Anda di halaman 1dari 4

Physiology & Behavior 134 (2014) 6669

Contents lists available at ScienceDirect

Physiology & Behavior


journal homepage: www.elsevier.com/locate/phb

The effect of holiday weight gain on body weight


Dale A. Schoeller
Nutritional Sciences, University of Wisconsin, United States

H I G H L I G H T S

Weight gain between the Nov. to Jan. holiday season is common in Western societies.
Holiday weight gain is highly variable and averages less than 1 kg.
Behavior modications have been successful at preventing holiday weight gain.

a r t i c l e i n f o a b s t r a c t

Article history: The topic of holiday weight gain has been a frequent subject of the lay media, however, scientic interest has only
Received 1 November 2013 been recent. Multiple studies in Western societies have reported average weight gains among adults during the
Received in revised form 12 March 2014 period between mid-November and mid-January that were about 0.5 kg. The range in individual weight changes
Accepted 13 March 2014
was large, however, and the already overweight and obese gain more weight than those who are healthy weight.
Available online 21 March 2014
When the average gain across the year was also measured, the holiday weight was the major contributor to
Keywords:
annual excess weight gain. Efforts patterned to increase awareness to energy balance and body weight have
Obesity been shown to be successful at reducing such gain. An exception to holiday weight gain being a major contributor
Energy balance to annual excess gain has been children, in whom summer weight gains have been observed to be the major
Summer weight gain contributor to average excess weight gain.
Overfeeding 2014 Elsevier Inc. All rights reserved.

Holiday weight gain has long been a topic in the lay press, but it was It is important to note that the reported individual weight gains dur-
not until Yanovski et al. [1] clearly demonstrated a seasonal increase in ing the holiday period were quite variable (range 7 to 4.1 kg) [2].
weight between mid-November and mid-January, that holiday weight Some of this variation could be accounted for by baseline weight status.
gain attracted signicant scientic attention. This study measured the Although the average holiday weight gain was not reported by weight
body weight of a convenience sample of 195 racially mixed adults four status, it was reported that the relative incidence of a large weight
to six times over the course of more than a year. They documented an gain (N 2.3 kg) increased from 5% in those who were healthy weight
average holiday weight gain of 0.4 1.5 kg (mean SD), which tended (body mass index b25 kg/m2), to 11% in those who were overweight
to be greater than 0.1 1.5 kg observed during the three months prior (BMI 25 and b 30), and to 17% in those who were obese (BMI 30).
to mid-Nov. and was not compensated for by the loss of 0.1 1.1 kg In an attempt to understand the alteration in energy balance leading
during the three month period following mid-Jan. It was noted that to the holiday weight gain, a survey of change in physical activity and
the holiday weight gain was much smaller than the 2 to 5 kg holiday hunger was administered along with the quarterly weight measures.
weight gain often cited by the lay press as the typical weight gain Results, unfortunately, indicated that either energy intake or expendi-
expected by the US public [1], but it should be noted that the gain was ture could explain the weight change. Specically, self-report question-
similar to the average annual weight gain of 0.3 kg that has fueled obe- naires indicated that holiday weight gain was greatest in those that
sity epidemic beginning in 1960 [1]. In this regard, the authors reported reported a decrease in physical activity during the holiday period, and
an average annual weight gain of 0.5 kg among this cohort. Moreover, also in those reporting the greatest hunger during the holiday period.
two-thirds of this annual average weight gain occurred between mid- While weight change averaged across several hundred individuals
Nov. and mid-Jan. and thus not a weight change that should be over a three month period is usually an indication of gain of fat, there
discounted. (See Table 1.) often is variability in the fraction of weight change that is fat mass [3].
To test whether holiday weight gain did reect gain in fat mass, a
follow-up study by Hull et al. [4] included a measure of body composi-
Support: No outside support.
University of Wisconsin, 1415 Linden Dr, Madison, WI 53706, United States. Tel.: +1
tion as assessed by repeated DXA measures. The study was performed
608 262 1082. in a younger cohort (23 5 years) of 82 college students, who were
E-mail address: dschoell@nutrisci.wisc.edu. mostly (66%) healthy weight and racially mixed. Unlike the ndings of

http://dx.doi.org/10.1016/j.physbeh.2014.03.018
0031-9384/ 2014 Elsevier Inc. All rights reserved.
D.A. Schoeller / Physiology & Behavior 134 (2014) 6669 67

Table 1
Summary of holiday weight gain studies.

