INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE ASSOCIATION BETWEEN HYPERTHYROIDISM and increased energy expenditure (EE) is well accepted (10, 16). There is a paucity of data, however, regarding how thyroid hormone excess influences spontaneous physical activity and nonexercise activity thermogenesis (NEAT). Patients with hyperthyroidism exhibit a characteristic resting tremor and self-report increased generalized physical activity, heat intolerance, and weight loss (13). When this is coupled to information that small, involuntary fidgeting-like movements in humans significantly increase EE by 20-50% above resting values (8), we became intrigued as to whether inducing hyperthyroidism would result in increased spontaneous physical activity and NEAT. This is of interest not only from a clinical perspective but also to gain insight into a mechanism by which NEAT is modulated (11). If hyperthyroidism proved to be associated with increased spontaneous physical activity and NEAT, it might help us understand how NEAT affects energy balance (7). On the other hand, if inducing hyperthyroidism proved not to be associated with increased spontaneous physical activity and NEAT, it might suggest that thyroxine (T₄) is not pivotal for modulating spontaneous physical activity and/or NEAT. To this end, we devised a study to address the hypothesis that hyperthyroidism is associated with increased spontaneous physical activity and activity EE. To address this hypothesis, spontaneous physical activity and activity EE were measured in rats that had been rendered hyperthyroid and in euthyroid controls.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. Twelve male Sprague-Dawley rats (Harlan Laboratories, Indianapolis, IN), 10-12 wk of age and weighing 300-350 g, were studied. Animals were delivered to our facility 7-10 days before study for acclimatization to our animal environment.

day 2

day 2

Protocol. On day 1, 24-h baseline measurements of EE and activity were performed as described below. On the morning of day 2, after these measurements were completed, animals were randomized to receive either infusion with thyroid hormone (n = 6; hyperthyroid group) or vehicle (n = 6; control group). Each animal then had a subcutaneous osmotic pump (Alza Scientific Products, Palo Alto, CA) implanted over the scapulae to provide either triiodothyronine (T₃; 200 µg/day, Spectrum Quality Products, Gardena, CA) or an isovolumic infusion of vehicle [20 mM NaOH, 50 mM Na₂CO₃, and 5% (wt/vol) bovine serum albumin]. Infusions of T₃ or vehicle lasted from day 2 through day 15 (14 days). On day 14, 24-h measurements of EE and activity were repeated, and on conclusion of these measurements (day 15) the animals were killed by intraperitoneal injection of pentobarbital sodium. Blood was drawn by cardiac puncture for measurement of T₃ and T₄ concentrations by using competitive chemiluminescent immunoassay with the ACS-180 automated immunoassay system (Bayer Diagnostics, Tarrytown, NY).

EE. For days 1 and 14, 24-h EE was calculated from measurements of O₂ consumption and CO₂ production obtained by using a customized, high-precision, single-chamber indirect calorimeter (Columbus Instruments, Columbus, OH) (2, 14). The calorimeter was housed in a sound- and lightproof, purpose-built room that received filtered air. The light-dark cycle, temperature, and humidity mimicked the conditions the animals were housed in throughout the experiment. Calibration of the calorimeter was performed before each measurement by use of a primary standard span gas (0.501% CO₂, 20.53% O₂), and 100% N₂ and cross-calibrations were performed against room air (reference gas for inspired O₂ and CO₂ fractions) every 60 min. Repeated measures of butane recovery for O₂ consumption for the calorimeter were 98.9 and 99.0% and for CO₂, 99.1 and 99.1%; the mean sign-corrected residuals for respiratory quotient were 0.3 ± 0.2% of total; and 25-h sensor drift was <0.3%.

Physical activity. Spontaneous physical activity was measured simultaneously with the EE measurements during days 1 and 14. Measurements were performed using customized, high-precision racks of collimated infrared activity sensors (Columbus Instruments) placed around the acrylic chamber. There were 45 collimated beams of infrared light crossing the 30-cm-diameter cage, allowing the detection of 1 cm of movement in three orthogonal axes. Photosensors registered an activity unit each time a beam was interrupted. In this fashion, activity was simultaneously detected in all three axes: forward-and-backward, side-to-side, and up-and-down. Data for spontaneous physical activity were summed for every minute and stored on the PC with use of the time stamp for identification. Data were thereby derived simultaneously for EE and physical activity, for each animal, minute-by-minute over the 24-h measurement period (Fig. 1).

