Perceived Competence and GLP‐1RA Weight Management Engagement: Mixed‐Methods Insights From Six Patient Groups

ABSTRACT

Objective

This study aimed to explore patients' perspectives on GLP‐1 receptor agonists (GLP‐1RAs) for weight loss across the continuum of contemplating use to discontinuation, with and without achieving weight loss goals.

Methods

From June to October 2025, participants completed a 30‐item survey and ~30‐question semi‐structured interview regarding their perceptions of GLP‐1RAs for weight loss. Interviews were tailored to six GLP‐1RA groups: (1) considering use; (2) < 3 months on therapy; (3) ≥ 3 months on therapy without achieving weight loss goal; (4) ≥ 3 months on therapy with achieving weight loss goal; (5) discontinued without achieving goal; and (6) discontinued after achieving goal. Qualitative coding and thematic analysis were conducted in Dedoose (9.0.107); quantitative analyses used SAS 9.4.

Results

A total of 185 participants consented, and 141 completed data collection (76%). Descriptive profiles suggested heterogeneity across psychosocial and behavioral domains. Groups 3 and 4 showed stronger profiles (higher perceived competence, intrinsic motivation, and self‐monitoring). After false discovery rate adjustment across 78 comparisons, perceived competence was the only statistically significant differentiator: Group 1 was lower (z = −3.15; q = 0.0347) and Group 4 higher (z = 4.44; q < 0.001).

Conclusions

Findings underscore the importance of tailoring behavioral support to individuals' stage of engagement in GLP‐1RA‐supported weight loss journeys.

Study Importance

• What is already known?

  • ○
    GLP‐1 receptor agonists (GLP‐1RAs) produce substantial weight loss, but patient experiences, behavioral variability, and psychosocial mechanisms driving engagement and adherence remain poorly understood.
  • ○
    Confidence, self‐efficacy, motivation, cravings, and self‐regulation are known to influence weight management behaviors, but their role among GLP‐1RA users has not been systematically characterized.
• What does this study add?

  • ○
    We identify distinct psychosocial profiles across GLP‐1RA user groups, with perceived competence emerging as the only statistically robust differentiator after false discovery rate correction.
  • ○
    Qualitative interviews reveal that confidence aligns with self‐efficacy, strategy use, craving control, and behavioral persistence, offering mechanistic explanations for variation in GLP‐1RA engagement.
• How might these results change the direction of research or the focus of clinical practice?

  • ○
    Findings support adding brief confidence/competence and craving‐related screening into routine GLP‐1RA care to tailor behavioral support by patient profile or stage of engagement.
  • ○
    Results highlight the need for mechanism‐informed, individualized behavioral interventions to optimize adherence, persistence, and long‐term outcomes for GLP‐1RA users.

1. Introduction

The rapid adoption of glucagon‐like peptide‐1 receptor agonists (GLP‐1RAs) has transformed modern obesity care, marking a shift toward biologically targeted treatments for weight management [1]. Quantitative studies document average weight loss and common adverse effects [2, 3]. Likewise, qualitative studies have described patient expectations and perspectives, particularly among those currently on GLP‐1RAs [4]. While controlled trials and real‐world studies have shown an average weight loss of approximately 15%–20% [2, 3], these summary estimates mask substantial heterogeneity in patient experiences of medication initiation [5], engagement [6], persistence [7], and weight loss outcomes [8].

Previous studies have suggested short and poor‐quality sleep habits [9], chronic and high perceived stress related to emotional eating, depression, food cravings, self‐weighing and monitoring [10, 11], low to no physical activity [12], and high alcohol intake [13] as factors linked to weight gain and impaired ability to lose or maintain weight loss. Psychological factors of intrinsic and extrinsic motivation as well as self‐regulation/self‐efficacy [14] can facilitate weight loss and maintenance but do not guarantee success. Interventions that enhance perceived competence in lifestyle changes have been associated with improved weight management‐related behaviors [15, 16]. Stress management‐based interventions show a context‐dependent association with weight loss and improved metabolic health [17]. Pre‐existing medical conditions, such as chronic pain or hypothyroidism, demonstrate highly condition‐specific and heterogeneous relationships with weight loss and long‐term weight maintenance, with outcomes shaped by functional limitations, symptom burden, treatment effects, and behavioral adaptations rather than by diagnosis alone [18, 19]. Finally, even with “good behaviors” such as frequent self‐monitoring, higher activity, healthier eating, and strong autonomous motivation, weight loss or maintenance is not guaranteed [20, 21]. To our knowledge, no study has comprehensively integrated behavioral data, mental health indicators, and demographic factors with patients' lived experiences across different GLP‐1RA treatment trajectories. Understanding these patterns is increasingly important as GLP‐1RAs move from specialty care into broader, more accessible obesity management.

