The NIH “Effort Preference” Controversy (2024-2025)
The 2024 NIH deep phenotyping study by Walitt et al. (Walitt et al. 2024) generated one of the most significant controversies in recent ME/CFS research history. While the study documented multiple objective biological abnormalities (catecholamine deficiency, immune dysfunction, autonomic abnormalities), its interpretive framing around “effort preference” sparked intense criticism from patients, clinicians, and researchers.
1 The Central Claim
The study concluded that “effort preference, not fatigue, is the defining motor behavior” of post-infectious ME/CFS. The authors proposed that ME/CFS patients have altered “effort preference”—defined as “how much effort a person subjectively wants to exert”—due to dysfunction of integrative brain regions, particularly the temporoparietal junction (TPJ).
2 Why the Framing Was Controversial
2.1 Language Echoing Psychogenic Models
The term “preference” implies volition and choice. Critics argued this framing echoed decades of psychogenic characterizations of ME/CFS that attributed symptoms to patients’ beliefs, behaviors, or psychological states rather than biological dysfunction. The language resonated uncomfortably with PACE trial rhetoric about “unhelpful illness beliefs.”
2.2 Methodological Problems with the EEfRT
The Effort-Expenditure for Rewards Task (EEfRT) was designed to measure motivation for rewards in psychiatric conditions, with an explicit requirement that tasks be easy enough that fatigue does not confound results. In the Walitt study:
- ME/CFS patients completed only 65% of hard trials vs. 96–99% for controls
- Seven of 15 patients performed below any control participant
- Physical function scores (SF-36) were 28.7 for patients vs. 97.5 for controls Kirvin-Quamme et al.’s reanalysis (Kirvin-Quamme, Davenport, et al. 2025) demonstrated a significant correlation between task completion ability and task choice, indicating the tool measured ability, not preference. Their conclusion: patients were “unable,” not “unwilling.”
2.3 Failure to Document PEM
The study used single-day CPET rather than the gold-standard two-day protocol, failing to objectively document post-exertional malaise—the defining feature of ME/CFS. PEM was mentioned only three times in the entire paper.
2.4 Selection Bias
By excluding severely affected patients (25% of the ME/CFS population), the study could not characterize the full spectrum of disease severity.
3 Published Academic Responses
3.1 Nature Communications Commentary
The most substantive published rebuttal is the Matters Arising by Davenport et al. in Nature Communications (Davenport et al. 2025). It argues that the effort-preference and deconditioning interpretation “risks reinforcing skepticism about the serious biological nature of [ME/CFS] and its hallmark of post-exertional malaise (PEM), as well as its potential misclassification as a mental health condition.”
The rebuttal rests on a core methodological point: the study used a single CPET, which cannot characterize PEM because it cannot measure the response to an initial exertion. The two-day CPET, with a second maximal test 24 hours later, is the standard method for documenting post-exertional metabolism in ME/CFS (Lim et al. 2020), and the Institute of Medicine cautioned that a single test may be insufficient. Because deconditioning and PEM are not mutually exclusive, evidence of deconditioning cannot by itself explain PEM (Davenport et al. 2025).
The critique also turns the physiological data against the deconditioning hypothesis. Deconditioning typically produces elevated exercise heart rate at a given workload, yet exercise HR was lower in post-infectious ME/CFS participants than in healthy volunteers (Davenport et al. 2025). Davenport et al. read this as inconsistent with deconditioning and more consistent with chronotropic incompetence and impaired oxidative metabolism — both established ME/CFS features. They further argued the study was underpowered (only 8 post-infectious ME/CFS and 9 healthy volunteers completed CPET), lacked deconditioned control matching, and did not document outcome-measurement order in the abstract. The full argument is developed in the mechanistic-studies chapter at The NIH Deep Phenotyping Study (Walitt et al. 2024).
This response, together with the authors’ reply below and the Frontiers reanalysis (Kirvin-Quamme, Davenport, et al. 2025), illustrates the wider academic contest over whether the same objective data should be interpreted as deconditioning or as intrinsic metabolic and autonomic limitation. The stakes are clinical: the interpretation directly affects whether patients are treated with activity-based therapies or energy-envelope pacing. Severity applicability: broad, from mild to severe, since the deconditioning rebuttal applies to the full severity spectrum.
3.3 Frontiers in Psychology Reanalysis
The Kirvin-Quamme et al. reanalysis (Kirvin-Quamme, Davenport, et al. 2025) provided detailed statistical evidence that the EEfRT data supported inability rather than altered preference, calling for proper task calibration in future studies.
4 NIH Clarification
Following criticism, NIH clarified that “preference” referred to “subconscious or unconscious or pre-conscious calculations by the brain” rather than conscious choice. Patient advocates responded that if unconscious brain dysfunction was the intended meaning, using “preference”—a word implying choice—was misleading and potentially harmful to patients.
5 Clinical and Research Implications
5.1 Potential Harms
- Reinforcement of psychogenic misconceptions among clinicians unfamiliar with ME/CFS
- Justification for continued use of graded exercise therapy despite harms
- Barriers to disability recognition if symptoms are framed as “preference”
- Psychological harm to patients from invalidating language
5.2 Lessons for Future Research
- Assessment tools must be appropriate for the population studied
- Language matters: terminology should not inadvertently pathologize or blame patients
- Peer review should include ME/CFS experts and patient representatives
- Objective measures of PEM (two-day CPET) should be standard
- Study designs should include severely affected patients
6 Separating Data from Interpretation
Despite the interpretive controversy, the Walitt study’s biological findings—CSF catecholamine deficiency, B cell population shifts, autonomic dysfunction, chronotropic incompetence—represent valuable contributions confirmed by other research. The challenge for the field is to build on these objective findings while rejecting framings that risk harm to patients. The NIH deep phenotyping study illustrates how a study can simultaneously advance biological understanding and generate harmful interpretations. The catecholamine findings alone—the first CSF neurotransmitter measurements in ME/CFS—provide crucial mechanistic insight. Future citations should specify which findings are being referenced (the objective biological data vs. the contested “effort preference” interpretation) to prevent misuse while preserving scientific value.