Law, James M.
(2021)
The thyroid and glycaemic endocrine influences on brown adipose tissue and its measurement using infrared thermography.
PhD thesis, University of Nottingham.
Abstract
Brown, and brown-like, adipose tissue provides a potential target for treatments to increase energy expenditure and the consumption of lipids and glucose. In humans, active depots have been shown to persist into adulthood where, previously, they were thought to atrophy in childhood. Due to the presence of uncoupling protein-1 (UCP1) within the inner mitochondrial membrane, brown adipocytes possess the capability to prevent the energy released during cellular respiration from being stored as adenosine triphosphate, instead allowing the energy to be released as heat. Since heat energy is, eventually, dissipated from the body, the result is energy expenditure. Why the body possesses such a tissue specifically designed to waste energy is explained by the benefit conferred by the ability to produce heat on demand, in response to cold challenge. The major alternative heat generating method, shivering, can respond instantly but produces vastly less heat for the same effort. Beyond the acute shivering response, which cannot be maintained for prolonged periods, brown adipose tissue is an essential contributor to thermal homeostasis.
Brown adipose tissue is innervated by the sympathetic nervous system (SNS), acting on B3-adrenoreceptors to stimulate intracellular lipolysis, increased free fatty acids (as well as glucose uptake), upregulation of UCP1 and, hence, heat production. This process is modulated both at the level of the hypothalamic SNS activity and at the intracellular level by thyroid hormones. Thyroid hormones act on the hypothalamus to increase brown adipose tissue stimulation and the biologically active thyroid hormone, triiodothyronine, concentrations are increased within the BAT cell by conversion from the inactive form, thyroxine, by 5’deiodinase-2.
To examine the efficacy of potential stimulators, an effective way of measuring brown adipose tissue activity is required. Positron emission tomography-computed tomography (PET-CT) measures glucose uptake and therefore, under cold stimulation, has allowed brown adipose tissue activity to be quantified using the surrogate measure of glucose uptake and for brown adipose tissue volumes to be calculated. However, PET-CT is unable to measure brown adipose tissue in the fed state due to glucose uptake of muscles, and the radiation exposure means it is not suitable for use in children, large studies or repeat measurements. It is of particular interest to understand the attributes of brown adipose tissue in children since this is the period of maximal brown adipose tissue activity and the time in which physiological parameters are established with lifelong effects.
Since the major output of BAT is heat, attempts have been made to measure the heat that is transferred from the BAT depot to the skin either by use of thermocouples, such as iButtons™, or infrared thermography. These methods aim to address some of the limitations of PET-CT as they are non-invasive, readily available and do not expose the participant to ionising radiation. Furthermore, they directly measure the major output of BAT activity, i.e. heat, rather than measuring an input, e.g. glucose uptake, which may not always reflect the activity of the tissue.
The thesis presented here aimed to validate IRT against PET-CT for the measurement of BAT activity and to improve the efficiency of IRT image analysis by automation before using IRT to demonstrate the effects of a) glucose dysregulation and b) thyroid hormone dysfunction on human BAT in vivo.
While infrared thermography was clearly able to demonstrate a supraclavicular hotspot, it was questioned whether this truly reflected BAT and whether changes in the thermal signature were indicative of BAT activation as many other factors influence skin temperature. In addition, image analysis was limited by the time-intensive process of manual image analysis. I, therefore, designed a semi-automated method of image analysis, coded using MATLAB™ by a colleague in the University’s Faculty of Engineering. This method was 86% faster than the previous method, without any increase in variation on repeated analysis.
I then undertook an analysis of thermal images and PET-CT scans from the same individuals and compared them. Following image alignment, the “hotspot” of glucose uptake on PET-CT overlaps with the thermal “hotspot” from infrared thermography and glucose uptake was correlated well with changes in the temperature of the skin overlying the BAT depot.
Having established and validated the image analysis method, a series of cold-water swimmers were imaged before and after their event. Exposure to this mixed maximal stimulation showed dramatic preservation of the BAT hotspot compared to the sternal reference region (3.5±1.6°C higher post-swimming compared to baseline), although further work is needed to identify the relative effects of the different stimulation components.
I have then shown that girls with autoimmune hypothyroidism have reduced increase in supraclavicular temperature relative to a sternal reference point compared to healthy controls (hypothyroid: 0.1±0.1°C; control: 0.2±0.2°C; p = 0.04) yet, within those participants with hypothyroidism, TSH concentration was associated with increased relative supraclavicular temperatures (r = 0.7, p = 0.01). In children with diabetes compared to healthy controls stimulated relative supraclavicular temperatures were lower (diabetes: 35.0±0.6°C; control: 35.4±0.5°C; p = 0.04) and a smaller change in relative supraclavicular temperature following stimulation, after adjusting for BMI (diabetes: 0.1±0.1°C; control: 0.2±0.2°C; p = 0.03). In these studies, data was not available on thyroid function test results or glycaemic status for healthy volunteers which future work should look to address in addition to measuring and controlling for pubertal status.
In conclusion, this thesis demonstrates that infrared thermography is a valid and reproducible way to measure BAT activity in the non-invasive and affordable way. The increase in image analysis rate provides the opportunity for future work to look at larger groups, longer imaging periods or increased image frequency even to the level of thermal videos and, thereby, reveal finer detail about the mechanisms and potential stimulators. Further advances in image analysis techniques will aim to fully automate the process.
In children with diabetes, there is an indication of a similar reduction in BAT activity, potentially due to the differences between exogenous and endogenous insulin, such as endogenous insulin not being suppressed by sympathetic nervous system activity. Further work is also required to better understand the complex interaction between BAT and the thyroid-axis with many questions remaining. In girls with autoimmune thyroid disease, BAT activity appears to be reduced but TSH appears to stimulate BAT activity suggesting that the disease process itself may have a negative effect on BAT response but this may be mitigated by high TSH levels.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
