Factors Associated with Low Body Mass Index in Huntington's Disease: A Spanish Multicenter Study of the European Huntington's Disease Registry

Abstract

Background

Patients with Huntington's disease (HD) are at risk for body weight loss and increased risk for institutionalization, morbidity, and mortality. The aim of this study was to determine the factors associated with low body mass index (BMI) in patients with HD.

Methods

In this national, observational, cross‐sectional study of the European Huntington's Disease Network, the frequency of food consumption, calories, and nutrient intake in patients with HD was assessed using questionnaires validated for the Spanish population and were calculated using the software package Alimentación and Salud (Diet and Health), version 2.0. Nutritional status was estimated using the BMI, and disease severity was assessed using the Unified Huntington's Disease Rating Scale and a total functional capacity (TFC) score. Linear regression models were performed using BMI as the dependent variable and using energy balance (energy caloric intake − energy expenditure); the TFC score; the presence of a caregiver; dysphagia; cytosine, adenine, guanine (CAG) repeats; comorbidities; intake of supplements; pharmacologic treatments; age; gender; education; and physical activity as the independent variables.

Results

Two hundred twenty‐four patients with HD were included (59% women), and their mean age was 47.41 ± 14.26 years, a median TFC score of 9 (range, 3–13), normal BMI in 124 patients (55.4%), and low BMI in 13 patients (6.7%). In the linear regression model, older age (β = 0.003; P = 0.01), male gender (β = 0.13; P = 0.003), and lower energy balance (β = −0.0001; P = 0.0003) were associated with a higher log‐transformed BMI.

Conclusions

Younger female HD patients are at risk for low BMI. To counteract the influence of the HD gene mutation on decreased BMI, an increase in kilocalories per day should be encouraged.

1.

Patients with neurodegenerative diseases, and particularly those with Huntington's disease (HD), are at risk of unintentional body weight loss and the subsequent increased risk of institutionalization, morbidity, and mortality.1 In HD, it is not uncommon to observe a decrease in anthropometric measurements, such as the body mass index (BMI) of multifactorial origin, during the course of the disease.2 Weight loss frequently leads to general weakening and a decline in the quality of life of patients with HD3; conversely, a higher BMI has been associated with a slower rate of disease progression.4 The cause of weight loss in HD is unknown, but the most likely contributing factors are sympathetic hyperactivity and the signaling provided by insulin5, 6; chorea, which might involve significant energy expenditure; chewing and swallowing difficulties; and an intrinsic hypermetabolic state.2 On the other hand, the use of neuroleptics can facilitate weight gain and increase the risk for obesity and metabolic syndrome.7 Because many factors contribute to weight loss in HD, the aim of this study was to analyze the factors associated with low BMI after adjusting for confounding variables.

Patients and Methods

Design

This was an Observational, cross‐sectional, national multicentre study of patients participating in the European Huntington's Disease Registry. We have reported a more detailed account of the background and methods of the survey and of the study population elsewhere.8 Briefly, the participants were selected from a Spanish cohort of patients who participate in the European Huntington's Disease Network (EHDN) registry study.9 The registry is a multicenter research observational project for individuals affected by HD. Inclusion criterion was adult‐onset HD mutation carriers in a borderline to advanced stage. All participants signed an informed consent form to participate. For participants who lacked the capacity to consent, study sites adhered to specific guidelines, and a legal tutor signed the consent form.

The study was approved by the Ethics Committee of Complejo Universitario Burgos (Burgos, Spain). Patients who were diagnosed with HD and individuals who were HD premanifest gene carriers and were included in the Spanish EHDN study were invited to participate in the study during the period from 1 June 2012 to 31 August 2013. Participants who were identified as premanifest HD mutation carriers were those rated by a movement disorder specialist as not meeting the clinical criteria (a motor Unified Huntington's Disease Rating Scale [UHDRS] score of either 0 or 1, indicating the presence of nonspecific motor signs for HD). Those participants with higher motor UHDRS scores were classified as manifest HD mutation carriers. To facilitate patient recruitment, this study was announced in different Spanish neurological and patient association meetings from 2011 to 2013. Patient anthropometrics and information on clinical rating scales were obtained during their annual visit to the European Huntington's Disease Registry.