Reference Subjects Weight gain Comments

Shahar et al. (1999) n = 94 ca. 1.5 kg % of reported energy intake from fat increased in winter.
BMI = 26 4 kg/m2 Summer to winter
Yanovski et al. (2000) n = 195, 1982 years 0.4 1.5 kg Gain tended to incr. with BMI.
BMI = 26 5, kg/m2 Mid Nov. to Jan. Gain tended to incr. with reported
energy intake and sedentary behavior.
Hull et al. (2006) n = 82, 23 5 years N.S. weight change %fat increased, %fat tended to increase with increasing BMI.
BMI: 34% N 25 kg/m2 Mid Nov. to early Jan.
Ma et al. (2006) n = 593, 2070 years 0.5 kg greater in early Feb. Peak reported energy increase in Nov.
BMI = 27 5 kg/m2 Peak physical activity in June.
Watras et al. (2007) n = 20, 1844 years Weight tended to increase Weight increase was signicant compared to a group treated
BMI: 25 to 30 kg/m2 Sept. to Jan./Mar. with CLA for weight loss.
Branscum et al. (2010) n = 90, 3rd5th grade No increase BMI %ile increased.
BMI %ile = 74 26 Early Dec. to mid Jan Relative weight tended to increase with increasing BMI.
Cook et al. (2012) n = 443, 4069 years 0.7 1.4 kg Low energy expenditure did not predict weight gain.
BMI: 61% N 25 kg/m2 Sept./Oct. to Jan./Feb.
Stevenson et al. (2013) n = 148, 1865 years 0.8 1.3 kg Weight gain increased with BMI.
BMI = 25 5 kg/m2 Mid Nov. to early Jan. %fat increased.

Uncertainties are expressed as one standard deviation.

