Parents of children with poor growth with or without underlying medical conditions often complain about their children not having appetite. These children do not ask for food, eat small quantities, and consume a limited variety of foods, resulting in difficult mealtimes and poor growth (1). Despite efforts at improving mealtime behaviors, giving nutritional advice, and involvement of occupational therapy, these children often remain poor feeders and either do not gain weight (2) or gain some weight with extreme maternal efforts. Cyproheptadine (CY), a known antiserotoninergic and antihistaminergic agent (3,4), has been used as an appetite stimulant (5,6) and in patients with gastrointestinal problems (7), dyspepsia (8), anorexia nervosa (9), pulmonary consumptions (eg, cystic fibrosis) (10–12), asthma (13), cancer and treatment-related cachexia (13,14), and for stimulant-induced weight loss in children with attention-deficit/hyperactivity disorder (15). The exact mechanism by which CY improves growth is not well understood; it may affect appetite via receptors in the ventromedial hypothalamus through its antihistamine and antiserotonin function (16), or it may influence the growth hormone and insulin-like growth factor axis (17). Several investigations have attempted to understand the mechanisms of action of CY as an appetite stimulant (17,18). Although the efficacy and safety of this medication have been documented in the pediatric population, its use in poorly growing children age ≤2 years and on specific feeding behaviors, has not been investigated. Although rare, the most common adverse effects reported are mild drowsiness and/or irritability that tend to wear off after the first few days of use.
We hypothesized that the use of CY in young children with poor growth is effective and safe and that this medication improves feeding behaviors and weight gain. To test our hypothesis we performed a retrospective analysis of all of the patients treated with CY at the pediatric feeding program (PFP) at the Montreal Children's Hospital.
Population and the PFP
From January 2007 to July 2011 the medical records of all of the patients studied at the PFP with poor growth and treated with CY were reviewed. This multidisciplinary clinic includes pediatric psychologists, occupational therapists, dietitians, a behavioral specialist, and a pediatric gastroenterologist. The PFP follows patients with feeding disorders secondary to poor appetite (with or without poor growth), difficulties with food textures and variety, and patients with enteral feedings who require the discontinuation of gavage. In this study, all tube-fed patients were excluded. The program receives >200 new consults per year, with initial weekly or biweekly follow-up by the psychologists. Behavioral and interactional approaches along with input from other team members are used to improve mealtime behaviors and achieve weight gain. The study was approved by the institutional review board of the Montreal Children's Hospital.
PFP Interventions and Management
Anthropometric measurements were taken at the time of initial assessment and at follow-up appointments. For children of age <2 years, the weight and height were taken without clothing and using an infant scale (Seca Model 727, Hamburg, Germany) and a wooden length board. Children of age >2 years were weighed and measured in light clothing without shoes using a standing Detecto scale (Detecto, Webb City, MO). At each consult, the weight was plotted on the Centers for Disease Control and Prevention (CDC) growth chart (19). The length was not measured systematically in the follow-up visits. As a first-line treatment, the psychologist provided psychological education and behavioral modification, with nutritional and/or occupational therapists’ involvement as needed. If there was no weight gain or the weight gain continued to be minimal, and the behavioral approach did not influence mealtime behaviors after a period of 1 to 2 months, CY was prescribed as an adjunct therapy at a dose of 0.25 mg · kg−1 · day−1 body weight (divided into 2 times per day). During the first week, any reaction to the medication was closely monitored by telephone followed by a clinic visit within 2 to 4 weeks. Follow-up appointments ensured that the dosage and cycling of the medication remained effective while at the same time mealtime behaviors, food quantity, and variety were reinforced. The cycling of the medication depended on the child's sensitivity to the medication and varied from 1 to 3 weeks of use and 4 days to 1 week of pause. This regimen was based on the observation made by our senior author in the last 25 years that a pause in taking the medication renewed interest in eating, which was followed by weight gain.