Assessments

Anthropometrics and Nutritional Status

Anthropometrics, such as BMI and weight/circumference, were collected as nutritional status estimators. In accordance with World Health Organization international standards,10 a BMI ranging from 18.5 to 24.9 kg/m² was considered nutritionally normal, whereas a BMI <18.5 kg/m² was associated with a deficient nutritional status, and a BMI >25 kg/m² was associated with overweight and obesity.10 The cutoff value for waist circumference associated with obesity for the European/North American population was >102 cm for men and >88 cm for women.11, 12 Dietary intake was assessed by 24‐hour recall and 3‐day dietary record questionnaires that were validated for the Spanish population.13 For individuals who completed the 3‐day recall questionnaire, an additional measure of adherence to the traditional Mediterranean diet was assessed using a 10‐point Mediterranean diet scale that incorporated the salient characteristics of this diet in which scores ranged from 0 to 9, with higher scores indicating greater adherence.14 Data on macronutrients, micronutrients, and energy calorie intake obtained from the dietary intake questionnaires were calculated using the software package Alimentación and Salud (Diet and Health), version 2.0 (Institute of Nutrition, Granada, Spain).

In individuals whose weight is stable, caloric intake should equal total energy expenditure (TEE). To calculate the energy needs of this population, total energy expenditure (TEE) was based on basal energy expenditure, which represents the minimal energy required for body vital function maintenance, plus the thermal energy of food, and physical activity. Because indirect calorimetry was not performed to estimate basal energy expenditure, we considered the caloric intake an estimate of TEE.15 The energy expenditure was calculated with the Harris Benedict equation, which originally defines the energy expenditure for body vital function maintenance,15 including a modified factor of 1.72 to account for food processing and a midstage‐HD physical activity.16 Therefore, TEE was calculated as follows: (1) men: 66.4730 + 13.7516 × weight in kilograms + 5.0033 × height in centimeters − 6.7550 age in years; and (2) women: 655.0955 + 9.5634 × weight in kilograms + 1.8496 × height in centimeters − 4.6756 age in years. Energy balance was calculated as energy caloric intake − TEE.

HD Severity and other Sociodemographic Clinical Data

Participants who were identified as premanifest HD mutation carriers were rated by a movement disorder specialist as not meeting the clinical criterion, which was a motor UHDRS score of either 0 or 1, indicating the presence of nonspecific motor signs for HD.17 Participants who had higher motor UHDRS scores were classified as manifest HD mutation carriers. Duration of HD and data on CAG repeats were obtained from the EHDN database.9 Comorbidity information was collected using the Cumulative Illness Rating Scale‐Geriatric, with higher scores indicating higher comorbidity.18 In this study, the sample was operationally classified as having a high rate of comorbidities based on published cutoff scores >8.19 Disease severity was assessed using a total functional capacity (TFC) score,20 which was derived from reports by the participant and his or her companion and quantified the ability of a participant to perform both basic and instrumental activities of daily living. In this measure, scores range from 0 to 13, with higher scores indicating more intact functioning. The severity of psychiatric symptoms was assessed using the Problems Behavior Assessment Scale‐short form (PBA‐S), in which higher scores indicate greater severity.21 Dysphagia was assessed using the Eating Assessment Tool (the EAT‐10 questionnaire) validated for the Spanish population with a cutoff score of ≥3 for dysphagia.22 Caregiver burden was assessed using the Caregiver Burden Inventory,23 with higher scores indicating higher caregiver burden; and quality of life was assessed using the Medical Outcomes Study 36‐item Short Form Health Survey,24 with higher scores indicating higher quality of life.25 The level of physical activity was assessed using the Global Physical Activity Questionnaire developed by the World Health Organization, which comprises 19 questions and calculates physical activity in terms of high, moderate, and low levels of physical activity.26 Intake of nutritional supplements included percutaneous endoscopic gastrostomy.

Statistical Analysis

Data Collection

BMI as an Independent Variable

During their annual visit to the EHDN registry, body length and weight were measured to compute the BMI.