Yanovski et al. [1], weight change between mid-Nov. and mid-Jan. and 69 years old who were most Caucasian and roughly equally distrib-
averaged 0.1 kg, and reected the greater prevalence of healthy uted between healthy weight, overweight, and obese. These subjects
weight individuals in the Hull study. Similar to the ndings of Yanovski were an average 15 years older than those of Yanovski et al. [1], but of
et al., however, Hull et al. reported that subjects who were overweight similar weight characteristics and locale (Bethesda, MD). The average
or obese were much for likely to gain 2 kg or more (8 of 28) and holiday weight gain (0.7 1.4 kg) was somewhat larger than that reported
weight gain in that group averaged 0.8 kg compared to an average by Yanovski et al. [1], Stevenson et al. [5], and Hull et al. [4]. The objec-
weight change (loss) of 0.4 kg in those of healthy weight. Of note tive of this study by Cook et al. extended beyond documenting holiday
however, both the BMI classes of healthy weight and excess weight weight gain by testing whether baseline energy expenditure could pro-
gained fat mass (0.5 and 1.3 kg, respectively) over the holiday period in- tect against the gain. Cook et al. [6] used doubly labeled water to mea-
dicating positive energy balance in both weight status groups. Thus, the sure baseline total expenditure and it was found that neither having a
DXA measures indicated that the gain in fat mass in both groups was higher total energy expenditure adjusted for fat-free mass nor higher
fully or partially masked by a decrease in fat-free mass mostly due to estimated PAL (total energy expenditure/resting metabolic rate) at
fat-free mass loss in the legs. It is possible that this pattern of change baseline reduced the average holiday weight gain. Thus, neither a higher
in body composition is more likely among college students, given the typical energy expenditure relative to fat-free mass in the study by Cook
large amount of walking performed by college students during the on et al., nor a high self-reported level of min/wk of moderate exercise re-
campus (non-holiday) portion of the semester. In both weight status ported by Stevenson et al. [5] was protective against subsequent holiday
groups, weight change was highly correlated with change in fat mass weight gain.
(r = 0.76), and thus, while the loss of fat-free mass masked the gain Two epidemiologic studies while not keying on holiday weight gain
in fat mass among those with healthy weight, greater weight gain was per se did investigate winter weight gain. These, studies, do not present
indeed due to greater gain in fat mass and thus would raise concern data in a manner that isolate the holiday period, but they do investigate
about the health consequences of holiday weight gain over time as the seasons of the year in a not dissimilar manner. These studies report
suggested by Yanovski et al. [1]. higher weights in winter than in summer and provide additional diet
In a study by Stevenson et al. [5] that was also performed in a group and physical activity data, although both are derived from self-report.
of 148 mostly college students, it was conrmed that increases in body Shahar et al. [8] surveyed 94 adult (mean age 43 years, mean BMI
weight were indeed largely due to increases in fat mass, although 26 kg/m2) in January and July. Winter BMI values were 0.5 kg/m2
body composition change in this study was measured by bioelectrical higher, which was a difference of about 1.5 kg. Energy intake as self-
impedance, which is less precise than DXA. In this cohort of slightly reported was not signicantly different between January and July, but
older (33 11 years) subjects recruited on college campus, average tended to be 69 kcal/d higher (p = 0.39). Unfortunately, self-reported
holiday weight gain was 0.8 1.1 kg. This cohort was largely Caucasian, energy intake is known to be inaccurate [9] and thus this evidence is
two-thirds female, and 56% healthy weight, with 26% overweight and weak. Possibly, more insightful was that there were signicant increases
18% obese. More similar to Yanovski, however, all groups by weight sta- in fat intake at the expense of carbohydrate, with both saturated and
tus gained weight; however, as reported by Hull et al. [4], both groups monounsaturated fat intake increasing. These differences did predict
gained fat and the gain in body fat was greater among the obese than an increase in plasma lipids that was in the same direction of the change
among those of healthy weight (increase in percent body fat of 1.6 vs in intake. Of note, this study was performed in Israel and thus closer to
0.2 percentage points), with overweight being intermediate (+0.5 per- the equator than most of the other studies reported herein, reducing the
centage points), although not different from either of the other weight likelihood that the weight increases reported above were not driven by
status groups. Participants who reported more than 150 min/wk of season effect disorder.
moderate exercise were less likely to gain fat mass, but this difference A second epidemiologic study of winter weight change was conduct-
disappeared after controlling for baseline body weight suggesting that ed by Ma et al. [10] among subjects from Central Massachusetts. Sub-
weight status was also a predictor of exercise time. Thus the primary jects were seen quarterly. Also in this study, subjects with seasonal
predictor of holiday gain was weight status and that those with excess effect disorder were excluded, leaving 593 subjects of which 47% were
weight tended to be less likely to be exercises. female. Average age was 48 years and average BMI was 27 kg/m2. Win-
With a similar aim of investigating the role of habitual energy expen- ter (Dec 21Jan 20) weight was 0.5 kg higher than summer, which is
diture in the amount of holiday weight gain, Cook et al. [6] analyzed the very consistent with reported holiday weight gains, but it was not pos-
three month weight gain across the holiday period in the participants of sible to isolate the holiday season weight data as presented. Winter en-
the OPEN study [7]. This study was performed in 443 adults between 40 ergy intake was higher by self-report and physical activity was lower by
68 D.A. Schoeller / Physiology & Behavior 134 (2014) 6669

