clinical characteristics. Correlates of interest available in the database are examined, particularly those with theoretical links to food‐related cognitions, including GLP‐1 medication use (given patient reports of reduced food noise with these medications) [6, 15], disordered eating behaviors (given their conceptual overlap with intrusive food thoughts) [1, 8], mental health (given the theoretical link between experiencing food noise and experiencing negative mental health outcomes) [3], and attitudinal variables related to eating behavior, including perceived genetic and social influences on eating and weight, as well as standard demographic characteristics.

2. Methods

2.1. Study Design and Data Source

This study is a secondary analysis of cross‐sectional data. Data were drawn from the Society of Behavioral Medicine (SBM) Omnibus Survey (study #2026‐049), an online survey administered by Verasight in May, 2026. Respondents were recruited via email from the Verasight Community, a panel built through random address‐based sampling, person‐to‐person text messaging, and dynamic online targeting. Eligibility was limited to U.S. adults aged 18+ who passed data quality checks. Community members were verified via multi‐step authentication, including SMS confirmation and bot detection (Google reCAPTCHA v3). Respondents showing low‐quality patterns (e.g., straight‐lining, speeding) were removed. Post‐collection checks confirmed U.S. IPs, removed duplicates, and excluded those failing attention checks or completing the survey in < 30% of median time. Verasight is a member of the AAPOR Transparency Initiative. Access to the dataset was provided to attendees of the 2026 SBM Meeting as a courtesy of Verasight.

Data were weighted to match the March 2026 Current Population Survey on age, race/ethnicity, sex, income, education, region, and metropolitan status, and to align with a three‐year average of partisanship from Pew NPORS surveys and 2024 presidential vote benchmarks. All analyses used these post‐stratification weights.

2.2. Measures

2.2.1. Food Noise

Assessed using a single item designed to capture respondents' beliefs of experiencing “food noise,” framing it similarly to how it has been portrayed in lay terms in media and social media reports: [1, 5] “‘Food Noise’ is a term popularized in media and social media to describe things like ‘constantly thinking about food, even when not hungry,’ or feeling like one's life revolves around food. How often do you experience ‘Food Noise’?” Responses ranged from 1 (Never) to 5 (Always). This measure captures self‐reported experiences consistent with widespread understandings of the term rather than a clinically or psychometrically validated construct. Given that the research question concerns population‐level prevalence of self‐identification with food noise as a lay concept, a single item framed in lay language was used rather than validated clinical scales such as the Food Noise Questionnaire (FNQ) and the RAID‐FN, which are designed to quantify food noise severity in clinical populations. A perceived frequency‐based answer scale was used because it captures whether someone believes they experience it (“never” vs. other answer options) and adds granularity for those who respond with options other than “never.” For descriptive analyses, responses were dichotomized into frequent (Often/Always) versus infrequent (Never/Rarely/Sometimes).

GLP‐1 medication use. Participants were asked: “Have you been taking a GLP‐1 medication (e.g., Ozempic, Mounjaro, Wegovy)?” Responses: (1) past week, (2) past month, (3) past 12 months but > 1 month ago, (4) not in past 12 months. These were grouped as current (week/month; n = 124), former (past year, not recently; n = 53), and non‐users (n = 823).

Weight and eating behaviors. Participants selected all that applied in the past month: loss of control eating, eating less to change weight/shape, vomiting, using diuretics/laxatives, or none. They also reported whether they had intentionally tried to lose weight in the past 12 months (yes/no).

Ordinal Likert‐scale correlates. Three variables used five‐point scales: Dietary quality: “How would you rate your healthy eating behaviors on a scale of 1–5, with 1 being poor and 5 being great?”; Social support: “When trying to change how I eat, support from others plays an important role” (1 = Strongly disagree, 5 = Strongly agree); Genetic influence: “Thinking about all of the things that influence a person's body weight–to what extent does a person's genetic makeup influence their weight?” (1 = Not at all, 5 = Extremely; “I'm not sure” excluded); Physical and mental health were each assessed with single items (1 = Excellent, 5 = Poor) and reverse‐coded so higher values indicate better health.