Fixed Regression Parameters

Fixed regression parameters were gender, age (years), education (≤11/>11 years), and low‐level physical activity. These demographic factors were selected because of their commonness and the amount of evidence on their association with BMI that existed before our study.

Backwards Regression Parameters

The following backwards regression parameters were included in the linear regression model: energy balance, TFC, the presence of a caregiver, CAG repeats (large allele), presence of dysphagia, presence of high comorbidity, intake of nutritional supplements, and intake of antidopaminergics or antiepileptics. The selection was based on previous evidence (higher comorbidity, TFC score, CAG repeats, and pharmacologic treatments),2, 4, 7 and the investigators’ criteria (energy balance, the presence of a caregiver and dysphagia, and intake of nutritional supplements). In this regard, low BMI should be the logical consequence of an energy expenditure that is higher than the energy consumption, which is a negative energy balance. However, the association between energy balance and low BMI in individuals with HD had not been quantified and described. Likewise, an attentive caregiver may help to overcome nutritional obstacles. Finally, we also were interested in investigating the extent to which nutritional supplements might be able to prevent low BMI.

Statistical Analysis and Presentation of Results

The participant flow from recruitment to analysis was presented as a flow diagram. To analyze the characteristics of participants from the Spanish EHDN study compared with those from the non‐Spanish EHDN, the TFC score and BMI also were compared. Comparing participants and nonparticipants, BMI was described as the mean and standard deviation, as well as the median, the lower quartile, and the upper quartile (the interquartile range). The statistical software package IBM‐SPSS version 19 (IBM Corporation, Armonk, NY) was used for data analysis.

The fixed and backwards regression and descriptive parameters were described in separate tables comparing complete and incomplete cases and nonparticipants. Binary parameters were described as counts and percentages, age was described as means and standard deviations, and other parameters were described as medians and interquartile ranges.

We studied histograms of BMI, backwards and fix regression parameters, as well as scatterplots of univariate associations with BMI. We transformed the parameters to maximize the linearity of the associations. The fixed and backwards regression parameters were regressed against logarithmized BMI as dependent variable. Case wise deletions were adopted for missing values. The regression coefficients of the final regression model, their standard errors, P values, and 95% confidence intervals were presented in a table. Finally, the biologic and clinical relevance of factors that were statistically significant at an α level of 0.05 was estimated using the antilogarithm of the linear regression equation.

Results

Characteristics of Participants and Nonparticipants

Of 445 HD gene carrier participants in the EHDN Registry from 22 centers in Spain in 2012, in total, of 224 subjects (50.3%) from 10 centers (45% participation) were included in the study (168 manifest HD, 56 premanifest HD), and 114 had complete clinical information (98 manifest HD, and 16 premanifest HD) (Fig. 1). Compared with nonparticipants patients from the Spanish EHDN, non‐significant differences were found in terms of TFC (P = 0.15), or BMI (P = 0.39). Likewise, compared to the Spanish participants of the EHDN, non‐Spanish participants of the EHDN (n = 5304), had similar BMI (P = 0.35), but lower TFC (P < 0.0001) (Table S1).

Figure 1.

Flow of patients asked to participate in Spanish centers of the European Huntington's Disease Registry.

Characteristics of the Included Participants

Information on BMI was available for all participants. In terms of age, the mean age varied between 46 and 49 years across complete cases, participants excluded from the analysis, and nonparticipants (Table 1). The proportion of persons with a TFC score > 6 was similar in nonparticipants and participants who were excluded from the analysis but was lower in complete cases (Table 2). Manifest HD participants were more frequent in the group of participants who were included in the analysis (Table 3). Tables S2 and S3 list the basic clinical and demographic characteristics of the participants.

Table 1.

Fixed Regression Parameters in Complete and Incomplete Cases and Nonparticipants

Parameter	Participants		Nonparticipants, N = 221
	Included in the Analysis, N = 114	Excluded from the Analysis, N = 110
Women: No. (%)	67 (58.8)	63 (57.3)	124 (56.4)
Age: Mean ± SD, y	49.2 ± 15.5	45.6 ± 12.7	48.4 ± 14.3
Education >11 y: No. (%)	44 (38.6)	54 (54)a
Low physical activity: No. (%)	61 (53.5)	64 (59.3)b

^aTen participants had missing values (the denominators of percentages are the nonmissing values).