self-report. Again, however, both measures are of limited accuracy [9, over the following quarter(s), it appears that this average weight gain
11]. Sine wave analysis indicated that the period of greatest positive en- during the holidays is retained and probably a major contributor to
ergy balance was between Oct. 27th and Jan. 7th, which is consistent the average annual weight gain observed in the United States.
with highest weight occurring in the holiday period. Percent of energy It is worth noting that that holiday weight gain indicates only a small
from fat and saturated fat were both highest in fall, although the differ- energy imbalance. Assuming the gain of 0.5 kg was adipose tissue with
ences were less than 0.5% of energy. an energy density of 7800 kcal/kg, the energy stored over 90 days
Only one study of preadolescent children was identied. The study would be 43 kcal/d [21]. To store 43 kcal/d, however, requires an excess
by Branscum et al. [12] enrolled 90 children (35th grade, 74th BMI per- energy intake that also includes the energy expended in metabolizing
centile) for measurement of change in weight status for the 37 days en- the excess energy intake. This inefciency of energy storage depends
capsulating winter break. Weight did increase 0.6 kg, however, height on the composition of tissue accrued, the macronutrient composition
increased 0.8 cm such that there was no change in BMI percentile of the diet and other factors [22]. However, if we assume 80% of the tis-
(pre-holiday 73.9 vs. 73.3 post-holiday, NS). There was, however, a sig- sue accrued is adipose tissue, and an efciency of 0.79 can be assumed
nicant interaction with weight classication indicating a greater rela- [22], then, the positive excess energy intake would be 55 kcal/d in
tive weight gain in those children who were already classied as order to store 43 kcal/d.
overweight or obese. This pattern is similar to what had been observed While small, the holiday weight would be sufcient to explain the
in adults, but it was not possible to determine if the excess weight gain average annual weight gain that accounts for the obesity epidemic. Tak-
was similar magnitude to that observed in adults. Further research is ing the average weights of those between the ages of 20 and 29 in the
needed to identify similarities or dissimilarities between children and United States in early 1960 (NHANES I) and comparing that to the aver-
adults as this would help to determine if the factors driving holiday age weight in 6069 year olds indicate an average weight gain of only
weight gain among adults are associated with the home or other envi- 0.4 kg/year across the four decades.
ronmental factors. It is interesting to compare these observations with the results of
Although the single study above is suggestive, it is insufcient to overfeeding studies that included a measure of weight retained after
conclude that the holiday period is high risk period for weight gain in the end of the overfeeding intervention. Two studies meet these criteria
children, excess weight gain in children has been repeatedly been re- [23,24], both of which involved overfeeding by as much larger
ported during the summer period [1316]. These gains in BMI-z score 1000 kcal/d. Roberts et al. [23] overfed young males for 21 days and
between May and September [1316] suggest that the school environ- thus the total excess energy intake was almost 21,000 kcal over three
ment is far less obesogenic than is often suggested [17]. These studies weeks, which is four times that of the estimate for holiday weigh gain
do indicate, however, that the risk for excessive weight gain among chil- above and thus might be likened to a super holiday period. Not sur-
dren is greatest in the summer months and thus seasonally opposite prisingly, the weight gain in the overfeeding study was larger than the
that of adults and indicate that further public health attention is needed average holiday weight gain (2 kg). What was a surprise was that within
during the summer months. Further efforts are needed to better charac- 6 weeks after the end of the overfeeding period the participants' aver-
terize the obesogenic factors associated with excessive weight gain age weight loss was 1 kg with all subjects having a lowest weight that
among children during the summer months. Such research is needed was not different from their starting weight. The spontaneous weight
in order to determine the primary targets for interventions to prevent loss following overfeeding was due to a self-selected energy intake
