Restoring body weight stands as an urgent therapeutic imperative in the clinical management of severe anorexia nervosa, yet establishing the precise energetic threshold necessary to rebuild depleted tissues has challenged clinicians for decades. In a retrospective observational study conducted at Kitakyushu Medical Prison and analyzed through the Department of Psychosomatic Medicine at Kyushu University, researchers discovered that female patients recovering from severe anorexia required an average excess caloric intake of 12,776.8 kilocalories to achieve a single kilogram of weight gain. This figure substantially exceeds historical caloric assumptions, pointing toward an elevated metabolic hurdle during nutritional rehabilitation. Rather than adhering to conventional formulas derived from healthy populations, the clinical data demonstrate that the physiological cost of physical recovery varies markedly across individuals and is strongly dictated by baseline bodily tissue distribution.
The quantitative ambiguity surrounding nutritional rehabilitation has persisted despite prominent international guidance outlining general rates of recovery. Guidelines issued by the National Institute for Health and Care Excellence have historically recommended a target weight gain between 0.5 and 1.0 kilogram per week, calculated upon an estimated theoretical surplus of roughly 3,500 to 7,000 extra kilocalories weekly. Prior research had reported widely conflicting caloric targets for single-kilogram gains, complicated by disparate study cohorts that rarely exceeded twenty individuals. By examining an institutional setting characterized by continuous supervision, the Kyushu University investigative team sought to eliminate confounding behavioral variables that frequently undermine nutritional monitoring in conventional inpatient facilities.
The study protocol evaluated forty-one female patients treated within the specialized medical prison between February 2013 and July 2018, selected from an overall cohort of sixty-four individuals detained with verified diagnoses under the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition. The facility structure provided an unprecedented window into biological dynamics because strict custodial oversight made surreptitious food disposal, covert self-induced vomiting, and excessive exercise nearly impossible. Unlike typical hospital wards where patients can evade behavioral protocols, this rigorous setting isolated purely biological energetic equations from behavioral interference.
Nutritional rehabilitation adhered to a structured cognitive-behavioral framework coupled with graded behavioral limitation. Initially, patients were confined to their rooms in recumbent posture until physiological vitals and dietary tolerances stabilized, receiving a baseline refeeding regimen calibrated to approximately 30 kilocalories per kilogram of body weight each day. To prevent lethal refeeding syndrome, clinicians administered an oral liquid supplement supplying 100 kilocalories per day alongside concentrated phosphorus and essential trace elements for approximately one month. Dietary intake was cautiously escalated in 200-kilocalorie increments as tolerated, never exceeding 2,100 kilocalories daily, a cap matching standard guidelines for healthy adult Japanese women. Intravenous infusions were restricted to supportive situations, representing just 1.2 percent of aggregate calories across the cohort and concentrating primarily within three individuals, while nasogastric feeding tubes were avoided entirely.
To rigorously trace metabolic expenditure over time, the team employed multifrequency bioelectrical impedance analysis using an InBody 770 device to chart body mass, total body water, extracellular water, soft lean mass, and body fat mass. Assessments were conducted weekly across the initial six weeks, transitioning to biweekly intervals through week twelve, and monthly thereafter. Daily excess energy intake was quantified by subtracting daily resting energy expenditure, derived through Scalfi’s validated formulation for female anorexia patients multiplied by a standardized sedentary activity factor of 1.3, from actual ingested calories. Linear regression lines plotting cumulative excess energy against changes in measured mass yielded the excess energy needed for one kilogram of weight gain, denoted as EE1. The analysis isolated the active refeeding phase from subsequent plateau periods, defined as intervals where mass remained static or fell for more than two months.
Baseline characteristics underscored a uniquely vulnerable and chronic cohort whose demographics differed from typical adolescent presentations. Three-quarters of the enrolled women presented with the binge-eating and purging subtype of anorexia nervosa, and the mean cohort age was 44.05 years with a standard deviation of 8.38 years. At enrollment, participants exhibited an average body weight of 32.22 kilograms, a mean body mass index of 13.12, and an average body fat percentage of only 7.04 percent. The duration of active weight restoration averaged 302 days, reflecting prolonged inpatient care relative to average Japanese psychiatric admissions. Over the entire therapeutic course, thirty-eight participants successfully gained more than 5 kilograms, and twenty-five attained a healthy body mass index exceeding 18.5 before discharge, although one individual gained only 0.2 kilograms over a 651-day admission.
Statistical evaluation identified two extreme outliers exhibiting values greater than three times the seventy-fifth percentile, leaving thirty-nine participants for final energy calculations. Among these patients, the mean excess energy necessary to deposit one kilogram of tissue reached 12,776.8 kilocalories, with a median of 12,120 kilocalories and a standard deviation of 3,334.8 kilocalories. Substantial interpersonal variation emerged across the cohort, with individual requirements ranging between 6,636 and 22,064 kilocalories per kilogram. This variation corresponded to a coefficient of variation of 26.1 percent, a figure noticeably tighter than the 25.4 to 95.4 percent ranges documented across prior smaller investigations. The finding underscores that even when physical activity, dietary compliance, and purging are strictly governed, metabolic demands diverge widely among individual patients undergoing identical therapeutic protocols.