^bTwo participants had missing values.

SD, standard deviation.

Table 2.

Backwards Regression Parameters in Complete and Incomplete Cases and Nonparticipants

Parameter	Participants		Nonparticipants, N = 221
	Included in the Analysis, N = 114	Excluded from the Analysis, N = 110
Energy balance: Median [lower, upper quartile], kcal/d	438.1 [−135.9, 1,176.4)	409.7 [−179.1, 1,089.4]
Total functional capacity >6: No. (%)	59 (51.8)	65 (70.7)a	155 (70.8)b
Presence of a caregiver: No. (%)	58 (50.9)	43 (47.3)c
Count of CAG repeats (large allele): Median no. [lower, upper quartile]d	44.0 [42.0, 46.9]	42.5 [40.0, 46.3]
Presence of dysphagia: No. (%)	53 (46.5)	30 (36.1)e
>2 Points on the comorbidity index: No. (%)	26 (22.8)	25 (27.8)f
Nutritional supplements: No. (%)	38 (33.3)	24 (22.9)g
Antidopaminergics or antiepileptics use: No. (%)	85 (74.6)	43 (72.9)h

^aEighteen participants had missing values (the denominators of percentages are the nonmissing values).

^bOne nonparticipant had a missing value.

^cNineteen participants had missing values.

^dCAG repeats were included in the model as 1 divided by CAG.

^eTwenty‐seven participants had missing values.

^fTwenty participants had missing values.

^gFive participants had missing values.

^hFifty‐one participants had missing values.

CAG, cytosine, adenine, guanine.

Table 3.

Descriptive Parameters in Participants Included and Excluded from the Analysis

Parameter	Included in the Analysis, n = 114	Excluded from the Analysis, n = 110
Manifest Huntington′s disease: No. (%)	98 (88.3)a	70 (73.7)b
Duration of manifest disease: Median [lower, upper quartile], yc	3.0 [1.0, 6.75]	2.0 [0.0, 6.75]
Current smoker: No. (%)	33 (28.9)	25 (28.4)d
Antiepileptics: No. (%)	25 (21.9)	11 (19)e
Benzodiacepins: No. (%)	54 (47.4)	22 (37.3)f

^aThree participants had missing values (the denominators of percentages are the nonmissing values).

^bFifteen participants had missing values.

^cThere were no patients with nonmanifest disease.

^dTwenty‐two participants had missing values.

^eFifty‐two participants had missing values.

^fFifty‐one participants had missing values.

Dietary Intake, Nutritional State, and Severity of HD in Participants

Dietary information (24‐hour recall questionnaire) was available for 198 participants (88.3%), and TFC scores were available for 216 participants (96.4%). Only 13 participants (5.80%; 95% CI 2.51–9.08) had a low BMI. Characteristics of the dietary composition in terms of macronutrients/micronutrients, dietary pattern, environmental characteristics, and nutritional status information of our sample have been published elsewhere.8

BMI and energy/calorie intake were similar when participants in the premanifest HD group were compared with participants who had manifest HD [BMI: 23 kg/m² (interquartile range, 26.25; 21.1323 kg/m²) vs 23.83 23 kg/m² (interquartile range, 27.25; 21.3623 kg/m²), respectively; P = 0.33; energy/calorie intake: 1893 ± 599.58 kcal/day vs 2084.25 ± 701.71 kcal/day, respectively; P = 0.12]. BMI was not significantly associated with UHDRS motor scores, cognitive scores, quality of life, PBA (psychosis, anger, and executive) dysfunction scores, disease duration, or comorbidity. BMI was weakly correlated with the PBA depression score (ρ coefficient = 0.23; P = 0.001), education (ρ coefficient = −0.19; P = 0.004), and energy balance (ρ coefficient = −0.34; P = 0.0001). Patients with HD who had higher educational background, women, patients on antidopaminergic or antiepileptic drugs, and patients with lower physical activity were had low BMI. Smokers also had a trend toward lower BMI (Table 4).