childhood obesity. that was almost 500 kcal/d less than their intake at baseline as measured
As part of the frequent lay press attention to holiday weight gain, in laboratory setting. Bouchard et al. [24] overfeed for a longer period of
suggestions for prevention of the gain are also published frequently, time, 100 days, which is similar to the length of time over which holiday
but relatively few clinical trials have been reported. Never-the-less the weight gain has been documented. Due to greater total excess energy
few published trials have been positive. Boutelle et al. [18] tested a intake, the average weight gain was also greater (8.2 kg). At four
self-monitoring intervention added to standard behavioral therapy for months after the end of the overfeeding period, the weight had de-
treatment of obesity and reported that those in the intervention group creased by 7.9 kg or 1.3 kg above baseline. On an interesting note,
increased their level of self-monitoring and managed their weight bet- during the ensuing 4.5 years, their weight increased by 3.6 kg or an
ter than those performing standard behavioral modication. Recently annual weight gain of 0.8 kg, slightly larger, but not dissimilar to
Squires [19] introduced a work-place intervention involving education the annual weight that drove the obesity epidemic as calculated
on relative weight and the importance of diet and physical activity in above [24]. Neither the decrease at 4 months nor the increase over
achieving energy balance, along with web chat advisors and the addi- the ensuing 4.5 years was under the supervision of the research
tion of a low calorie lunch special in the employee cafeteria during the staff and thus classied as spontaneous.
holiday period and reported that participants actually reduced weight In both of the overfeeding studies, there was compensatory weight
by 1 kg during the holiday period. Detailed information on the partici- loss and most of the weight gain during supervised overfeeding was
pant characteristics was not presented. Finally, Watras et al. [20] con- lost. In light of these two studies, it is not clear why the holiday weight
ducted a weight loss trial of conjugated linoleic acid among forty gain as measured during the study of Yanovski et al. [1] was not also lost
adults (33 years with a BMI of 28 kg/m2) for 6 months. Conjugated during the following quarter. Indeed, based on thermodynamic models
linoleic acid is a partitioning agent that increases fat oxidation. The of energy balance, weight gains even as modest as those associated with
daily dose was 3.2 g. During the holiday period (NovDec), controls holiday periods should lead to small increases in energy expenditure
gained 0.6 kg, similar to other holiday weight gain reports; while the secondary to increased body size, however, this process is estimated
conjugated linoleic acid treatment group did not gain weight. to require three years [25,26]. Roberts et al. [23] and Bouchard et al.
Holiday weight gain among adults has been observed in a number of [24] demonstrated a faster rate of rate loss after controlled positive en-
studies. It is reported and conrmed from these studies that weight ergy balance due to excess energy intake. Of these two studies, only
change over the holiday period is highly variable between individuals, Roberts et al. included laboratory based measures of energy intake
but that there is an average weight gain over the holidays in many co- after the overfeeding and found that the weight loss resulted from com-
horts. That average gain is small with a mean of 0.5 0.4 kg in the six pensatory undereating. Neither of these studies, however, address the
diverse studies reported here. Studies that included a measure of body individual variation in weight gain during the holiday period. To this
composition indicated that the weight gain was largely due to fat gain. end, a recent study by Orsama et al. [27] reported variability in retention
It was also consistently reported that the holiday weight gain was great- of weight gained during a period as short as a weekend. Specically, it
er than 0.5 kg in those who are already obese and smaller in those who was reported that this short-term weight gain more likely to retained
are at a healthy weight. In the few studies that followed the participants among those with greater BMIs. Based on these weight retention
D.A. Schoeller / Physiology & Behavior 134 (2014) 6669 69