To address this gap, we examined patient journeys into, through, and off GLP‐1RA treatment, conceptualizing these experiences as distinct pathways shaped by behavioral patterns, mental health factors, and subjective experiences. The objective of this study was to develop an integrated mixed‐methods typology explaining how behavioral, psychological, and social factors correlate with patient pathways through the stages of GLP‐1RA medication use. This typology advances understanding of why individuals initiate GLP‐1RA therapy, how they experience treatment, what facilitates or hinders weight loss, and what drives decisions to continue or discontinue medication. To quantify the relative salience of these factors, we constructed a heat map comparing z ‐scored profiles for 13 key characteristics across patient groups. This visualization clarifies which dimensions are most and least influential within each pathway and provides a foundation for evidence‐based, stage‐aligned recommendations to strengthen clinical support across GLP‐1RA trajectories. These analyses are intended to describe patterns of association and should not be interpreted as evidence of the directionality of relationships among the identified factors. The heat map uses a diverging color gradient, with red indicating lower (less favorable) standardized scores and blue indicating higher (more favorable) standardized scores for each characteristic. To deepen interpretation of the heat map results, we integrated participants' quantitative responses with their qualitative accounts. These narratives provide contextual grounding for the quantitative patterns, illustrating how and why particular characteristics relate to individuals' motivation, treatment experiences, and decisions regarding continuation or discontinuation.

2. Methods

We used an explanatory convergent mixed methods approach to characterize six naturally occurring GLP‐1RA patient pathways: (1) considering GLP‐1RA medication use; (2) < 3 months on GLP‐1RA; (3) ≥ 3 months on GLP‐1RA without achieving a weight loss goal; (4) ≥ 3 months on GLP‐1RA with achieving a weight loss goal; (5) discontinued GLP‐1RA without achieving a weight loss goal; and (6) discontinued GLP‐1RA after achieving a weight loss goal. Eligibility criteria included age ≥ 18 years, no history of pancreatitis, type 2 diabetes, or bariatric surgery or personal or family history of multiple endocrine neoplasia type 2 (MEN2) or medullary thyroid carcinoma, and willingness to participate in a semi‐structured interview. All participants provided written informed consent in accordance with the Declaration of Helsinki. The study was approved by the University of Missouri Institutional Review Board (IRB #2125541). Participants received a $20 electronic gift card following interview completion. Data were collected remotely using REDCap surveys and telephone interviews conducted between June and October 2025.

