Resting metabolic rate (RMR) of 22 individuals with Down syndrome was compared to that of 20 nondisabled control individuals of similar age (25.7 and 27.4 years, respectively). Using a ventilated hood system, we measured RMR in the early morning after an overnight fast. Peak aerobic capacity (VO2peak) and body composition were also determined. Resting metabolic rate was not different between groups. Adjusting RMR for body weight or body surface area did not change these findings. Using stepwise regression for the total population and each subgroup, we found that only body surface area was a significant predictor of RMR. These data show that individuals with Down syndrome do not have lower RMR than their nondisabled peers, suggesting that reduced RMR does not explain the high incidence of obesity in this population.
This study was supported in part by a grant from the American Heart Association and Grant H133040323 from the National Institute on Disability and Rehabilitation Research to the first author.
The most common form of developmental disability in Western society is mental retardation, with an estimated prevalence of 1% to 3% (Baird & Sadovnick, 1989; Fernhall, Pitetti, & Guerra, 2003; Roizen & Patterson, 2003). Although there are multiple causes of mental retardation, Down syndrome is the major genetic cause in North America (Baird & Sadovnick, 1989). Individuals with this syndrome have a lower life expectancy than their nondisabled peers, even in the absence of congenital heart disease (Baird & Sadovnick, 1989; Roizen & Patterson, 2003). In addition, they have an accelerated rate of common medical problems, and, after the age 40, the mortality rate is almost twice that of their nondisabled peers (Hayden, 1998; Roizen & Patterson, 2003).
Obesity is a major health problem that is more common in individuals with Down syndrome than in nondisabled populations (Pitetti, Rimmer, & Fernhall, 1993; Roizen & Patterson, 2003). This is of particular importance for individuals with Down syndrome because obesity is related to many other health consequences, including diabetes, cardiovascular disease, and cancer (Bray, 2003). Obesity is also related to reduced aerobic capacity in this population, which may contribute to sedentary life styles, poor work performance, and poor quality of life (Fernhall & Pitetti, 2001). Although the underlying etiology of obesity in persons with Down syndrome is not clear, decreased resting metabolic rate (RMR) has been suggested (Allison et al., 1995; Chad, Jobling, & Frail, 1990; Luke, Rozien, Sutton, & Schoeller, 1994). If RMR is reduced, individuals will spend less energy throughout the day compared to someone without reduced RMR, thus increasing their risk of obesity. However, not all researchers have found lower RMR in individuals with Down syndrome. Shapiro and Rapoport (1989) found no difference in basal metabolic rate (expressed per unit of fat free mass between individuals with Down syndrome and controls, but this approach of expressing the data has been criticized (Allison et al., 1995). Indeed, Allison et al. reported lower RMR in individuals with Down syndrome after covarying for fat free mass, which is statistically more correct, suggesting that even after controlling for fat free mass, RMR is reduced in persons with Down syndrome.
A careful evaluation of the available data on RMR in individuals with Down syndrome reveals several interesting factors that may have influenced the results. First, there appears to be a difference between studies on children versus adults with Down syndrome. Children with Down syndrome have consistently been found to exhibit reduced RMR compared to nondisabled controls (Chad et al., 1990; Luke et al., 1994), whereas the data on adults are more equivocal (Allison et al., 1995; Schapiro & Rapoport, 1989). Furthermore this population often exhibit hypothyroidism, which could explain their lower RMR (Allison et al., 1995; Luke et al., 1994). However, when researchers have controlled for hypothyroidism, RMR does not appear to be different between individuals with Down syndrome and controls (Allison et al., 1995). Finally, the protocols used when evaluating RMR have varied greatly among studies, with only one study group (Luke et al., 1994) using early morning measures following a 12-hour fast while measuring RMR using a ventilated hood (mouth piece or face mask methods may not be as valid). However, Luke et al. had a problem with excessive subject movement during measurements.