observations and the various holiday weight gain studies cited herein, it [11] Colbert LH, Matthews CE, Havighurst TC, Kim K, Schoeller DA. Comparative validity
of physical activity measures in older adults. Med Sci Sports Exerc 2011;43:86776.
is speculated that holiday weight gain results from a serial short-term [12] Branscum P, Kaye G, Succop P, Sharma MJ. An evaluation of holiday weight gain
weight gain during the holiday period where festive occasions are fre- among elementary-aged children. Clin Med Res 2010;2:16771.
quent. It is not known, however, if the holiday periods are also unique [13] von Hippel P, Lynch J. A simplied equation for adult BMI growth, and its use to
adjust BMI for age. Int J Epidemiol 2012;41:88890.
in terms of the macronutrient content of the festive foods, their high, [14] Moreno JP, Johnston CA, Woehler D. Changes in weight over the school year
palatability, or social factors accompanying the festive occasions. and summer vacation: results of a 5-year longitudinal study. J Sch Health
These remain important questions for further research, but results of 2013;83:4737.
[15] Kato N, Sauvaget C, Kato T. Large summer weight gain in relatively overweight pre-
multiple studies agree that in western societies, the winter holiday pe- school Japanese children. Pediatr Int 2012;54:5105.
riod is a high risk period for an average weight gain of 0.4 to 0.7 kg [16] Economos CD, Hyatt RR, Must A, Goldberg JP, Kuder J, Naumova EN, et al. Shape Up
and that gains of this magnitude are sufcient to drive the obesity Somerville two-year results: a community-based environmental change interven-
tion sustains weight reduction in children. Prev Med 2013;57:3227.
epidemic in the United States.
[17] Mannie B. Childhood obesity & school lunches. http://applevalley-rosemount.patch.
com/groups/dr-brandon-mannies-blog/p/bp-childhood-obesity-school-lunches.
References [accessed Oct. 25, 2013].
[18] Boutelle KN, Kirschenbaum DS, Baker RC, Mitchell ME. How can obese weight
[1] Nguyen TT, O'Neil PM, Sebring NG, Yanovski JA, Yanovski SZ, Sovik KN. A prospective controllers minimize weight gain during the high risk holiday season? By self-
study of holiday weight gain. N Engl J Med 2000;342:8617. monitoring very consistently. Health Psychol 1999;18:3648.
[2] Ogden CL, Fryar CD, Carroll MD, Flegal KM. Mean body weight, height, and body [19] Squires S. Controlling holiday weight gain: lessons from the Lean Plate Club. J Nutr
mass index, United States 19602002. Adv Data 2004;347:117. Educ Behav 2007;39(2 Suppl.):S512.
[3] Heymseld SB, Thomas D, Martin CK, Redman LM, Strauss B, Bosy-Westphal A, et al. [20] Watras AC, Buchholz AC, Close RN, Zhang Z, Schoeller DA. The role of conjugated
Energy content of weight loss: kinetic features during voluntary caloric restriction. linoleic acid in reducing body fat and preventing holiday weight gain. Int J Obes
Metabolism 2012;61:93743. (Lond) 2007;31:4817.
[4] Hull HR, Hester CN, Fields DA. The effect of the holiday season on body weight and [21] Heymseld SB, Thomas D, Martin CK, Redman LM, Strauss B, Bosy-Westphal A, et al.
composition in college students. Nutr Metab (Lond) Dec 28 2006;3:44. Energy content of weight loss: kinetic features during voluntary caloric restriction.
[5] Stevenson JL, Krishnan S, Stoner MA, Goktas Z, Cooper JA. Effects of exercise during Metabolism 2012;7:93743.
the holiday season on changes in body weight, body composition and blood pres- [22] Hall KD. Mathematical modeling of energy expenditure during tissue deposition. Br J
sure. Eur J Clin Nutr 2013;67:9449. Nutr 2010;104:47.
[6] Cook CM, Subar AF, Troiano RP, Schoeller DA. Relation between holiday weight gain [23] Roberts SB, Young VR, Fuss P, Fiatarone MA, Richard B, Rasmussen H, et al. Energy
and total energy expenditure among 40- to 69-y-old men and women (OPEN expenditure and subsequent nutrient intakes in overfed young men. Am J Physiol
study). Am J Clin Nutr 2012;95:72631. 1990;259:R4619.
[7] Kipnis V, Subar AF, Midthune D, Freedman LS, Ballard-Barbash R, Troiano RP, et al. [24] Bouchard C, Tremblay A, Desprs JP, Nadeau A, Lupien PJ, Moorjani S, et al. Overfeeding
Structure of dietary measurement error: results of the OPEN biomarker study. Am in identical twins: 5-year post-overfeeding results. Metabolism 1996;45:104250.
J Epidemiol 2003;158:1421. [25] Thomas DM, Weedermann M, Fuemmeler BF, Martin CK, Dhurandhar NV, Bredlau C,
[8] Shahar DR, Froom P, Harari G, Yerushalmi N, Lubin F, Kristal-Boneh E. Changes in et al. Dynamic model predicting overweight, obesity, and extreme obesity preva-
dietary intake account for seasonal changes in cardiovascular disease risk factors. lence trends. Obesity 2014;22:5907.
Eur J Clin Nutr 1999;53:395400. [26] Hall KD, Sacks G, Chandramohan D, Chow CC, Wang YC, Gortmaker SL, et al. Quan-
[9] Trabulsi J, Schoeller DA. Evaluation of dietary assessment instruments against dou- tication of the effect of energy imbalance on bodyweight. Lancet 2011;378:82637.
bly labeled water, a biomarker of habitual energy intake. Am J Physiol Endocrinol [27] Orsama A-L, Mattila E, Ermes M, van Gills M, Wansink B, Korhonen I. Weight
Metab 2001;281:E8919. rhythms: weight increases during weekends and decreases weekdays. Obes Facts
[10] Ma Y, Olendzki BC, Li W, Hafner AR, Chiriboga D, Hebert JR, et al. Seasonal variation 2014;7:3647.
in food intake, physical activity and body weight in a predominantly overweight
population. Eur J Clin Nutr 2006;60:51928.