Quantitative measures include instrument for physical activity (IPAQ) [22], sleepiness (Epworth) [23], stress (Jackson Heart Study) [24], depression (PHQ‐2) [25], motivation (internal and external motivation, TSRQ); perceived competence (PCS) [26], eating behaviors and cravings (FCQ) [27], self‐monitoring (use of scale) [28], number of chronic conditions, pain (SF‐36 subscale) [29], and alcohol use (BRFSS source) [30]. Demographic variables included age categories (< 25, 25–44, 45–64, ≥ 65 years), sex, race (White, Black, Hispanic, multirace/another race), marital status (never married, married/long‐term relationship, widowed/divorced/separated), education attainment (high school/GED, some college/technical school‐no degree, 2‐year college degree/technical school degree, 4‐year college degree, postgraduate work or degree), employment status (working full time or part time, retired, disabled, homemaker, in school/not working, or unable to work), insurance status (Medicaid, Medicare, self‐pay, commercial/private), annual household income (< $60,000; $60,000–$99,000; $100,000–$149,000; ≥ $150,000), perception of duration of medication (lifelong, short‐term, follow doctor's advice, don't know/not sure), duration of medication use: current or have taken (in months), type of GLP‐1RA (FDA‐approved GLP‐1RA, compounded GLP‐1RA), weight before starting GLP‐1RA, and weight loss goal. Due to our grouping by achievement of the individual's goal and status of GLP‐1RA medication use, we characterized individual goals by group. We grouped weight loss goals into three categories: under average expectation (< 13%), on par with expectation (13%–20%), and over average expectation (> 20%). This grouping was informed by SURMOUNT‐5, the head‐to‐head trial of both semaglutide and tirzepatide at 72 weeks [31]. To facilitate comparison across psychosocial and behavioral domains measured on different scales, continuous variables were standardized to z scores. To control for multiple comparisons, we applied the Benjamini‐Hochberg false discovery rate (FDR) procedure. Standardized values were organized using principal component analysis (PCA) to structure visualization, and results were displayed in a heat map comparing 13 key characteristics across the six groups. Descriptive and statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC).

Qualitative data include in‐depth interviews covering themes such as safety, concerns, lifestyle changes, social norms, side effects, motivations, reasons for discontinuation, and experiences before, during, and after medication use. Details on the qualitative survey are described in a previous published manuscript [5]. Briefly, semi‐structured interviews were developed using constructs from the Health Belief Model (HBM) and designed to elicit participants' experiences, perceptions, motivations, and barriers related to GLP‐1RA use for weight management. Given the limited and focused study aims, we targeted thematic saturation with approximately 10–20 interviews per group [32].

We applied conventional content analysis to examine responses regarding GLP‐1RA use for weight loss [33]. To ensure transparency and rigor, reporting adhered to the Consolidated Criteria for Reporting Qualitative Research (COREQ‐32) [34]. While HBM informed the interview guide and interpretation, coding in Dedoose version 9.0.107 (Los Angeles, CA), a cloud‐based platform for qualitative and mixed‐methods analysis, was inductive, allowing themes to emerge from participant narratives rather than imposing a predefined framework, an approach well suited for areas with limited prior research [35]. Intercoder reliability was assessed using the platform's Testing Center, with pooled kappa scores calculated to evaluate agreement beyond chance.

3. Results

To assess the reliability of the qualitative coding process, an interrater reliability check was conducted on a subset of 10 codes applied to a random sample of 59 transcripts. Two coders (J.M., lead coder, and C.M.) independently coded these transcripts. Across the full dataset, J.M. and C.M. coded all transcripts either independently or as part of a coding pair. Cohen's kappa was calculated for each selected code, and the overall aggregated kappa value was 0.81, indicating excellent interrater agreement [36].

Between June and October 2025, a total of 232 individuals expressed interest in the study. Of these, 80% (n = 186) completed the consent process; however, 37 did not continue to the interview stage. This resulted in 149 completed interviews. A total of 8 interviews were later excluded, 6 due to reliability concerns and 2 because participants were deemed ineligible during the interview, leaving a final analytic sample of 141 interviews (Figure 1).

FIGURE 1.

Participant flow diagram.

The number of interviews ranged from 15 to 31 across groups with approximately 60% female and 40% male. Interviews averaged approximately 30 min in length (Table 1).

TABLE 1.

Characteristics of participants among six groups of GLP‐1RA medication use for weight loss.