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Example of energy expenditure and physical activity data stream for a rat. Data are averaged for each 10 min. AU, activity units.

Statistical analysis. Simultaneous total EE (kcal/h) and physical activity [activity units (AUs)] values were averaged for each minute of the 24-h measurement period. Activity EE was defined as the mean EE for when the activity sensors read greater than zero. Nonactivity EE was defined as the mean EE for when the activity sensors read zero. Data are expressed as means ± SD. Within- and between-group analyses were performed by using ANOVA with post hoc testing using paired and unpaired t-tests, respectively. When appropriate, linear regression was used. Statistical significance was defined at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In the hyperthyroid group, the mean T₃ concentration was 755 ± 137 (range 574-919) ng/dl and in the control group, 59 ± 0.5 (58-59) ng/dl. T₄ concentrations were appropriately suppressed (<1 µg/dl) in all of the animals in the hyperthyroid group and normal in the controls (2 ± 0.2 µg/dl). Thus each animal in the hyperthyroid group had been rendered hyperthyroid, whereas the controls remained euthyroid.

Animals in the hyperthyroid group failed to gain weight compared with controls. Over the 14-day experimental period, weight gain in the hyperthyroid group was 3 ± 21 g and 56 ± 10 g in the controls (P < 0.0001; Table 1). Food intake in the hyperthyroid group increased from 23.0 ± 2.8 to 36.4 ± 5.4 g/day (P = 0.003), whereas food intake in the controls did not increase significantly (Table 1).

Total daily EE increased to a greater magnitude in the hyperthyroid rats compared with the controls. In the hyperthyroid group over the 14-day experiment, total daily EE increased from 29 ± 2 to 41 ± 4 kcal/day, whereas total daily EE in the controls increased from 28 ± 1 to 31 ± 1 kcal/day (P = 0.04). The increase in total daily EE for the hyperthyroid group (13 ± 4 kcal/day) was fivefold greater than for the controls (2 ± 2 kcal/day; P < 0.001). The increase in total daily EE in the hyperthyroid group (posttreatment over baseline) was still evident when EE was expressed relative to body weight (Table 2). The expected increase in total daily EE in the control animals is likely to have reflected their increasing body size (Table 2).

To address our primary hypothesis, we wanted to know whether spontaneous physical activity increased with hyperthyroidism. Over the 14-day treatment period, spontaneous physical activity increased in the hyperthyroid rats (from 24 ± 7 to 36 ± 6 AU/min) but did not increase in the controls (23 ± 7 vs. 22 ± 4 AU/min) (Table 2). Interestingly, the change in activity in the hyperthyroid rats was predominantly through increased vertical (rearing) activity. We do not have objective data as to what specific behavior the increased rearing represented. It is unlikely to have been associated principally with increased food gathering because the food bowl height and diameter did not necessitate substantial rearing. Also, rats in general spill their pelleted food onto the cage floor, and eating in this fashion would not have triggered the vertical sensors. In the hyperthyroid rats, the fraction of the day spent at rest (0 AUs) decreased over the 14-day experiment from 0.38 ± 0.07 to 0.28 ± 0.06 (P = 0.01), whereas the controls showed no change (0.41 ± 0.08 vs. 0.41 ± 0.08). Thus hyperthyroidism was associated with increased physical activity and less rest.

We next wanted to address the hypothesis that the increased spontaneous physical activity in the hyperthyroid rats was associated with increased activity EE. We were able to separate the EE of physical activity (>0 AUs) from nonactivity EE (0 AUs) and so could calculate activity EE and nonactivity EE. Over the 14 days, daily activity EE increased in the hyperthyroid rats from 8 ± 3 to 20 ± 5 kcal/day (P < 0.001) but did not increase in the controls (9 ± 4 vs. 9 ± 2 kcal/day; not significant) (Fig. 2A). Over the 14 days, daily nonactivity EE did not change significantly in either the hyperthyroid (21 ± 2 vs. 22 ± 2 kcal/day) or the control rats (19 ± 2 vs. 21 ± 1). Changes in activity EE therefore accounted for the majority of the changes in total EE that accompanied thyroid hormone excess (94 ± 24%). For the 12 animals we studied, the change in EE correlated with the change in physical activity (r = 0.68; P = 0.02). Moreover, the fraction of the time spent active correlated significantly with activity EE (r² = 0.7) and with total EE (r² = 0.4). Overall, hyperthyroidism appeared to be associated with increased activity EE.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Activity energy expenditure (EE) and nonactivity EE (means ± SD) in rats randomized to receive 14 days of continuous infusion with inactive vehicle (controls; n = 6) or triiodothyronine (hyperthyroid; n = 6). A: total daily EE apportioned as nonactivity EE or activity EE. B: time spent per day, in hours, engaged in nonactivity (rest = 0 activity units) and activity (>0 activity units). C: energy costs of nonactivity per hour and activity per hour. Baseline vs. posttreatment, *P < 0.05, **P < 0.005, ***P < 0.001. Hyperthyroid vs. controls, #P < 0.05, ##P < 0.005, ###P < 0.001.