Table 4.

Clinical Characteristics of Participants (N = 224)

Characteristic	Body Mass Index Comparison
	No. of Samples	Median (IQR)	P
Current smoking			0.07
No	144	24.25 (21.79; 26.93)
Yes	58	22.72 (20.23; 26.25)
Gender			0.03
Men	94	24.26 (26.58; 22.49)
Women	130	22.86 (26.77; 20.59)
Presence of dysphagia: EAT‐10 score >10			0.63
No	114	23.96 (26.49; 21.43)
Yes	83	23.54 (27.55; 20.5)
Presence of caregiver			0.25
No	104	24.95 (21.51; 26.81)
Yes	101	24.23 (20.67; 26.72)
High education background: >11 y			0.03
No	116	24.42 (21.87; 27.46)
Yes	98	23.11 (21.02; 25.57)
Total functional capacity score >6			0.15
No	82	22.87 (20.41; 26.77)
Yes	124	23.85 (21.77; 26.72)
Manifest HD			0.62
No	45	23.21 (21.24; 26.43)
Yes	160	23.71 (21.30; 26.98)
Physical activity level			0.30
Low	125	23.53 (20.81; 27.27)
Medium	65	23.67 (21.18; 26.14)
High	32	24.05 (22.71; 27.03)
High comorbidity: CIRS‐G score >8			0.76
No	153	23.54 (21.29; 26.62)
Yes	51	23.92 (20.58; 27.48)
Use of amantadine			0.96
No	135	23.74 (26.99; 21.22)
Yes	38	23.88 (27.05; 21.67)
Use of antidopaminergics
No	55	25.54 (28.01; 22.68)	0.006
Yes	118	23.34 (26.28; 20.97)
Use of antidepressants			0.11
No	80	24.55 (27.63; 21.59)
Yes	92	23.66 (26.18; 20.92)
Use of antiepileptics			0.01
No	136	24.48 (27.38; 21.44)
Yes	36	22.08 (24.47; 20.63)
Use of benzodiazepines			0.58
No	97	23.92 (26.98; 21.19)
Yes	76	23.92 (27.17; 21.35)
Use of nutritional supplements			0.09
No	142	24.16 (27.34; 21.38)
Yes	30	22.90 (25.89; 20.45)

IQR, interquartile range; EAT‐10 the Eating Assessment Tool questionnaire; HD, Huntington's disease; CIRS‐G, Cumulative Illness Rating Scale‐Geriatric.

Likewise, compared with patients who had HD with central obesity, those without central obesity had similar HD severity [median TFC score, 7.5 (interquartile range, 3;12) vs. 9 (interquartile range, 4.7;13); P = 0.15, respectively] and intake of antidopaminergic drugs (33 patients with central obesity [62.3%] vs 85 patients without central obesity [70.8%]; P = 0.29). Waist circumference was not correlated with CAG repeats (ρ coefficient = −0.30; P = 0.68).

Factors Associated with BMI in HD

Table S4 shows the univariate linear regression model conducted in the group of complete cases (n = 114). In the multivariate linear regression model, older age (β = 0.003; P = 0.01), male gender (β = 0.13; P = 0.003), and lower energy balance (β = −0.0001; P = 0.0003) were associated with a higher log‐transformed BMI (Table 5). This model indicated 0.003 log‐BMI points more for each additional year of age, a 1.14 log‐BMI point decrease for female gender, and a 0.00015 log‐BMI point decrease for each kilocalorie (kcal) of energy balance.

Table 5.

