Adaptive coding of reward in schizophrenia, its change over time and relationship to apathy

• Published in: Brain (London, England : 1878) England : 1878), 2024-07, Vol.147 (7), p.2459-2470
• Main Authors: Kaliuzhna, Mariia, Carruzzo, Fabien, Kuenzi, Noémie, Tobler, Philippe N, Kirschner, Matthias, Geffen, Tal, Katthagen, Teresa, Böge, Kerem, Zierhut, Marco M, Schlagenhauf, Florian, Kaiser, Stefan
• Format: Article
• Language: English
• Subjects: Adaptation, Physiological - physiology; Adaptation, Psychological - physiology; Adult; Apathy - physiology; Brain - diagnostic imaging; Brain - physiopathology; Female; Humans; Magnetic Resonance Imaging; Male; Middle Aged; Motivation - physiology; Original; Reward; Schizophrenia - physiopathology; Schizophrenic Psychology
• ISSN: 0006-8950; 1460-2156; 1460-2156
• DOI: 10.1093/brain/awae112

Adaptive coding of reward is the process by which neurons adapt their response to the context of available compensations. Higher rewards lead to a stronger brain response, but the increase of the response depends on the range of available rewards. A steeper increase is observed in a narrow range and a more gradual slope in a wider range. In schizophrenia, adaptive coding appears to be affected in different domains, especially in the reward domain. Here, we tested adaptive coding of reward in a large group of patients with schizophrenia (n = 86) and control subjects (n = 66). We assessed: (i) the association between adaptive coding deficits and symptoms; (ii) the longitudinal stability of deficits (the same task was performed 3 months apart); and (iii) the stability of results between two experimental sites. We used functional MRI and the monetary incentive delay task to assess adaptation of participants to two different reward ranges: a narrow range and a wide range. We used a region-of-interest analysis to evaluate adaptation within striatal and visual regions. Patients and control subjects underwent a full demographic and clinical assessment. We found reduced adaptive coding in patients, with a decreased slope in the narrow reward range with respect to that of control participants, in striatal but not visual regions. This pattern was observed at both research sites. Upon retesting, patients increased their narrow-range slopes, showing improved adaptive coding, whereas control subjects slightly reduced them. At retesting, patients with overly steep slopes in the narrow range also showed higher levels of negative symptoms. Our data confirm deficits in reward adaptation in schizophrenia and reveal an effect of practice in patients, leading to improvement, with steeper slopes upon retesting. However, in some patients, an excessively steep slope may result in poor discriminability of larger rewards, owing to early saturation of the brain response. Together, the loss of precision of reward representation in new (first exposure, underadaptation) and more familiar (retest, overadaptation) situations might contribute to the multiple motivational symptoms in schizophrenia.