In addition to inconsistencies, few researchers examining RMR in individuals with Down syndrome have correctly controlled for other factors that may influence RMR, such as body weight, height, body mass index (BMI), body surface area, and fat free mass. Although the relationship between these variables and RMR are well-established in the general population, the relationship between these factors and RMR in individuals with Down syndrome has not been clarified. Also, investigators have recently shown that aerobic capacity is significantly related to RMR in a variety of populations (Forman, Miller, Szymanski, & Fernhall, 1998; Poehlman et al., 1992; Toth, Dardner, & Poehlman, 1995). Because aerobic capacity is significantly reduced in persons with Down syndrome (Fernhall & Pitetti, 2001; Fernhall et al., 1996), this reduction may also impact their RMR.
Considering the few available studies on RMR in persons with Down syndrome, and the wide variability in the methods used to evaluate it, plus the lack of information on how other variables related to RMR in the general population may influence RMR in persons with Down syndrome, we designed this study to (a) compare RMR in healthy participants with Down syndrome with normal thyroid function to age- and gender-matched nondisabled controls and (b) evaluate the relationship between factors known to affect RMR (body weight, height, body surface area, fat free mass, and aerobic capacity) in both groups and investigate whether this relationship differed between persons with Down syndrome and nondisabled controls.
Participants were 42 healthy individuals, 22 with Down syndrome (13 women, 9 men) and 20 controls (9 women, 11 men) without disabilities who volunteered for the study. They ranged in age from 17 and 39 years. All participants were sedentary or moderately active, but none were participating in any extensive exercise training. Participants were recruited from the local community and from local support groups for individuals with Down syndrome. Those with Down syndrome were classified with mild mental retardation. Before enrollment, all participants underwent a medical screening interview, which was conducted with both participants and parents for individuals with Down syndrome. We determined from their medical history that all participants were free of any overt disease. Participants with Down syndrome were included if they were apparently healthy and had undergone a thyroid screen during the past 12 months and were not diagnosed with hypo- or hyperthyroidism. They were excluded if they exhibited any of the following: (a) history of cardiovascular disease, (b) history of diabetes or other metabolic disease, (c) took medications that alter heart rate or metabolic responses, (d) were smokers, (e) had any contraindications to exercise, (f) had asthma or other respiratory disorders, (g) had congenital heart disease, or (h) had atlanto-axial instability. The nature of the study was explained and after the initial screening, all participants signed informed consent. For those with Down syndrome, parents also signed informed consent. The study protocol was approved by the Syracuse University Institutional Review Board.
Participants with Down syndrome were familiarized with all testing procedures a minimum of 2 days prior to data collection. Data were collected after each participant could satisfactorily perform each test, including comfortable walking on a motorized treadmill. The number and length of familiarization session were individualized according to the capability of each participant. Those without Down syndrome were familiarized with testing procedures on the day of testing and demonstrated they could satisfactorily complete each task before data collection.
Participants reported to the laboratory within 60 minutes of waking, following an overnight 12-hour fast. Participants were asked not to perform any exercise during the 24 hours prior to the test. They completed 30 minutes of supine rest in a dimly lit room, followed by 30 minutes of RMR measurements through indirect calorimetry for which we used a ventilated hood (SensorMedics, Yorba Linda, CA). The average of the last 20 minutes of the collection period was used as a measure of RMR.
Body weight was measured using a calibrated digital scale, and height was measured using a stadiometer. Body composition was assessed through air-plethysmograhy using the BODPOD® (Concord, CA). This method is similar to hydrostatic weighing, but we used it to measure body density through air displacement. The method is highly reliable and valid for body composition determination compared to hydrostatic weighing (Vescovi, Zimmerman, Miller, & Fernhall, 2002; Vescovi et al., 2001). We chose to measure body composition using this method because (a) of its ease of use for the participants and (b) it does not require water immersion and breath-holding.