Clinical Variables and Weight Measurements
The following data were extracted from the medical records: maternal age and education level, child's age, sex, and medical conditions, age CY started, reaction to the medication, and changes in mealtime behaviors (as spontaneously reported by parents). For each patient, weight was recorded at an age that was as close as possible to 6 months before the initiation of the medication to ensure that poor weight gain was long term. All of the children had their weight recorded at the time the medication was started and as close as possible to 3 and 6 months after initiation of CY. All weight measurements were converted into weight-for-age z scores (WtZ) for each of the 4 time points of weight measurement (ie, 6 months prior, at time treatment started, 3 and 6 months during treatment), using the nutrition program of EpiInfo version 184.108.40.206 (EpiInfo, Centers for Disease Control and Prevention, Atlanta, GA), based on the CDC growth reference (19,20).
In addition, based on their age at the time of initiation of CY, the children were divided into 3 age groups: 7 to 18, 19 to 30, and 31 to 80 months. These groups are based on guidelines for the development of feeding skills, as well as the understanding that the younger the child is, the more difficult it is to influence mealtime behaviors when his or her appetite is low.
Repeated-measures mixed model was used to analyze the pattern of change in mean WtZ over time (SPSS version 20.0, IBM SPSS Statistics, Armonk, NY). Analyses were performed within groups (pretreatment and during treatment time points) and between groups (regularly receiving CY vs comparison group; age and medical conditions groups). Post-hoc analyses were performed using repeated-measures analysis of variance between and within measurement time points to test for differences in mean WtZ across groups.
During the study period between 2007 and 2011, 941 children between ages 7 months and 6 years were referred to our PFP clinic for poor growth, gavage feeding, poor food texture tolerance, and/or difficult mealtime behaviors. Sixty percent of the children were referred by community pediatricians and 35% by hospital specialty clinics. Of the 941 children, 152 (16%) received CY: 127 with the goal of improving weight gain and 25 with the goal of increasing food texture tolerance or aiding the training off gavage feedings. All of the tube-fed children were excluded from this study. Of the 127 patients in treatment owing to poor weight gain who received CY, 82 took the medication regularly and were considered in our analyses as an experimental group. Of the remaining 45 patients, 8 parents stopped giving CY within 1 week because, according to them, their children became too irritable or sleepy; 7 parents gave CY irregularly, 9 parents never used CY stating that they rather got involved with exceptional efforts and time to increase food intake, and 21 did not return after the prescription was given. These 45 patients were therefore used in our analyses as a naturally conceived comparison group. The demographic characteristics of these 82 patients who received the medication regularly and the comparison group are presented in Table 1. The demographic characteristics did not differ between the experimental (n = 82) and the comparison groups (n = 45). As Table 1 indicates, 24 patients of the experimental group were healthy and 58 had a medical condition. Of these 58 patients, 22 were receiving medication for gastroesophageal reflux and/or delayed gastric emptying.
Of the 127 patients who received CY, a total of 18 patients presented with adverse effects (14%). As stated above, significant adverse effects were reported by parents of 8 children, whereas parents of 6 children reported mild irritability and 4 parents reported that their children were sleepier than usual. These latter 2 adverse effects were, however, mild and wore off within a few days.
Of the 82 parents whose children continued treatment regularly, 96% reported a change in their children's feeding behaviors. The specific behavioral changes are detailed in Table 2.
Analyses of mean WtZ changes were performed for the experimental and the comparison groups. Because WtZ did not differ between boys and girls, a patient's sex was not considered in the following analyses. For the experimental group (n = 82), there were fewer children at the second measurement time point during treatment (n = 54), as they were discharged or the family decided to terminate further visits when growth improved and mealtimes became easier. For the comparison group (n = 45), weight measurements were available in the medical records of our hospital for 33 patients in the first and 13 patients in the second time points during treatment.
Pretreatment weight measurement time point was on average 5.5 ± 1.1 months (range 3–7 months) before medication was received, whereas first and second measurement time points during the course of the treatment were on average 3.0 ± 0.7 months (range 1–4 months) and 6.5 ± 1.2 months (range 5–10 months) after treatment started, respectively.