Demographics: Age was collected continuously (range: 18–89 years) and grouped into four bands (18–29, 30–44, 45–59, 60+) for descriptive presentation in Table 1; the continuous variable was used in multivariable analyses. Other variables included sex (male, female, other/non‐binary), education (high school or less; some college/associate; bachelor's+), income (< $15k; $15k–$74,999; $75k–$149,999; ≥ $150k), and race/ethnicity (White non‐Hispanic, Black non‐Hispanic, Hispanic, Asian non‐Hispanic, other/multiracial).

TABLE 1.

Sample characteristics and perceived food noise frequency distribution.

Characteristic	n	Never (%)	Rarely (%)	Sometimes (%)	Often (%)	Always (%)	Often/always (%)	Difference
Full sample (reference)	1000	17.03	32.07	32.91	11.87	6.11	17.98	—
Sex
Male	469	20.20	33.76	33.35	10.37	2.31	12.68	−5.30pp (p < 0.001 *** )
Female	522	14.01	30.66	32.64	13.10	9.60	22.70	+4.72pp (p < 0.001 *** )
Other	9	21.53	22.75	25.39	21.19	9.15	30.34	+12.36pp (p < 0.001 *** )
Age group
18–29	137	11.80	25.98	35.73	17.56	8.94	26.50	+8.52pp (p = 0.002 ** )
30–44	220	17.89	29.53	33.02	12.47	7.09	19.56	+1.58pp (p = 0.002 ** )
45–59	257	18.44	30.79	35.16	9.23	6.38	15.61	−2.37pp (p = 0.002 ** )
60+	386	18.61	38.59	29.32	10.01	3.48	13.49	−4.49pp (p = 0.002 ** )
Education
HS or less	300	19.94	32.53	31.99	8.56	6.98	15.54	−2.44pp (p = 0.3)
Some college or associate	277	16.40	30.83	33.11	13.18	6.47	19.65	+1.67pp (p = 0.3)
Bachelor or higher	423	14.57	32.48	33.70	14.25	4.99	19.24	+1.26pp (p = 0.3)
Household income
Less than $15k	87	28.41	15.75	37.20	13.18	5.46	18.64	+0.66pp (p = 0.53)
$15k–$74k	430	16.83	33.61	31.96	9.91	7.69	17.60	−0.38pp (p = 0.53)
$75k–$149k	341	14.99	34.82	33.64	12.57	3.98	16.55	−1.43pp (p = 0.53)
$150k or more	141	14.85	31.57	31.31	15.39	6.89	22.28	+4.30pp (p = 0.53)
Race/Ethnicity
White non‐Hispanic	669	15.36	35.59	33.21	11.32	4.52	15.84	−2.14pp (p = 0.007 ** )
Black non‐Hispanic	94	22.19	30.23	31.75	8.60	7.23	15.83	−2.15pp (p = 0.007 ** )
Hispanic	142	14.83	29.93	32.65	14.18	8.41	22.59	+4.61pp (p = 0.007 ** )
Asian non‐Hispanic	49	12.02	14.49	41.27	20.07	12.15	32.22	+14.24pp (p = 0.007 ** )
Other or multiracial	46	36.84	24.40	24.39	7.91	6.46	14.37	−3.61pp (p = 0.007 ** )
GLP‐1 medication use
Non‐user	823	18.36	32.90	32.45	10.54	5.76	16.30	−1.68pp (p = 0.009 ** )
Former user	53	13.81	20.01	39.40	20.80	5.98	26.78	+8.80pp (p = 0.009 ** )
Current user	124	9.06	32.20	32.98	17.02	8.74	25.76	+7.78pp (p = 0.009 ** )
Weight and eating behaviors (among endorsers, past month)
Tried to lose weight, past 12 months	557	13.59	28.59	33.49	14.85	9.48	24.33	+6.35pp (p < 0.001 *** )
Loss of control over eating	138	3.62	14.17	25.44	31.44	25.33	56.77	+38.79pp (p < 0.001 *** )
Eating less to change weight/shape	321	9.43	27.52	35.22	16.82	11.02	27.84	+9.86pp (p < 0.001 *** )
Vomiting to change weight/shape	23	18.88	15.57	40.95	14.06	10.53	24.59	+6.61pp (p = 0.4)
Laxative or diuretic use	24	10.63	13.32	36.40	32.98	6.66	39.64	+21.66pp (p = 0.005 ** )