Aerobic capacity was measured using an individualized treadmill protocol, designed for each individual, based on his or her own capabilities. The protocol started with a comfortable walk for 3 minutes followed by fast walk for 2 minutes. Thereafter, the grade was increased 2.5% every 2 minutes until a grade of 12.5% was achieved. From that point, we increased the treadmill speed 1 mph every minute until exhaustion. Participants wore a face mask and expired air was analyzed for oxygen, carbon dioxide, and volume; we used a computerized breath-by-breath metabolic system (Cosmed, Rome, Italy). Oxygen uptake (VO2) was calculated and expressed as 20-second averages. Using a Polar heart rate monitor, we measured heart rate continuously throughout the test, which was terminated when the participant could no longer keep up with the treadmill speed or showed signs of volitional fatigue. The test was considered to be a valid peak effort if VO2 plateaued (less than a 150 ml/min increase) with an increase in work rate or if there was a plateau in heart rate (less than 3 beat per minute increase) with an increase in work rate, concomitant with a respiratory exchange ratio of more than 1.0. This protocol is valid and reliable for use with participants with and without Down syndrome (Fernhall, 2003; Fernhall, Millar, Tymeson, & Burkett, 1990; Fernhall & Tymeson, 1987).
We calculated descriptive statistics for all variables. Potential group differences were evaluated using analyses of variance (ANOVA). Analyses of covariance (ANCOVA) were conducted to compare RMR between groups, controlling for body weight, BMI, body surface area, fat free mass, and aerobic capacity. Pearson correlations were calculated to evaluate the relationship between RMR and body weight, BMI, body surface area, fat free mass, and aerobic capacity. Correlations were calculated for the total sample as well as separately for each group. To evaluate which of the above variables best predicted RMR, we conducted a stepwise multiple regression using the variables described above. Statistical significance was set at the .05 level throughout.
The descriptive data are shown in Table 1. Individuals with Down syndrome and controls were of similar age, but persons with Down syndrome weighed more, were shorter, had higher BMI yet similar body surface area, and had lower aerobic capacity, p < .05. Due to equipment problems, we only had body composition data for 12 participants in each group. From this subgroup analysis, percentage of body fat was higher in participants with Down syndrome, p < .05, but fat free mass was not significantly different between groups. Unadjusted RMR was similar between groups, and adjusting for body weight and body surface area did not alter these results (see Figure 1). When BMI was adjusted, RMR was significantly greater in the controls, p < .05. However, when we adjusted for fat free mass (n = 12 for each group for fat free mass analysis), we found that RMR was significantly greater in participants with Down syndrome, p < .05 (see Figure 1). We also adjusted for gender and aerobic capacity (data not shown), and there were no differences in RMR between groups after these adjustments.
The relationships between RMR and body weight, VO2peak, body surface area, and fat free mass for the entire sample are shown in Figure 2. All the correlations were statistically significant, ps <.05. The same relationships for the group with Down syndrome are shown in Figure 3, and the scatter-plots for the control group are shown in Figure 4. For individuals with Down syndrome, body weight, body surface area, and fat free mass were significantly related to RMR, p <.05, but VO2peak was not. All of the variables were significantly related to RMR for the control group, ps <.05. The stepwise multiple regression revealed that only body surface area was retained as a significant predictor of RMR, R2 = .70, p <.05.
Our main finding in this study was that RMR was similar between individuals with Down syndrome and control participants. Covarying for body weight or body surface area did not alter these results. Although covarying for BMI elicited reduced RMR in persons with Down syndrome, we found that RMR was significantly greater in persons with Down syndrome when RMR data were adjusted for fat free mass. These data suggest that individuals with Down syndrome without hypothyroidism exhibit similar RMR to controls of similar age. Furthermore, both control participants and individuals with Down syndrome exhibited similar relationships between RMR and body weight, body surface area, and fat free mass. However, aerobic capacity was only significantly related to RMR in the control group, suggesting that aerobic capacity is not a determinant of RMR in individuals with Down syndrome.
Several investigators have shown RMR to be decreased in persons with Down syndrome (Allison et al., 1995; Luke et al., 1994), consistent with the notion that low levels of RMR contribute to the high incidence of obesity in this population. In contrast, Shapiro and Rapoport (1989) did not find reduced RMR in persons with Down syndrome, consistent with our findings. Although Shapiro and Rapoport have been criticized for expressing their data as energy expenditure per unit of body weight, fat free mass, or body surface area, their findings are similar to ours, and we did adjust for these variables using an ANCOVA, as suggested by Allison et al. (1995). Our data are actually consistent with those of Alison et al. as well. Although they found that RMR was decreased in individuals with Down syndrome (almost a 21% decrease), thyroid function was vastly different between their participants with Down syndrome and controls. After controlling for thyroid function, they found that RMR was no longer different between groups. Because all of our participants with Down syndrome were diagnosed with normal thyroid function, our results are also consistent with those of Allison et al.