	All		Group 1		Group 2		Group 3		Group 4		Group 5		Group 6
	N = 141	%	N = 31	%	N = 22	%	N = 30	%	N = 23	%	N = 20	%	N = 15	%
Age (years)
18–24	9	6.4	6	19.4	1	4.6	—	—	1	4.4	—	—	1	6.7
25–44	61	43.3	10	32.3	8	36.4	16	53.3	12	52.2	7	35	8	53.3
45–64	59	41.8	12	38.7	12	54.6	13	43.3	8	34.8	9	45	5	33.3
65+	12	8.5	3	9.7	1	4.6	1	3.3	2	8.7	4	20	1	6.7
Marital status
Never married	29	20.6	8	25.8	2	9.1	6	20	6	26.1	3	15	4	26.7
Married‐long‐term relationship	93	66	20	64.5	16	72.7	20	66.7	14	60.9	14	70	9	60
Widowed/divorced/separated	18	12.8	2	6.5	4	18.2	4	13.3	3	13.1	3	15	2	13.3
Income
< $60,000	33	23.4	12	38.7	2	9.1	5	16.7	5	21.7	5	25	4	26.7
$60,000 to 99,999	34	24.1	7	22.6	7	31.8	6	20	7	30.4	3	15	4	26.7
$100,000 to $149,999	31	22	3	9.7	8	36.4	8	26.7	2	8.7	5	25	5	33.3
≥ $150,000 or higher	36	25.5	7	22.6	4	18.2	10	33.3	8	34.8	5	25	2	13.3
Prefer not to answer	7	5	2	6.5	1	4.6	1	3.3	1	4.4	2	10	—	—
Educational attainment
High school or GED	5	3.6	—	—	2	9.1	2	6.7	—	—	1	5	—	—
Some college or technical school‐no degree	16	11.4	6	19.4	3	13.6	2	6.7	3	13	1	5	1	6.7
2‐year college degree/technical school degree	14	9.9	2	6.5	1	4.6	2	6.7	5	21.7	3	15	1	6.7
4‐year college degree	46	32.6	14	45.2	6	27.3	9	30	4	17.4	7	35	6	40
Postgrad work or degree	59	41.8	9	29	10	45.5	14	46.7	11	47.8	8	40	7	46.7
Employment status
Working full or part time	119	84.4	26	83.9	20	90.9	25	83.3	21	91.3	14	93.3	13	65
Retired, disabled, homemaker, in school not working, or unable to work	20	14.2	5	16.1	2	9.1	4	13.3	1	4.4	1	6.7	7	35
Insurance status
Medicaid, Medicare, self‐pay	21	14.9	6	19.4	2	9	3	10	3	13	6	30	1	7
Commercial, private	119	84.4	24	77.4	20	90.9	27	90	20	87	14	70	14	93.3
Race
White	117	83	23	74.2	21	95.5	26	86.7	20	87	17	85	10	66.7
Hispanic	8	5.7	1	3.2	—	—	2	6.7	1	4.4	2	10	2	13.3
Multirace, another race	9	6.4	5	16.1	—	—	2	6.7	1	4.4	1	5	—	—
Black	7	5	2	6.5	1	4.6	—	—	1	4.4	—	—	3	20
Sex
Male	55	39	9	29	8	36.4	15	50	9	39.1	10	50	4	26.7
Female	86	61	22	71	14	63.6	15	50	14	60.9	10	50	11	73.3
Perception of length
Lifelong	37	26.2	4	12.9	3	13.6	13	43.3	9	39.1	4	20	4	26.7
Short‐term	45	31.9	12	38.7	12	54.6	3	10	5	21.7	6	30	7	46.7
Follow doctor advice	17	12.1	5	16.1	—	—	3	10	1	4.4	6	30	2	13.3
Don't know/not sure	42	29.8	10	32.3	7	31.8	11	36.7	8	34.8	4	20	2	13.3
Type of GLP‐1RA
FDA‐approved GLP‐1RA	93	66	—	—	18	81.8	28	93.3	19	82.6	17	85	11	73.3
Compounded GLP‐1RA	13	9.2	—	—	3	13.6	1	3.3	3	13	3	15	3	20
Missing	35	24.8	31	100	1	4.6	1	3.3	1	4.4	—	—	1	6.7
Duration of medication use: current or have taken (months), mean (SD)	110	16.3 (13.2)	—	—	22	1.8 (1.3)	30	11.5 (6.0)	23	13.0 (6.8)	20	6.4 (5.2)	15	8.2 (7.2)
Length of interview, mean (SD)	141	31 (11.7)	31	24 (9.1)	22	33 (9.5)	30	36 (11.8)	23	36 (14.8)	20	30 (6.6)	15	25 (9.4)

Note: Group definitions: 1—considering taking a GLP‐1RA medication, 2—recently started a GLP‐1RA in past 3 months, 3—has been on a GLP‐1RA for over 3 months without reaching goal weight, 4—has been on a GLP1‐RA for over 3 months and reached goal weight, 5—has not reached goal weight and gone off GLP‐1RA medication, 6—has reached goal weight using a GLP‐1RA and gone off medication.