Note: n, unweighted count. Percentages are weighted to match the U.S. adult population (age, sex, race/ethnicity, income, education, region, metropolitan status). Often/Always = combined weighted percent reporting experiencing food noise often or always. Difference = percentage‐point difference versus full‐sample rate of 17.98% (positive = higher than average, negative = lower). p‐values: for demographic variables, weighted chi‐square test of whether experienced food noise distribution differs across subgroups; for weight/eating behavior variables, weighted two‐proportion z‐test comparing endorsers versus non‐endorsers. For weight/eating behavior rows, n and percentages reflect endorsers only.

*p < 0.05.

^** p < 0.01.

^*** p < 0.001.

2.2.2. Statistical Analysis

Weighted frequencies and percentages were computed for categorical variables. For each subgroup, the weighted percentage reporting frequent food noise (Often/Always) and percentage‐point differences from the full sample were calculated. Differences across demographic groups were tested using weighted chi‐square tests; binary eating/weight behaviors used weighted two‐sample proportion z‐tests.

Differences in ordinal correlates between frequent and infrequent food noise groups were assessed using weighted Welch t‐tests. Multivariable associations were examined using a weighted proportional odds model (ordinal logistic regression) with perceived food noise frequency as the outcome, with age entered as a continuous variable and all other covariates as categorical. Given the exploratory nature, p‐values were unadjusted and should be interpreted as hypothesis‐generating. Analyses were conducted in R (v4.5) using tidyverse and MASS, with weights applied throughout.

3. Results

The sample included 1000 U.S. adults (Table 1). Most respondents identified as White non‐Hispanic (n = 669), and educational attainment and income varied widely across the sample. Approximately 12.40% reported current GLP‐1 use (n = 124), 5.30% were former users (n = 53), and 82.30% reported no GLP‐1 use in the past 12 months (n = 823). Overall, 17.98% of respondents reported experiencing food noise often or always (Table 1). The most common responses were “Rarely” (32.07%) and “Sometimes” (32.91%). 17.03% reported never experiencing food noise.

Perceived frequency of food noise differed significantly by sex (p < 0.001; Table 1), as women reported higher rates of frequent (often/always) food noise (22.70%) compared to men (12.68%), and by age group (p = 0.002), with younger adults showing the highest prevalence (18–29: 26.50%; 30–44: 19.56%) and older adults the lowest (45–59: 15.61%; 60+: 13.49%). Food noise frequency differed significantly by race/ethnicity (p = 0.007), with Asian non‐Hispanic respondents reporting the highest prevalence (32.22%) and Other/Multiracial respondents reporting the lowest prevalence (14.37%). Perceived food noise frequency differed significantly across GLP‐1 use groups (p = 0.009; Table 1; Figure 1). 25.76% of current users and 26.78% of former users reported frequently experiencing food noise (non‐users: 16.30%).

FIGURE 1.

Perceived food noise frequency by GLP‐1 medication use status. Weighted estimates, U.S. adults, Module A (N = 1000). Percentages are weighted to match U.S. adult population benchmarks. Current user = GLP‐1 use in the past week (n = 92) or past month (n = 32), combined n = 124. Former user = GLP‐1 use more than 1 month ago but within the past 12 months (n = 53). Non‐user = no GLP‐1 use in the past 12 months (n = 823).

Among endorsers of weight and eating‐related behaviors, loss of control over eating showed the strongest association with frequent food noise (56.77%; p < 0.001; Table 1). Frequent food noise was also significantly more common among those who reported eating less to change their weight or shape (27.84%; p < 0.001) and those who had attempted weight loss in the past 12 months (24.33%; p < 0.001).