Investigation into physiological markers revealed striking associations between caloric needs and baseline body composition rather than diagnostic categories. The team found no statistically significant correlation between EE1 and anorexia subtype, nor did initial body mass index dictate energy costs. Instead, significant associations emerged across four core tissue metrics: absolute body fat mass, body fat percentage, soft lean percentage, and the ratio of body water to total mass. When dividing patients across interquartile thresholds, participants presenting with an initial body fat mass of 3 kilograms or less exhibited significantly higher EE1 values than those above that mark. Similar disparities appeared for body fat percentages at or below 8 percent, soft lean percentages exceeding 85 percent, and body water ratios surpassing 0.665, confirming that tissue proportions govern the energetic cost of recovery.
These observations indicate that the ratio between adipose stores and metabolically active lean mass drives the elevated caloric threshold. In severely depleted individuals where adipose depots have been virtually exhausted, these vital viscera represent an unusually large proportion of remaining total mass, dramatically inflating resting energy expenditure per kilogram of living tissue. When nutrition is reintroduced, individuals possessing a low fat-to-lean proportion burn considerable energy maintaining organ integrity and resisting protein catabolism.
Several parallel physiological mechanisms help explain why recovering anorexia patients require vastly more caloric surplus than the approximately 10,500 kilocalories per kilogram observed during overfeeding trials in healthy adults. Prior research demonstrates that resting energy expenditure climbs disproportionately in recovering patients relative to raw gains in fat-free mass. Furthermore, diet-induced thermogenesis—which accounts for roughly ten percent of daily energy dissipation—can become markedly accentuated following meal ingestion in malnourished states. In addition, established alterations in the intestinal microbiota of anorexia patients, which have been shown in animal fecal transplants to diminish nutritional extraction efficiency, likely depress net energy absorption and elevate downstream caloric demands.
The study also highlights meaningful parallels with earlier Japanese hospital literature while clarifying discrepancies found in Western studies. The observed mean requirement of 12,777 kilocalories per kilogram aligns closely with findings from Yamashita and colleagues, who monitored Japanese patients over extended admissions using dual-energy X-ray absorptiometry. Both trials evaluated cohorts consuming diets calibrated close to standard maintenance volumes over months of observation, contrasting with historical Western studies where participants received aggressive caloric loads between 3,500 and 5,500 kilocalories daily over shorter timeframes. Conversely, earlier small comparisons suggesting healthy male controls demanded more energy than anorexia patients suffered from unbalanced sex distributions and limited sample sizes, confounding broader physiological comparisons.
Despite its unique behavioral controls, the authors noted several key methodological limitations that contextualize their conclusions. The investigative team did not utilize indirect calorimetry or wearable actigraphy, creating potential discrepancies between estimated total energy expenditure and true biological energy expenditure. Furthermore, Scalfi’s predictive equation for resting metabolic rate was originally developed and validated for women under thirty years of age, meaning its application to a cohort averaging forty-four years carried unmeasured estimation error. The older demographic profile also means the findings cannot be generalized directly to adolescents and young adults who comprise the traditional epidemiological peak of the disorder. Finally, because several participants never entered an extended weight plateau, data from stabilization phases were omitted, underscoring the need for larger prospective metabolic investigations.
Subject of Research: Caloric intake requirements and body composition determinants of weight gain in hospitalized patients with anorexia nervosa
Article Title: Factors related to the amount of energy required for weight gain in patients with anorexia nervosa under strict behavioral control: a study in a Japanese medical prison
Article References: Suzuyama, C. A., Takakura, S., Takii, M., Kishimoto, J., Toda, K., Yamashita, M., Hata, T., & Sudo, N. (2025). Factors related to the amount of energy required for weight gain in patients with anorexia nervosa under strict behavioral control: a study in a Japanese medical prison. BioPsychoSocial Medicine, 19(1), Article 9. https://doi.org/10.1186/s13030-025-00332-0
Image Credits: AI Generated
DOI: 10.1186/s13030-025-00332-0
Keywords: anorexia nervosa, weight restoration, caloric intake, energy expenditure, body composition, bioelectrical impedance analysis, body fat mass, soft lean mass, resting energy expenditure, diet-induced thermogenesis, refeeding, Kitakyushu Medical Prison, Scalfi formulation, malnutrition, clinical nutrition, body water ratio, hypermetabolism
News Source: Daisy Hatcher. (November 11, 2026). Severe Anorexia May Require Massive Caloric Surpluses for Weight Restoration. Scienmag.