Was the increased activity EE seen in the hyperthyroid rats because they increased their duration of physical activity, expended more energy per unit time of physical activity, or both? To answer this, we calculated the activity EE and nonactivity EE per unit time; the data are shown in Fig. 2, B and C. The hyperthyroid rats not only were active for longer each day as described above but also their EE per unit time while they were active was substantially greater than that of the controls (1.1 ± 0.2 vs. 0.7 ± 0.1 kcal/h, P < 0.001). Also, as expected in the hyperthyroid rats, nonactivity EE per unit time was significantly greater than in the controls (3.0 ± 0.7 vs. 2.2 ± 0.4 kcal/h, P = 0.04). Thus the increased activity EE seen in the hyperthyroid rats was associated with both greater time per day spent active and greater EE per unit time while active.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Hyperthyroidism is known to be associated with elevated EE in both humans and animals (11, 16). However, little is known regarding the mechanism by which this occurs. Patients with thyroid hormone excess have tremor, and many patients report increased spontaneous physical activity, heat intolerance, and weight loss (13). Measuring spontaneous physical activity and NEAT in humans and animals is complex and is likely to be the explanation for why so little information is available. Technological advances, however, now enable this question to be addressed. The question is important not only for understanding the negative energy balance that accompanies thyrotoxicosis but also for understanding how NEAT impacts the modulation of energy balance. The present study addressed the hypothesis that hyperthyroidism is associated with increased spontaneous physical activity and NEAT. To investigate this, we rendered rats hyperthyroid by using continuous infusions of T₃, measured physical activity and EE, and compared the results with euthyroid controls infused with inactive vehicle. Thyroid hormone excess was associated with significant and overt increases in spontaneous physical activity and NEAT.

The data we gathered strongly support our primary hypothesis that spontaneous physical activity and NEAT are increased with hyperthyroidism. First, total EE increased in the hyperthyroid animals over the 14-day experimental period to a fivefold greater degree than in controls. Second, physical activity increased in the T₃-treated animals but did not increase in the controls. Third, the time spent motionless decreased in the hyperthyroid rats but not in the controls. Fourth, activity EE doubled in the T₃-treated animals but did not increase in the controls. Fifth, we were able to identify that not only were the hyperthyroid rats active for longer each day but also they expended more energy per unit time of activity. Finally, there was a direct correlation between the change in physical activity and the change in total EE.

Do the changes we describe in activity EE correspond to changes in NEAT? NEAT is an entity described in humans (4) and is akin to the EE of spontaneous physical activity. For a human it would be the EE of performing all of our daily tasks such as walking, talking, yard work, and fidgeting. For a free-living rat, NEAT would include scavenging for food, grooming, and habitat building. For a laboratory rat, NEAT principally includes grooming, walking around the cage, rearing, the locomotion of feeding, and other less definable movements such as head shaking. In this experiment, we deliberately did not provide the rats with an exercise wheel. We did provide them with relatively large cages, however. We were specifically interested in the effects of thyroid hormone excess on spontaneous physical activity rather than exercise-related activity. If exercise is defined as purposeful physical activity initiated for the purpose of health, it is clear that in this experiment the rats did not exercise. We would, therefore, argue that the effect of thyroid hormone excess in these animals was to increase the EE associated with spontaneous physical activity that we term NEAT (7). It would be interesting to repeat this experiment and include an exercise wheel to determine whether rats would volitionally increase wheel running, for example. One could also investigate the interplay between imposed exercise and spontaneous physical activity in hyperthyroidism. We would argue that the increases in activity EE observed in this experiment are akin to NEAT. This is an important concept because thyroid hormone might then be viewed as a putative mediator of NEAT, which appears to be an important counterbalance to fat gain as humans are overfed (7).