Participants had a mean age of 45 years (range 19–79). The sample was predominantly White (83%). Most respondents were employed (84%), carried private health insurance (90%), and were married (66%). Nearly one quarter (23%) reported an annual household income below $60,000. The majority (88%) reported use of an FDA‐approved GLP‐1RA, including medications such as tirzepatide (Zepbound, Mounjaro), semaglutide (Wegovy, Ozempic), and liraglutide (Saxenda) (Table 1).

Across the six groups, descriptive z score profiles suggested meaningful heterogeneity in all psychosocial and physical health domains with within‐group standard deviations (SD) ranging from 0.61 to 1.36 (Table 2). Because z scores reflect a standardized reference distribution (SD = 1.0), values below 1 indicate tighter clustering and values above 1 indicate greater dispersion within a group. Certain constructs, including intrinsic motivation, perceived competence, weighing behaviors, depressive symptoms, and food cravings, show the tightest clustering in specific groups (SD 0.61–0.67), whereas sleepiness, depressive symptoms, stress, food cravings, number of comorbidities, and intrinsic motivation exhibited wider dispersion (SD 1.22–1.36; Table 2, Figure 2).

TABLE 2.

Health‐related behavioral and psychosocial z scores (mean of 0; standard deviation [Std] of 1) indicators by GLP‐1RA groups.

Behavioral and psychosocial scales	Group classification
	1 (n = 31)		2 (n = 22)		3 (n = 30)		4 (n = 23)		5 (n = 20)		6 (n = 15)
	Mean	Std	Mean	Std	Mean	Std	Mean	Std	Mean	Std	Mean	Std
Sleepiness (Epworth) a	−0.06	1.22	−0.06	0.89	0.01	1.00	0.04	0.85	0.30	0.77	−0.28	1.17
Food cravings (FCQ) a	−0.44	1.02	0.09	1.26	0.20	0.87	0.43	0.66	−0.12	1.03	−0.12	0.91
Perceived competence	−0.52	0.92	0.20	0.79	0.40	0.78	0.62	0.67	−0.69	1.17	−0.04	1.06
Stress (Jackson stress) a	−0.01	1.02	0.13	0.97	0.02	0.98	0.02	0.96	0.20	0.85	−0.51	1.26
% Obesity social network a	−0.02	0.96	0.11	1.09	0.05	1.10	0.04	0.91	−0.18	0.91	−0.04	1.13
Depressive symptoms (PHQ) a	−0.25	1.32	0.13	0.61	0.05	1.07	0.10	0.89	0.08	0.89	−0.04	0.91
Number of comorbidities a	0.18	1.00	−0.13	0.96	−0.12	0.81	−0.07	1.18	0.11	0.97	0.04	1.22
Food strategies	−0.12	1.13	0.17	1.00	−0.13	0.90	0.33	1.05	−0.24	0.92	0.07	0.92
Physical activity (IPAQ)	0.04	0.97	−0.13	0.84	−0.34	1.12	0.09	1.06	0.09	1.05	0.52	0.70
Weighing behaviors	−0.48	1.17	−0.44	1.13	0.38	0.73	0.36	0.61	0.09	0.93	0.22	0.94
Pain (SF‐36) a	0.45	1.15	−0.22	1.11	−0.01	0.86	−0.30	0.85	0.05	0.77	−0.21	1.05
Intrinsic motivation (TSRQ)	−0.48	1.10	0.07	0.62	0.35	0.85	0.47	0.66	−0.40	1.02	−0.01	1.36
Extrinsic motivation (TSRQ)	0.04	0.91	0.06	0.98	0.15	0.98	−0.32	1.12	0.15	1.14	−0.18	0.88

^a Reverse‑scored: Higher (positive) values indicate healthier behaviors or psychosocial functioning; lower (negative) values indicate less healthy behaviors or poorer psychosocial functioning.