Multivariate Backwards Linear Regression of Log‐Transformed Body Mass Index (N = 114)^a

Variable	Coefficient	Standard Error	P	95% Confidence Interval
				Lower Limit	Upper Limit
Constant	−3.096	0.067	0.000	2.964	3.228
Age, y	−0.003	0.001	0.010	0.001	0.005
Women	−1.140	0.043	0.003	4.630	22,222
Education >11 y	−0.032	0.036	0.381	−0.104	0.040
Low physical activity	−0.029	0.035	0.415	−0.098	0.041
Energy balance, kcal/d	−0.00015	0.000	0.000	0.000	0.000

^aThis model was adjusted for age, gender, education background, level of physical activity, energy balance, total functional capacity, presence of caregiver, 1/CAG (cytosine, adenine, guanine) repeats, dysphagia, intake of supplements, and use of antidopaminergic/antiepileptic drugs. Univariate backwards linear regression analyses are shown in Table S3 (see supporting information).

Antilog transformation led to the following formula:

$$\begin{array}{cc}\hfill \text{BMI}=& \frac{1.003(\mathrm{age}\mathrm{in}\mathrm{years})}{1.00015(\mathrm{energy}\mathrm{balance}\mathrm{in}\mathrm{kcal}/\mathrm{day})}\hfill \\ & \times 22.109(\mathrm{for}\mathrm{women})\mathrm{or}25.204(\mathrm{for}\mathrm{men})\hfill \end{array}$$

For example, a change in BMI from 22 to 23 kg/m² corresponded to an increase in age of 14.8 years as well as a decrease in energy balance of 300 kcal per day.

Discussion

Low BMI is not a consistent feature of HD. In this clinical, national, ambulatory HD sample, only a small percentage of patients were underweight compared with other HD studies that had a higher frequency of low BMI.1, 2, 27 We found that women and younger individuals were associated with lower BMI. Likewise, BMI was not associated with HD severity measured by the TFC, PBA, UHDRS cognitive and motor scores, presence of dysphagia, and carrier status (premanifest vs manifest HD). Our results obtained from the Spanish EHDN were representative of the entire EHDN population.

These exploratory findings deserve a confirming study, because they suggest that low BMI and subsequent decline is not caused by a decreased calorie intake. On the contrary, in agreement with other studies,1, 2 HD patients with lower BMI seem to have a higher calorie intake compared with those who have a normal BMI, suggesting that lower BMI in HD is more likely caused by higher basal energy expenditure. Previous studies have investigated the effects of the HD mutation on the hypothalamic‐pituitary axis and energy homeostasis and have demonstrated mild hyperactivity of the thyrotrophic axis, disturbed regulation of the lactotropic axis, changes in the regulation of growth hormone and ghrelin secretion, and lower insulin sensitivity.5, 27, 28

In contrast to other studies, we did not observe an association between BMI and the number of CAG repeats.2 However, why was the prevalence of low BMI relatively low in our HD sample? The most likely explanation would be lifestyle (adherence to the Mediterranean diet, physical activity), environmental conditions, or the presence of family caregivers. Conversely, in a recent study analyzing the anthropometrics characteristics of 3200 Spanish Caucasian subjects,29 the prevalence of overweight and obese individuals was relatively higher (46.8%) compared with our sample (37.9%), suggesting that rapid changes in the prevalence of obesity in the general population might counteract the HD mutation. Supporting our results, other cross‐sectional studies conducted in different countries have also found a low prevalence of low BMI in patients with HD.16, 30, 31 However, we cannot exclude the possibility that the lack of an association of BMI with CAG repeats could be due to the size of our sample.

Interestingly, contrary to expectations, the use of antidopaminergic drugs and antiepileptics was associated with lower BMI compared with their use in other pathologies, in which neuroleptics have been associated with a higher risk for metabolic disorders, including obesity.32 Surprisingly, waist circumference, which is classically considered a better index for the metabolic state,33 was similar when the group of participants on antidopaminergics was compared with those not taking antidopaminergic drugs, indicating that the waist circumference was less sensitive compared with the BMI in this population. Because of the cross‐sectional design of this study, we could not establish a relationship between BMI or waist circumference and the intake of antidopaminergic drugs, and only longitudinal data may help to resolve these discrepancies.