For the group of patients who received CY regularly, a significant change in mean WtZ was observed over time (F = 19.2, df = 216.5, P < 0.00). The pattern of change in mean WtZ after treatment started was linear, with a considerable increase at the first time point during treatment. The mean WtZ was significantly higher at the first (F(1,81) = 120.9, P < 0.00) and second time points during treatment (F(1,53) = 32.2, P < 0.00), when compared with the mean WtZ at the time CY started. There was no significant difference in the mean WtZ between the 2 time points during treatment (F(1,53) = 2.9, P = 0.09).
For the comparison group, no significant differences in mean WtZ were observed over time (F = 0.1, df = 90.6, P = 0.96). As shown by Figure 1, the only time point of measurement with a significant difference in mean WtZ between the experimental and the comparison groups was at the time medication started (F(1,126) = 4.6, P < 0.05). Therefore, the mean WtZ of the patients who took the medication regularly was significantly lower than the mean WtZ of the comparative group at the time CY was prescribed.
Within the group of patients with medical conditions who regularly received CY (n = 54), there was no difference in mean WtZ between patients who were taking other medications and those who were not at each of the 4 time points (all F ≤ 1.9, P ≥ 0.17). There was also no significant difference in mean WtZ for all 4 measurement time points between patients with or without medical conditions who regularly received CY (all F ≤ 3.1, P ≥ 0.09). Finally, Figure 2 illustrates that there was a significant difference for WtZ among the 3 age groups within the patients who regularly received the medication: at the time CY started and at the first time point during treatment, the mean WtZ of the youngest group was significantly lower than that of the 31- to 80-month-old group (all F(1,81) ≤ 4.1, P < 0.05); at the second time point during treatment, the mean WtZ of the youngest group was significantly lower than those of the 2 older groups (F(1,53) = 12.2, P < 0.00).
The results of this retrospective review demonstrate that the use of CY under the supervision of a specialized PFP was safe and effective in infants and young children with poor growth. In combination with a multidisciplinary approach, CY administration resulted in significant changes in feeding behaviors and weight gain in children with or without medical conditions, and this effect was independent of the chronological age at which the medication was initiated. The major limitation of this study was its retrospective design, with all of its limitations in terms of data collection. As an example, the length was not measured systematically in the follow-up appointments and could not be used to show the effect of CY. This anthropometric measure will be used in future research projects to better evaluate the results of CY.
We were able to use a quasicontrol group when we analyzed the available data of the group of children who did not use CY regularly or never used the medication. We found that the experimental group had poorer WtZ than the comparison group at the time the medication was received, but not at the pretreatment time point (3–7 months before the treatment). This suggests that the group who took CY regularly—experimental group—was composed of patients whose weight-for-age was severely declining with time. Therefore, parental compliance to the treatment may be related to the severity of the WtZ at the time CY was prescribed.
There was no sex difference on the degree of weight gain in our study population of children of age <6 years. Twice as many children had medical conditions in our study, indicating that these children are more likely to have comorbid feeding difficulties and poor weight gain. Our results indicate that the youngest age group had the lowest rate of weight gain before treatment with CY, supporting the notion that when young children have diminished appetite they refuse to eat regardless of parental efforts. The 2 older groups of children included some children with borderline daily rate of weight gain, relative to the minimum recommended by the CDC (19,20). These children were included because their borderline rate of weight gain was the result of lengthy mealtimes and extreme parental efforts at getting their child to eat. As the results suggest, not only did these older children respond well to the administration of CY in terms of weight gain but also their mealtime behaviors improved substantially. Based on these results, we have developed a short questionnaire regarding patients’ feeding behaviors to be completed by the mothers during their first visit after starting the medication. The results of this questionnaire will be reported in a separate article.
In our routine practice we used a daily dose of 0.25 mg/kg, which is >0.1 mg/kg used by Mahachoklertwattana's group (17), but <0.3 mg/kg used by Rerksuppaphol and Rerksuppaphol (21). The effect of our dosage was sustained for an extended period owing to continuous adjustment to the increased weight of the child, as well as to the cycling of the medication. The continuous effect of the medication when it was cycled was observed by the senior author and has been used as the standard of care in our institution. The younger the children were, the more caution was applied in administering CY. Most of the adverse effects were prevented by starting an initial daily dosage of one-fourth of the daily dosage required for all of the children, and the parents were required to call the clinic during the first 4 days to report any adverse effect. If there was any adverse effect, the dosage was decreased for that child. In our study, children who were receiving antireflux medication and/or motility agents also experienced a significant increase in their rate of weight gain when treated with CY, indicating that those medications did not influence the appetite-stimulating properties of CY. In conclusion, although children tolerated 0.25 mg · kg−1 · day−1 of CY, we were more cautious with children of age <2 years because the effects were not known.