Although the data from our study and findings from the literature suggest that RMR is not reduced in adults with Down syndrome with normal thyroid function, children with Down syndrome appear to have reduced RMR. The literature contains only two studies in which the researchers evaluated RMR in children with Down syndrome. Chad et al. (1990) found that children with Down syndrome exhibited reduced RMR, but they compared their findings to published norms and not directly with a control group. Luke et al. (1994) also found reduced RMR in children with Down syndrome, and they included only individuals with normal thyroid function. Consequently, available data clearly suggest that RMR is reduced in children, but not in adults with Down syndrome. Interestingly, BMI was not different between the children with and without Down syndrome in the investigation by Luke et al., showing that the reduced RMR in children with Down syndrome was not associated with obesity in their participants. However, reduced RMR in childhood is a risk factor for future weight gain (Griffiths, Payne, Stunkard, Rivers, & Cox, 1990; Ravussin, 1995; Schoeller, 1998) and may explain why adults with Down syndrome often become obese. It is not unusual to find normal RMR in obese individuals (Ravussin et al., 1988; Schoeller, 1998); thus, obesity per se is not necessarily associated with reduced RMR, but low RMR predisposes individuals to become obese (Ravussin et al., 1988).
It is not known why children, but not adults, with Down syndrome exhibit reduced RMR, but skeletal muscle hypotonicity has been suggested (Luke et al., 1994). Skeletal muscle hypotonicity is present in most children with Down syndrome (Diamond & Moons, 1961; Fernhall, 2004; Morris, Vaughan, & Vaccaro, 1982), but muscle tone improves as children mature (Parker & James, 1985). Researchers have shown that improvements in muscle tone may be related to improvements in sympathetic nerve activity and tonic sympathetic activity is related to RMR (Bell, Pitetti, Jones, & Seal, 2003; Monroe et al., 2001; Seals & Esler, 2000). Populations with low sympathetic nervous activity exhibit higher levels of obesity (Ravussin, 1995), and sympathetic activation has been shown to be reduced in individuals with Down syndrome (Fernhall & Otterstetter, 2003). Other possible explanations may include alterations in cellular metabolism, but this is probably less likely because decreases in cellular metabolism has been found in both children and young adults with Down syndrome (McCoy & Enns, 1978). Many investigators have shown that RMR is linearly related to both body weight and fat free mass, but fat free mass is usually a stronger correlate (Buchholz, McGillivray, & Pencharz, 2003; Jorgensen, Vahl, Dall, & Christiansen, 1998; Smith et al., 1997; Sparti, DeLany, de la Bretonne, Sander, & Bray, 1997; van der Ploeg & Withers, 2002). Our study is consistent with these findings, because we found that fat free mass is highly correlated with RMR. This was true for the total population (Figure 2) as well as when the data were separated by group (Figures 3 and 4). Thus, for both individuals with and those without Down syndrome, fat free mass is a major correlate of RMR, and this relationship was not altered by the presence of Down syndrome. Our findings are similar to those of Luke et al. (1994), who also showed fat free mass to be strongly correlated with RMR in children with Down syndrome. Interestingly, when we corrected RMR for fat free mass, our participants with Down syndrome exhibited slightly higher RMR compared to controls, which would suggest higher metabolic activity of fat free mass in people with Down syndrome. However, because data on fat free mass were only available on a subgroup of participants, this finding needs to be interpreted with caution.
It is interesting to note that we observed disparate results when adjusting RMR for different measures of body size (Figure 1). When RMR is adjusted for BMI, participants with Down syndrome exhibited significantly lower values compared to controls. However, because individuals with Down syndrome are typically short, as was the case in our study, BMI may not be the best measure of body size in this population (Guerra, Llorens, & Fernhall, 2003). Instead, it may be more appropriate to adjust RMR for body surface area, as has been shown in other studies (Refsum, Holter, Lovig, Haffner, & Stadaas, 1990; Shetty, 1984; Swinamer et al., 1990), as body surface area is often used as a covariate or denominator for various metabolic measurements. In our study, body surface area was not different between groups, and adjusting RMR for body surface area did not produce any group differences. Thus, when adjusting for an appropriate measure of body size, RMR was not reduced in participants with Down syndrome. In fact, adjusting our data either for body weight, body surface area, or fat free mass showed that participants with Down syndrome did not exhibit reduced RMR. A shown in Figures 2, 3, and 4, body surface area was highly correlated with RMR in the total population and in the separate group analyses. In addition, only body surface area was retained as a significant predictor of RMR in the multiple regression analysis, showing the efficacy of adjusting RMR for body surface area.