According to Nance and Sanders,3 patients with HD require a diet that provides from 3000 to 4000 kcal/day to maintain or increase body weight, although there is little scientific evidence for this specific range.34 In the present study, with a mean energy calorie intake of 2062.86 ± 679.75 kcal/day, most of our patients were able to maintain their BMI, indicating that there is no need to add oral nutritional supplements to their regular diet. In this regard, in small human sample and animal model studies of dietary interventions in HD using a ketogenic diet,35 supplements with extra calorie intake,1 triheptanoin,36 and extra virgin olive oil in conjunction with hydroxytyrosol to reduce lipid peroxidation37 have been studied with moderate success.

Limitations of the current study include the lack of corrections or queries for missing data, the use of a clinic‐based sample, and the use of a cross‐sectional design, which makes it more susceptible to confounding variables. Conversely, strengths of the current study include results based on a representative sample of the EHDN community, the exhaustive collection of environmental characteristics, and variables that impact on BMI, such as dysphagia, eating patterns, and physical activity, using validated questionnaires.

In conclusion, younger female patients with HD are at risk for weight loss. To counteract the influence of the HD gene mutation, an increase of at least 300 kcal/day for a 1‐point decrease in BMI should be encouraged. Other factors, such as lifestyle, adherence to the Mediterranean diet, and other environmental variables, deserve further investigation as protective factors to prevent weight loss in patients with HD.

Author Roles

1. Research Project: A. Conception, B. Organization, C. Execution; 2. Statistical Analysis: A. Design, B. Execution, C. Review and Critique; 3. Manuscript Preparation: A. Writing the First Draft, B. Review and Critique.

E.C.: 1A, 3A

J.R.: 1B, 1C, 3B

N.M.: 2C, 3C

A.M.: 1C, 3B

D.A.: 1C, 3B

R.J.C. 1C, 3B

Disclosures

Funding Sources and Conflicts of Interest: This study was sponsored by the European Huntington Disease Network (EHDN, Seed Fund 338) and by Nutricia and Advanced Medical Nutrition Fresenius‐Kabi Pharmaceuticals. Data were analyzed independently by the funding sources. Esther Cubo received travel funding from Abbvie, Allergan, and UCB Pharmaceuticals and received research support from the Movement Disorder Society, the World Federation of Neurology and Junta de Castilla y León. Rafael J. Cámara received travel compensation from the European Huntington's Disease Network and the Federal Office for Radiation Protection of Germany. The remaining authors report no sources of funding and no conflicts of interest.

Financial Disclosures for the previous 12 months: The authors reported no disclosures.

Appendix

Spanish members of the European Huntington Disease Network who participated in this project: Itziar Gastón Zubimendi (Hospital Virgen del Camino, Pamplona); Dolores Martínez (Hospital Virgen del Camino, Pamplona); Maria Antonia Ramos (Hospital Virgen del Camino, Pamplona); Ines Legardaba (Hospital Universitario Son Espases, Mallorca); Barbara Vives Pastor (Hospital Universitario Son Espases, Mallorca); Raquel Pérez (Centro de Rehabilitación el Hayedo, Madrid); Juan Antonio Burguera (Hospital La Fé, Valencia); Francisco Casterá (Hospital la Fé, Valencia); Misericordia Floriach Robert (Hospital Mare de Déu de la Mercé, Barcelona); Jose M Ruiz (Hospital Mare de Déu de la Mercé, Barcelona); Jesús Pérez and Saül Martínez‐Horta (Hospital de la Santa Creu I San Pau, Barcelona); Esteban Muñoz (Hospital Clinic, Barcelona); Carmen Durán (Hospital Infanta Cristina, Badajoz); José Manuel García (Hospital Virgen Macarena, Sevilla); Carolina Méndez (Hospital Virgen Macarena, Sevilla); María Fuensantas Noguera (Hospital Virgen de la Arrixaca, Murcia); María Angeles Acera (Hospital Cruces, Bilbao); and Juan Carlos Gomez Esteban and Koldo Berganzo Gonzalez (Hospital Cruces, Bilbao).

Supporting information

Table S1. Clinical Characteristics of Participants Versus Nonparticipants

Table S2. Clinical and Demographic Characteristics*

Table S3. Nutrition and Environmental Characteristics

Table S4. Univariate Backwards Regression Analysis of Log‐Transformed Body Mass Index (N = 114)