As shown in Table 2, compared to those with infrequent food noise, respondents reporting frequent food noise rated their eating quality lower (mean 2.91 vs. 3.30, difference = −0.40, p < 0.001), reported greater importance of social support in dietary change (mean 3.69 vs. 3.31, difference = +0.38, p < 0.001), attributed greater influence to genetics on weight (mean 3.72 vs. 3.32, difference = +0.40, p < 0.001), and reported poorer mental health (mean 2.98 vs. 3.49, difference = −0.51, p < 0.001). Physical health differences were smaller and significant (mean: 3.04 vs. 3.31, difference = −0.27, p = 0.002).

TABLE 2.

Weighted mean (SE) of ordinal correlates with perceived food noise frequency group.

Variable	Never‐sometimes mean (SE)	Often/always mean (SE)	Difference	p value
Self‐rated eating quality (1 = poor, 5 = great)	3.30 (0.03)	2.91 (0.08)	−0.40	< 0.001 ***
Importance of social support in dietary change (1 = strongly disagree, 5 = strongly agree)	3.31 (0.04)	3.69 (0.08)	+0.38	< 0.001 ***
Perceived genetic influence on weight (1 = not at all, 5 = extremely)	3.32 (0.03)	3.72 (0.07)	+0.40	< 0.001 ***
Physical health (1 = poor, 5 = excellent)	3.31 (0.03)	3.04 (0.08)	−0.27	0.002 **
Mental health (1 = poor, 5 = excellent)	3.49 (0.04)	2.98 (0.09)	−0.51	< 0.001 ***

Note: Often/Always: perceived food noise reported often or always (n = 178). Never‐Sometimes: perceived food noise reported never, rarely, or sometimes (n = 822). Difference = Often/Always mean minus Never‐Sometimes mean; positive values indicate higher scores in the perceived frequent food noise group. p‐values from weighted Welch t‐test.

Abbreviation: SE, standard error.

*p < 0.05.

^** p < 0.01.

^*** p < 0.001.

In the multivariable proportional odds model (N = 938 complete cases; Table 3), loss of control over eating was the strongest independent predictor of frequent food noise (OR = 5.95, 95% CI: 4.05–8.75, p < 0.001), followed by Asian non‐Hispanic race/ethnicity (OR = 2.19, 95% CI: 1.26–3.83, p = 0.006), current GLP‐1 use (OR = 1.60, 95% CI: 1.08–2.37, p = 0.019), eating less to change weight or shape (OR = 1.57, 95% CI: 1.19–2.07, p = 0.002), and self‐identifying as being female (OR = 1.47, 95% CI: 1.14–1.88, p = 0.002). Older age was independently associated with lower odds of frequent food noise (OR = 0.99 per year, 95% CI: 0.98–1.00, p = 0.012), consistent with the descriptive pattern. Perceived importance of social support in dietary change (OR = 1.42, 95% CI: 1.26–1.60, p < 0.001), perceived genetic influence on weight (OR = 1.33, 95% CI: 1.16–1.52, p < 0.001), and poorer mental health (OR = 0.81, 95% CI: 0.70–0.93, p = 0.004) were independently associated with frequent food noise. Physical health was not independently significant after accounting for mental health and other covariates (OR = 1.04, p = 0.635).

TABLE 3.

Weighted ordinal logistic regression predicting perceived food noise frequency.