Several researchers have reported a positive relationship between RMR and VO2peak (Ballor & Poehlman, 1992; Forman et al., 1998; Toth et al., 1995), suggesting that higher aerobic capacity is related to higher RMR. Others investigators have failed to show such a relationship (Broeder, Burrhus, Svanevik, & Wilmore, 1992; Smith et al., 1997). In our study, VO2peak was significantly related to RMR in the control participants, r = .77, p < .05, but not in the individuals with Down syndrome, r = .27, suggesting that aerobic capacity differentially affects RMR in controls and participants with Down syndrome. However, VO2peak was not included as a significant predictor in the multiple regression analyses, showing that when accounting for body surface area, aerobic capacity was not related to RMR. This finding is similar to the results reported by Smith et al. (1997), who found that VO2peak was not independently related to RMR in young adult women. Our data suggest that reduced RMR does not account for the high incidence of obesity in these adults with Down syndrome. It is likely that overfeeding is a major contributor to the high obesity rates, concomitant with low levels of physical activity, as suggested by others (Allison et al., 1995; Chad et al., 1990; Luke et al., 1994; Pitetti et al., 1993). There are probably several factors contributing to possible overfeeding and decreased physical activity, but the deinstitutionalization movement may actually have contributed. Cross-sectional evidence suggests that as people with intellectual disabilities moved from very controlled, state-owned facilities to the least controlled community dwellings, rates of obesity increased (Pitetti et al., 1993). Pitetti et al. (1993) hypothesized that increased food availability, less supervision, and less direction have contributed to both overfeeding and decreased physical activity. Recent data on activity levels of persons with Down syndrome living in the community support this notion; most participants were classified as sedentary (Pitetti et al., 1993).
Our findings suggest several important considerations that may have policy implications. First, for adults with Down syndrome, the high rates of obesity and increasing rates of weight gain with age are probably not a result of decreased metabolism in this population, but are likely to be related to lifestyle choices, such as poor diets and sedentary lifestyles. Thus, in the future, it is important to develop appropriate programs aimed at better eating habits and increased physical activity levels for individuals with Down syndrome. Second, based on our data and those of other researchers, thyroid function clearly impacts RMR in this population. Thus, more frequent check-ups and better medical management of thyroid function is an important considerations for individuals with Down syndrome. This point is especially important because the incidence of thyroid disease in adults with Down syndrome is over 15% (Roizen & Patterson, 2003), and has been reported to be as high as 30% (Pozzan et al., 1990). Finally, it is possible that RMR is reduced in children with Down syndrome, predisposing them to become obese adults. This topic needs much more investigation in order to enhance our understanding of this phenomenon and, ultimately, to design effective intervention programs aimed at reducing the level of obesity in persons with Down syndrome.
In summary, our data show that individuals with Down syndrome do not exhibit reduced RMR and that correlates of RMR are similar in controls and persons with Down syndrome. Multiple regression analysis showed body surface area was most strongly related to RMR in both groups. Consequently, the high incidence of obesity observed in populations with Down syndrome is most likely due to life style factors, at least in those with normal thyroid function.
Authors: Bo Fernhall, PhD (email@example.com), College of Applied Life Studies, University of Illinois at Urbana-Champaign, 1206 S. Fourth St. MC-586, Champaign, IL 61820. Arturo Figueroa, MD, PhD, Scott Collier, MS, Styliani Goulopoulou, MS, Ifigenia Giannopoulou, PhD, and Tracy Baynard, MS, The Human Performance Laboratory, Exercise Science Department, Syracuse University, Syracuse, NY 13244