FIGURE 2.

Heat map illustrating behavioral and psychosocial z score above and below mean of 0. [Color figure can be viewed at wileyonlinelibrary.com]

We found that more than half of participants had a weight loss goal above 20%, over average expectation, with the majority of these people in the group that had not yet achieved this weight loss goal (Figure 3).

FIGURE 3.

Weight loss goals by groups. [Color figure can be viewed at wileyonlinelibrary.com]

Qualitative interviews helped contextualize these quantitative profiles (Table 3, Figure 2). Groups 1 and 5 described lower confidence in their ability to lose weight, mirroring lower perceived competence z scores. This difference was statistically significant for Groups 1 and 4 after FDR adjustment and directional for Group 5. Group 1 participants described persistent self‐doubt (“I'm not successful even if I do [list of weight loss strategies]”), and Group 5 participants highlighted age‐related challenges (“The older I get, it's more and more difficult to lose weight and to maintain”). In contrast, Group 4 expressed strong confidence and mastery of new habits (“I've definitely learned a lot, and I think they're lasting habits”). Competence variability was also observed in other groups; for instance, a Group 2 participant noted (“I get frustrated with myself if I don't see anything coming out of my work”), illustrating mid‐range competence consistent with near average quantitative scores. Individuals who achieved their weight loss goal appear to have higher or more normalized intrinsic motivation after discontinuing GLP‐1RA therapy compared with those who did not achieve their goal. This may suggest that successful weight loss is associated with stronger internalized motivation to regulate eating or health behaviors once medication is stopped. Finally, participants receiving GLP‐1RA therapy for more than 3 months tended to exhibit more favorable profiles across established weight loss–related factors, including perceived competence, intrinsic motivation, reduced food cravings, and consistent weighing behaviors.

TABLE 3.

Joint display integrating quantitative indicators and qualitative themes across groups.

Domain	Quantitative trends	Key qualitative themes	Exemplar quotes	Integrated interpretation
Perceived competence/confidence (FDR‐significant)	G1 ↓, G4 ↑	Competence/self‐efficacy; Confidence in weight management	“…frustrating when you do everything you can, and the needle doesn't move.” (G1) “Inability to actually lose weight.” (G1) “I'm confident my changes are going to do it for me.” (G4)	Confidence differentiates groups; reflects SDT competence/self‐efficacy mechanisms that likely drive engagement.
Weighing behaviors/self‐monitoring	G1 ↓ (trend); G3–G4 ↑ (trend)	Lifestyle change; Routine maintenance	“I try to weigh myself every week.” (G3)	Routine tracking supports weight management execution.
Food noise/craving burden	G1 ↑ (trend); G4 ↓ (trend)	Food‐related cognitive load; Appetite regulation	“GLP‐1 s quiet some of that food noise.” (G1) “Food chatter is completely cut out—I eat when my body needs to.” (G4)	Appetite/cue suppression mediates adherence.
Physical activity	G4 ↑ (trend)	Lifestyle change; Intentional activity	“Consistent and intentional with exercise routines.” (G4)	Increased activity complements medication effects and reflects proactive behavior.
Pain/function	G4 ↑ (trend)	Symptom relief vs. functional limitations	“Decreased pain has let me be more active.” (G4)	Pain improvements may support routines.
Intrinsic motivation	G1 ↓ (trend); G3–G4 ↑ (trend)	Internal motivators; Health‐driven change	“I had to do something—menopause was not a friend of mine.” (G1)	Internal, autonomous motivation aligns with better engagement.

Participants across all groups described difficulty losing and maintaining weight, with internal motivators more salient than external pressures. Improved energy, family engagement, and health concerns were highly emphasized motivators. A Group 3 participant shared, “I want to be able to be more active with my kid,” and a Group 4 participant cited multiple health concerns (“Blood pressure going up… cholesterol… sleep apnea”). Some referenced seeing GLP‐1RA benefits among relatives (“She quit drinking… now goes out with her grandbabies”).