Predictor	OR	95% CI	p value
GLP‐1: Former user	1.43	0.86–2.38	0.172
GLP‐1: Current user	1.60	1.08–2.37	0.019 *
Age (years, continuous)	0.99	0.98–1.00	0.012 *
Sex: Female	1.47	1.14–1.88	0.002 **
Sex: Other	0.41	0.11–1.54	0.188
Education: Some college or associate	0.97	0.71–1.34	0.867
Education: Bachelor or higher	0.98	0.72–1.33	0.901
Income: $15k–$74k	1.39	0.87–2.22	0.163
Income: $75k–$149k	1.39	0.85–2.28	0.195
Income: $150k or more	1.94	1.11–3.40	0.021 *
Race: Black non‐Hispanic	0.97	0.64–1.47	0.880
Race: Hispanic	0.99	0.72–1.37	0.953
Race: Asian non‐Hispanic	2.19	1.26–3.83	0.006 **
Race: Other or multiracial	0.35	0.19–0.62	< 0.001 ***
Tried to lose weight (past 12 mo)	1.37	1.05–1.78	0.020 *
Loss of control over eating	5.95	4.05–8.75	< 0.001 ***
Eating less to change weight	1.57	1.19–2.07	0.002 **
Vomiting to change weight	1.17	0.54–2.54	0.683
Laxative or diuretic use	2.00	0.94–4.24	0.071 a
Importance of social support in dietary change	1.42	1.26–1.60	< 0.001 ***
Perceived genetic influence on weight	1.33	1.16–1.52	< 0.001 ***
Physical health (higher = better)	1.04	0.89–1.21	0.635
Mental health (higher = better)	0.81	0.70–0.93	0.004 **

Note: N = 938 complete cases. Proportional odds model (MASS: polr) with post‐stratification weights. OR > 1 indicates higher odds of more frequent food noise, holding all other variables constant. 95% CI = Wald confidence interval. Reference categories: GLP‐1 = Non‐user; Sex = Male; Education = HS or less; Income = less than $15k; Race = White non‐Hispanic.

Abbreviations: CI, confidence interval; OR, odds ratio.

^a p < 0.10.

^* p < 0.05.

^** p < 0.01.

^*** p < 0.001.

4. Discussion

In this nationally representative sample of U.S. adults, approximately 1 in 2 reported experiencing food noise at least sometimes, and nearly 1 in 5 reported experiencing it often or always, suggesting that self‐reported experiences labeled as “food noise,” such as constantly thinking about food or feeling that one's life revolves around food, are relatively common in the general U.S. population. Perceptions of frequent food noise were more prevalent among female‐identifying participants, younger adults, and Asian non‐Hispanic respondents, and were most strongly associated with reported loss of control over eating (over half of those endorsing loss of control over eating also reported frequent food noise, and this remained the strongest independent predictor in multivariable analyses). These patterns align with existing literature on food cue reactivity and food noise, which similarly maps onto sex, age, and disordered eating constructs [1, 2, 16]. The strong co‐occurrence of reported perceived food noise with loss‐of‐control eating, restrictive eating behaviors, poorer mental health, and lower self‐rated diet quality collectively suggests that people who experience psychological and behavioral burdens around eating are more likely to believe they experience food noise more frequently, consistent with studies in which food noise is perceived as a distressing experience [5, 6].

Current and former GLP‐1RA users in this sample reported higher rates of frequent perceived food noise than non‐users. This finding may seem counterintuitive at first glance given the frequent anecdotal reports and emerging evidence supporting that GLP‐1RAs might reduce food noise [1, 5, 6, 9, 15] and recent cross‐sectional data showing lower levels of food cue responsivity in individuals who had been using GLP‐1RAs for 6 months compared to individuals who weren't yet using a GLP‐1RA but were clinically eligible (i.e., treatment candidates) for such treatment [17]. Such contrast is likely explained by selection rather than treatment effect: individuals living with obesity or insulin resistance–conditions associated with heightened food cue reactivity [1, 18]–are precisely those most likely to seek or be prescribed GLP‐1RAs [19]. In this context, higher rates of perceived frequent food noise among current users may reflect the underlying clinical profiles that drive treatment initiation rather than a paradoxical medication effect. Longitudinal designs with pre‐ and post‐assessments will be essential to disentangle baseline vulnerability from medication‐related changes in perceptions of food noise. Furthermore, it is possible that awareness of one's experiences of food noise might be greater among users of GLP‐1RAs because experiencing some level of alleviation from constantly thinking about food might make individuals more aware of it once they realize they can contrast their previous “normal” levels of food‐related intrusions with their post‐treatment experiences. This view is consistent with perspectives discussed by Barsky & Borus as early as the 20th century [20], which hold that individuals' levels of tolerance for symptoms may decline once new technologies become available to alleviate them, leading to relabeling previously accepted states as pathological once they become treatable—in the context of this study, it is possible that, by feeling reductions in food noise thanks to GLP‐1RAs, patients become more aware of it now that they know it can be managed.