Overall, our results have demonstrated that a combined multidisciplinary approach and the use of an appetite stimulant can significantly improve appetite, feeding behavior, and weight gain. It has proved to be safe to use even for younger children of age <2 years. Parental compliance to the treatment appears to be higher for patients whose weight gain is severely declining over time.
The authors thank Dr Guilherme Sant’Anna for the extensive review of the manuscript.
1. Tannenbaum GS, Ramsay M, Martel C, et al. Elevated circulating acylated and total ghrelin concentrations along with reduced appetite scores in infants with failure to thrive. Pediatr Res
2. Wright CM, Parkinson KN, Drewett RF. How does maternal and child feeding behavior relate to weight gain
and failure to thrive? Data from a prospective birth cohort. Pediatrics
3. Konstandi M, Dellia-Sfikaki A, Varonos D. Effect of cyproheptadine hydrochloride on ingestive behaviors. Pharmacol Res
4. Konstandi M, Kolijianni A, Sfikakis AD. Effect of cyproheptadine treatment on conditioned avoidance response in female rats. Gen Pharmacol
5. Silverstone T, Schuyler D. The effect of cyproheptadine on hunger, calorie intake and body weight in man. Psychopharmacologia
6. Kaplowitz PB, Jennings S. Enhancement of linear growth and weight gain
by cyproheptadine in children with hypopituitarism receiving growth hormone therapy. J Pediatr
7. Andersen JM, Sugerman KS, Lockhart JR, et al. Effective prophylactic therapy for cyclic vomiting syndrome in children using amitriptyline or cyproheptadine. Pediatrics
8. Rodriguez L, Diaz J, Nurko S. Safety and efficacy of cyproheptadine for treating dyspeptic symptoms in children. J Pediatr
9. Powers PS, Santana C. Available pharmacological treatments for anorexia nervosa. Expert Opin Pharmacother
10. Chinuck RS, Fortnum H, Baldwin DR. Appetite stimulants in cystic fibrosis: a systematic review. J Hum Nutr Diet
11. Epifanio M, Marostica PC, Mattiello R, et al. A randomized, double-blind, placebo-controlled trial of cyproheptadine for appetite stimulation in cystic fibrosis. J Pediatr
12. Homnick DN, Marks JH, Hare KL, et al. Long-term trial of cyproheptadine as an appetite stimulant in cystic fibrosis. Pediatr Pulmonol
13. Lavenstein AF, Dacaney EP, Lasagna L, et al. Effect of cyproheptadine on asthmatic children. Study of appetite, weight gain
, and linear growth. JAMA
14. Couluris M, Mayer JL, Freyer DR, et al. The effect of cyproheptadine hydrochloride (periactin) and megestrol acetate (megace) on weight in children with cancer/treatment-related cachexia. J Pediatr Hematol Oncol
15. Daviss WB, Scott J. A chart review of cyproheptadine for stimulant-induced weight loss. J Child Adolesc Psychopharmacol
16. Sakata T, Ookuma K, Fukagawa K, et al. Blockade of the histamine H1-receptor in the rat ventromedial hypothalamus and feeding elicitation. Brain Res
17. Mahachoklertwattana P, Wanasuwankul S, Poomthavorn P, et al. Short-term cyproheptadine therapy in underweight children: effects on growth and serum insulin-like growth factor-I. J Pediatr Endocrinol Metab
18. Calka O, Metin A, Dulger H, et al. Effect of cyproheptadine on serum leptin levels. Adv Ther
19. Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al. CDC growth charts: United States. Adv Data
20. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC Growth Charts for the United States: methods and development. Vital Health Stat 11
21. Rerksuppaphol S, Rerksuppaphol L. Effect of cyproheptadine on weight gain
in malnourished children: a randomized, controlled trial. Asian Biomed