“Food noise” emerged as a salient theme and aligned with quantitative trends. Group 1 participants anticipated relief from intrusive thoughts about food (“I was really hoping the GLP 1s would help turn that noise off… and whether [or not] it comes back after stopping”). Groups 3 and 4 described meaningful reductions that supported self‐regulation (“It's that food chatter…completely cut out. Now I eat when my body needs to eat”). Some participants in Groups 5 and 6 expressed concern about food noise returning as medication effects plateaued or doses changed.

Pain and physical activity also varied. Group 1 participants less often cited pain as a barrier, whereas several Group 4 participants described musculoskeletal pain impacting daily activity (“It was affecting my spine”; “Terrible arthritis… unable to exercise effectively”). Despite these barriers, Group 4 participants emphasized purposeful engagement in activity (“Consistent and intentional with exercise routines”; “Doing my regular exercise”). Regular self‐weighing also distinguished groups, with Groups 3 and 4 more often describing consistent weighing habits (“I usually try to do it every week”), aligning with their more favorable weight‐related z ‐scores.

In summary, Groups 3 and 4, both currently taking GLP‐1RA medications, demonstrated the strongest psychosocial profiles, including higher perceived competence, greater intrinsic motivation, and more consistent self‐monitoring behaviors. Groups 1 and 5 exhibited vulnerabilities centered on competence and motivation (with Group 1 also showing more favorable pain outcomes and Group 5 demonstrating lower perceived competence). Group 6 exhibited a heterogeneous profile, characterized by higher physical activity and self‐monitoring alongside elevated stress and poorer sleep. Group 2 presented with near average profiles and minor deviations. After controlling for FDR across 78 comparisons, only perceived competence remained statistically significant, with Group 1 significantly lower (z = −3.15; q = 0.0347) and Group 4 significantly higher (z = 4.44; q < 0.001). All other cross‐group differences should be interpreted as descriptive patterns. Taken together, the mixed methods findings indicate that perceived competence, a construct closely aligned with confidence within Bandura's Social Cognition Theory, is the most robust differentiator across groups, while qualitative narratives elucidate the behavioral processes underlying these patterns.

4. Discussion

Distinct psychosocial profiles across GLP‐1RA experience groups suggest meaningful opportunities for targeted, mechanism‐informed care. These profiles highlight heterogeneity across domains, with Groups 3 and 4 demonstrating the most robust psychosocial strengths; Groups 1 and 5 exhibiting perceived competence and motivation‐related vulnerabilities; and Groups 2 and 6 presenting more heterogeneous profiles. Group 2 reflects largely neutral early‐stage patterns, whereas Group 6 demonstrates a combination of strong behavioral engagement (e.g., physical activity, self‐monitoring) alongside elevated stress and sleep burden. Perceived competence uniquely differentiates groups at a statistically robust level and qualitative narratives reinforced its central role. Higher competence groups exhibited narratives consistent with greater self‐efficacy (e.g., consistent strategy use and perceived behavioral control), whereas lower competence groups described themes indicative of lower competence and greater behavioral struggle. Importantly, because these data are cross‐sectional, the observed associations cannot establish directionality. These profiles characterize psychosocial attributes of individuals at different stages of the weight loss journey but do not determine whether greater perceived competence contributes to more successful weight loss or emerges as a consequence of weight loss success.

Patterns around cravings, self‐regulation capacity, and autonomous motivation are broadly consistent with, though not uniformly reflective of, emerging literature on GLP‐1RA‐mediated modulation of reward and craving [37]. This aligns with longstanding evidence that autonomy support and perceived competence are core predictors of behavioral weight loss outcomes [38]. Drawing on established behavioral weight management literature, early‐stage and contemplating users may therefore benefit most from brief, structured interventions, such as motivational interviewing and autonomy‐supportive (choice‐rich) counseling, to clarify goals, address ambivalence, and build early micro‐successes. This aligns with trials showing that motivational interviewing enhances adherence and self‐monitoring in obesity treatment [39]. These stage‐specific needs and practice considerations are summarized in Figure 4, which outlines tailored strategies for each GLP‐1RA user group. Figure 4 should be interpreted as a conceptual framework derived from participant‐reported experiences and existing behavioral obesity literature rather than as an evidence‐based clinical algorithm. The proposed strategies are intended to generate hypotheses and inform future intervention development and testing.