It is important to acknowledge that, despite the widespread interest in researching food noise as a separate construct, the current disagreements on its definitions [9, 11] highlight how limited the research is in trying to provide it with the necessary construct validity so that it stands apart as a construct of its own. For instance, the existence of the FNQ and the RAID‐FN, both of which were developed by drawing item pools from anecdotal reports and expert panels, does not necessarily mean that food noise is a separate, novel construct representing an aspect of human experience that hasn't been measured before. It can be argued that food noise has likely always existed, as a product of the combination of individual factors that increase susceptibility to food cues and a toxic food environment dominated by ubiquitous, pervasive food marketing and the widespread availability of convenient, highly palatable, low‐nutrient foods [21, 22, 23, 24]. In that sense, it is likely that food noise has been researched under different names over time, a possibility that needs to be investigated by additional studies assessing its construct validity. Regardless of what food noise actually is or how it should be measured as a patient‐reported outcome measure in studies or clinical settings (questions that still require extensive research to be properly addressed, and responding to which is beyond the scope of this paper), estimating the proportion of US adults who believe they experience food noise (in other words, endorse using the term food noise to describe their experiences) provides valuable information to gauge the magnitude of this phenomenon—whether it represents a unique construct or is a term people found to convey experiences already captured by other constructs that are otherwise hard to describe.

In a recent publication by Alarfaj and colleagues [25], for example, the authors assessed levels of food noise by repurposing a psychometric tool published in 2010, the Food Thought Suppression Inventory (FTSI) [26, 27]. The FTSI includes items such as “I have thoughts about food that I cannot stop” and “sometimes I wish I could stop thinking about food,” which are adapted from the White Bear Suppression Inventory (WBSI) [28]. A close inspection of the items in the FTSI reveals a very close overlap between the idea of food‐related thought suppression and food‐related intrusive thoughts, preoccupation, and obsessive thoughts about food, all elements hypothesized as defining characteristics of food noise [1, 2, 3]. This methodological choice highlights that different research teams might hold different views on what food noise represents and how to measure it at this point. Prematurely approaching food noise as a separate, measurable construct just because of the existence of recent psychometric tools that claim to measure food noise, before having enough data to guarantee that it is not redundant with existing constructs, could contribute to the jangle fallacy, in which the same construct is referred to by multiple names, all of which with their separate psychometric scales, which fragments the literature, makes comparisons and meta‐analyses harder, and can produce misleading claims of novelty [29]. By employing a single‐item measurement that exclusively seeks to estimate the proportion of US adults who endorse experiencing food noise as described in lay terms, rather than attempting to measure it as a unique construct, the present study adds to the literature by gauging how widespread the identification with the term is in the adult US population, without treating it as a consolidated measurable construct [9, 10].

This analysis has several limitations that should be considered when interpreting its findings, including its cross‐sectional design; the limitations inherent to secondary analyses of existing databases, which limit the data available to that collected in the database, such as the use of single items to assess variables of interest such as perception of dietary quality rather than validated measures of dietary quality and the reliance on lay interpretations of questions rather than providing official definitions for concepts such as loss of control over eating. Likewise, the database didn't include potentially useful information, such as participant weight and history of weight cycling, both of which could be linked to perceptions of food noise. Therefore, readers should be cognizant that this analysis is exploratory in nature, and its findings are hypothesis‐generating. Additionally, the regression model relied on complete cases only (N = 938), and missing data were not imputed. Nevertheless, it suggests that self‐reported experiences labeled as “food noise” are common and associated with demographic and disordered eating–related factors, highlighting widespread identification with this emerging construct and the need for further research to define, measure, and understand its clinical significance.

Funding

Access to the database was provided free of cost by Verasight to attendees of the 2026 meeting of the Society of Behavioral Medicine (SBM).

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from Verasight. Restrictions apply to the availability of these data, which were used under license for this study. Data are available from the author(s) with the permission of Verasight.