Metabolic and lifestyle profiling of overweight female runners compared to lean counterparts: exploring the implications and causes of their elevated body weight

• Main Author: Leith, David
• Other Authors: Smith, James
• Format: Dissertation
• Language: English
• Published: University of Cape Town 2017
• Subjects: Exercise Science
• Online Access: http://hdl.handle.net/11427/22715

There appears to be an emerging phenotype of recreational runners who are overweight despite being regularly active. This conflicts with the common perception that exercise protects against weight-gain, and it may be caused by underlying insulin-resistance. Alternatively, recent research has brought attention to metabolically healthy obese (MHO) individuals, who have increased adiposity but no commonly associated metabolic abnormalities, such as insulin-resistance, hypertension, dyslipidaemia and systemic inflammation. This study aimed to determine whether overweight (OW, BMI ≥ 25 kg.m⁻²) female runners were at risk of developing metabolic pathology and compare the findings to lean (LN, BMI < 23 kg.m⁻²) counterparts. A secondary aim was to explore potential inherent or lifestyle factors that may have predisposed or contributed to weight-gain in OW runners. METHODS. Twenty (10 OW, 10 LN) female recreational runners (years of running experience 7.1 ± 4.4 OW; 8.0 ± 3.7 LN) matched for mean age (38.7 ± 4.6 OW; 37.7 ± 4.3 LN), current mileage in km.week⁻¹ (42.0 ± 10.9 OW; 44.5 ± 12.1 LN) and running calibre expressed as energy expenditure (kcal.min⁻¹) in their most recent half-marathon (9.0 ± 1.1 OW; 9.2 ± 1.1 LN) were recruited for this study. Body fat percentage (BF%) was determined using DXA. Participants completed questionnaires about health history, lifestyle and eating habits and validated questionnaires concerning recent sleep and stress. Their diet was recorded using 3-day diet records and analysed using the South African Food Data System (Medical Research Council of South Africa). Habitual sleep and physical activity were quantified using 7-day actigraphy (Actiwatch 2) and accelerometry (Actigraph GTX3+) respectively. Blood pressure and resting metabolic rate were measured after an overnight fast. Blood samples were analysed for cardio-metabolic parameters and an Oral Glucose Tolerance Test was performed for insulin-sensitivity. RESULTS. OW exhibited a greater body weight (74.4 ± 6.4 kg OW; 59.4 ± 7.8 kg LN, p < 0.001) but similar fat-free-mass (49.4 ± 5.6 kg OW; 45.4 ± 5.9 kg LN) to the LN group. OW had a higher BF% (32.1 ± 3.9 OW; 21.8 ± 3.9 LN, p < 0.0001), and systolic (118 ± 10 mmHg OW; 107 ± 5 mmHg LN, p < 0.05), but not diastolic (72 ± 6 mmHg OW; 68 ± 4 mmHg LN) blood pressure. There was no difference between groups in serum uric acid, alanine aminotransferase, % HbA1c, total cholesterol, HDL-cholesterol, triglycerides or free-fatty-acids. OW had higher levels of C-reactive protein (1.30 ± 0.97 mg.L⁻¹ OW; 0.59 ± 0.35 mg.L⁻¹ LN, p < 0.05), total cholesterol / HDL-cholesterol (2.70 ± 0.40 OW; 2.30 ± 0.42 LN, p < 0.05) and LDL-cholesterol (2.99 ± 0.65 mM OW; 2.43 ± 0.72 mM LN, p < 0.05), but these were within normal ranges. IDL-cholesterol constituted a significantly greater proportion of total cholesterol in OW compared to LN, but HDL- and LDL- cholesterol sub-fraction distributions were similar. Indices of hepatic (HOMA-IR, 1.06 ± 0.51 OW; 0.86 ± 0.24 LN), and whole-body (Matsuda, 7.84 ± 2.46 OW; 9.16 ± 2.28 LN) insulin-sensitivity were variable and similar between groups. Total area-under-the-curve of the OGTT insulin response tended to be higher in OW (p = 0.08). Two OW runners had insulin-resistance (Matsuda < 5); but no participants had the metabolic syndrome. RMR (kcal.kg FFM⁻¹.day⁻¹) was lower in OW (29.5 ± 2.1 OW; 31.6 ± 2.3 LN, p < 0.05), but there were no significant differences in lifestyle factors (diet, physical activity, sleep and stress). Total energy intake in kcal.day⁻¹ (1928 ± 354 OW; 2166 ± 489 LN) and % macronutrient composition as Protein/Fat/Carbohydrate/Alcohol (20/44/33/3 OW; 16/43/36/5 LN) were both similar between groups. OW and LN also exhibited similar activity in steps.day⁻¹ (10 742 ± 3552 OW; 12 073 ± 3273 LN) and percentage accelerometer wear-time spent in Sedentary/Light/Moderate-Vigorous physical activity (75/14/11 OW; 72/15/13 LN). Both groups attained circa 7 hours.night⁻¹ of sleep, with good sleep onset latency (7.3 ± 5.8 minutes OW; 5.8 ± 3.5 minutes LN) and sleep efficiency (91.6 ± 4.4% OW; 90.7 ± 2.8% LN), and they reported reduced to average levels of recent stress. DISCUSSION. OW runners presented with greater mean adiposity than LN counterparts, but the two groups were not as distinct as anticipated. OW runners did present with greater metabolic risk according to some traditional risk factors, including inflammation, systolic blood pressure, LDL-C and total cholesterol. However, the first three were within normal ranges and the clinical relevance of the latter is questionable. It was, therefore, concluded that on average the OW group was not at metabolic risk. Only two OW runners and no LN runners were insulin-resistant according to indices derived from the OGTT. These findings may primarily reflect the insulin-sensitising effects of regular exercise and the consequent fitness of the OW runners. Appetite-dysregulation is speculated to have played an integral role in their prior weight-gain. We did not identify any lifestyle discrepancies that could have explained this weight-gain. The cross-sectional nature of this study made it difficult to assess past behaviour during weight-gain, and inter-individual variation was considerable. In combination with the small sample size, these factors limited the generalisability of the results. Future exploration of the 'overweight-runner' phenotype is warranted to clarify the mechanisms of weight-gain in habitual runners and consequent lifestyle changes that may promote meaningful weight-loss.