FIGURE 4.

Stage‐specific considerations for GLP‐1RA clinical care. [Color figure can be viewed at wileyonlinelibrary.com]

Active users who have not yet met weight loss goals appear to require additional support to optimize food‐related strategies, self‐regulatory skills, and autonomy support consistent with evidence that autonomy‐supportive communication and self‐regulatory skill building are associated with larger and more durable weight losses [16, 40].

By contrast, successful, ongoing users and those who discontinued after achieving weight loss may benefit from structured maintenance and/or relapse‐prevention strategies, particularly given signals of elevated stress and sleep disruption that could undermine long‐term sustainability [41]. This reflects broader weight maintenance evidence suggesting that continued low‐burden support improves long‐term outcomes [42]. Meanwhile, discontinuers with more adverse emotional and physical symptom profiles underscore the importance of addressing pain, sleep, stress, and cost/access barriers early. These needs resonate with community reports of GLP‐1RA discontinuation drivers and with emerging engagement work showing that low‐burden digital or brief tools can increase uptake of intensive behavioral treatment [43, 44, 45, 46]. Embedding a brief cravings/competence/autonomy screener into routine GLP‐1RA care, paired with nurse or dietitian follow‐ups, represents a pragmatic, scalable strategy that is supported by behavioral obesity and GLP‐1RA literature [40].

Several themes emerged consistently across groups that may warrant attention during clinical encounters. Participants frequently described concerns regarding social judgment related to GLP‐1RA use, such as being told they are “cheating” by friends, colleagues, or family members. A second pervasive theme was diminished appetite among GLP‐1RA users, highlighting a potential area for guidance on intentional dietary choices, particularly prioritizing adequate protein intake. Finally, many patients expressed concern about losing access to GLP‐1RA therapy due to changes in insurance coverage, underscoring the importance of discussing contingency plans to maintain or continue weight loss efforts if medication access becomes limited or discontinued.

Quantitative domains beyond perceived competence should be interpreted as descriptive trends rather than confirmatory differences. Perceived competence uniquely differentiates Groups 1 and 4 at a statistically robust level, whereas other domains reflect profile patterns that warrant cautious interpretation and validation in larger samples.

Our findings regarding participants' individualized weight loss goals highlight that patients' goals and evidence‐based expectations may not be aligned. Patients who feel “unsuccessful” or that they haven't achieved adequate weight loss may be meeting research‐based realities but not their own goals. This is a space for doctor‐patient communications about the expected weight loss on GLP‐1RAs.

Several limitations warrant consideration. Although we conducted a robust number of qualitative interviews within each group, quantitative analyses were limited by relatively small group sizes, reducing power to detect differences across the 13 domains. A small proportion of participants reported using compounded GLP‐1RAs, which may introduce heterogeneity in experiences and limit generalizability to FDA‐approved formulations. The study was not powered to perform sex‐based analyses. Additionally, the sample was predominately White and privately insured, which may limit generalizability to more diverse populations and to those facing greater access barriers. Finally, the cross‐sectional design precludes conclusions regarding the directionality of associations between psychosocial characteristics (e.g., perceived competence) and weight loss outcomes. Further work in larger cohorts will be necessary to validate these profiles and test whether tailored interventions improve engagement, persistence, and outcomes.

5. Conclusion

Our findings underscore the importance of recognizing where individuals are in their weight loss journey with GLP‐1RAs and tailoring behavioral support to their specific psychosocial needs. Heterogeneity within and across groups suggests that a one‐size‐fits‐all approach may be insufficient. Mechanism‐informed, stage‐aligned strategies that explicitly build perceived competence may optimize adherence, persistence, and real‐world effectiveness of GLP‐1RA‐supported weight management.

Funding

The authors have nothing to report.

Conflicts of Interest

Regina DePietro reports honorarium from PeerView ECHO type 2 diabetes and CGM teaching supplies from FreeStyle Libre, DexCom, and EverSense. The other authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.