At a molecular level, thyroid hormone is known to contribute to brown fat thermogenesis. First, thyroid hormone induces a regulated proton leak via uncoupling proteins, especially uncoupling protein 1 in brown fat (12). Second, thyroid hormone stimulates the expression of elements in the norepinephrine-signaling pathway (11) via the ₁ thyroid hormone receptor isoform (15). Our study is not the first to suggest that thyroid hormone impacts other components of EE (i.e., nonresting EE). For example, thyroid hormone impacts tension-independent EE during contraction of isolated muscle preparations perhaps through altered ATP hydrolysis (6). Silva (11) suggests, from a molecular perspective, that nonresting EE may account for more than half of the thyroid hormone effects on EE. Perhaps this occurs via a series of mechanisms that include greater heat dissipation for a given stretch of skeletal muscle, which in turn might have profound effects on activity EE and NEAT and might independently affect the amount of physical activity performed. A complementary mechanism might be that thyroid hormone increases sympathetic outflow and thereby norepinephrine effects; the sympathetic nervous system has been linked to levels of spontaneous physical activity in humans (9). One could speculate further as to the potential importance of thyroid hormone excess on energetic efficiencies of tissues, such as muscle and adipose, that express uncoupling proteins 2 and 3 (1, 4). Thus there are provocative molecular links between thyroid hormone excess and NEAT particularly with reference to the marked increase in activity EE per unit time that our hyperthyroid animals demonstrated. Our data provide evidence that thyroid hormone excess increases nonresting or activity EE.

What might the neural mechanism of the interaction between thyroxine and physical activity or NEAT be? Very little is known regarding how spontaneous physical activity is centrally modulated, although the cerebellum (3, 5) and the energy balance regulation centers in the hypothalamus (17) are potential candidates. It was interesting that both EE and food intake increased with thyroid hormone excess and body weight remained stable. We would speculate that T₃ acts centrally, possibly via hypothalamic energy balance centers, to simultaneously increase EE, NEAT, and appetite. It is intriguing to speculate as to why rearing appeared to be important in the T₃-related increases in activity EE we observed. Perhaps this is a mechanism for increasing heat dissipation through increases in exposed surface area. Could rearing be akin to being bipedal? In humans, walking even at very slow velocity (1 mile/h) is a potent means of expending energy (8). Could it be that rodents exploit this effect as a means of expending energy? Future studies will tell.

Our study had limitations that we recognize. First, we did not account for the potential effect of hyperthyroidism on the thermic effect of food and/or resting EE. It could be argued that substantial changes in the thermic effect of food could have affected the changes in NEAT we reported, noting that these animals ate more. However, the thermic effect of food generally accounts for 10-15% of total EE. Hence it would take implausible changes in the thermic effect of food to impact on our findings. Moreover, not only did NEAT increase dramatically with hyperthyroidism but also so did independent measurements of spontaneous physical activity. Thus it is unlikely that measuring the thermic effect of food would have impacted on our primary observations and conclusions. It is unlikely that resting EE data would have impacted our findings or conclusions. By definition, all our measurements of spontaneous physical activity and NEAT were made while the animals were moving and hence not resting. We acknowledge that the effects of thyroid hormone excess on the thermic effect of food and resting EE would be important for formal energy-balance studies. A second limitation of this study was that T₃ was administered at levels compatible with human thyrotoxicosis rather than physiological perturbation. We felt that this was warranted because so little information was available regarding the effect of thyroid hormone excess on physical activity and NEAT. If our null hypothesis had been supported in this study, we would be hesitant to proceed with experiments using more physiological perturbations of the thyroid hormone axis over longer experimental periods. Because these data support rejection of the null hypothesis, future studies should assess the impact of physiological perturbations of thyroid hormone excess on NEAT (7). Finally, this experiment was performed in rodents. It would have been unethical to recapitulate this study in humans, but these results are sufficiently exciting and robust to spur an exploration into the role of thyroid hormone in human spontaneous physical activity and NEAT. Overall, despite the limitations noted, we think that these data allow us to readily test our hypotheses and draw firm conclusions within the confines of the experimental paradigm.

In conclusion, thyroid hormone excess in rats is associated with increased spontaneous physical activity and NEAT. Defining the central mechanism of this effect will advance our understanding of thyroid hormone action and energy balance modulation.