Body Composition, Hemodynamic and Biochemical Parameters in Young Female Normal-Weight Oligo-amenorrheic and Eumenorrheic Athletes and Non-athletes

Abstract

Aims

Low-weight hypogonadal conditions such as anorexia nervosa are associated with marked changes in body composition, hemodynamic and hematological parameters, and liver enzymes. The impact of athletic activity in normal-weight adolescents with/without amenorrhea on these parameters has not been assessed. Our aim was to examine these parameters in normal-weight athletes and non-athletes and determine any associations of body composition, oligo-amenorrhea and exercise intensity.

Methods

We assessed vital signs, complete blood counts, liver enzymes, and regional body composition in 43 oligo-amenorrheic athletes (OAA), 24 eumenorrheic athletes (EA) and 23 non-athletes 14-21 years of age.

Results

BMI was lower in OAA than EA. Systolic and pulse pressure, and temperature were lowest in OAA. Blood counts did not differ among groups. AST was higher in both groups of athletes, while ALT was higher in OAA than EA and non-athletes. Total and regional fat was lower in OAA than other groups, positively associated with heart rate and inversely with liver enzymes.

Conclusions

Athletic activity is associated with higher AST, whereas menstrual dysfunction is associated with lower total and regional fat and higher ALT. Higher liver enzymes are associated with reductions in total and regional fat.

Introduction

The number of young women participating in athletics has increased markedly in recent decades [1]. Exercise has many benefits but carries risks of energy deficit when excessive [2, 3]. The female athlete triad of low energy availability, menstrual dysfunction and low bone mineral density (BMD) is now increasingly recognized [4], with ∼24% of adolescent athletes demonstrating menstrual dysfunction [5]. While excessive exercise has known associations with adverse reproductive and bone outcomes [3, 6], there are few data regarding other adverse effects of excessive exercise and amenorrhea.

Conditions of extreme undernutrition, such as anorexia nervosa (AN), are characterized by changes in body composition, hemodynamic and hematological parameters and liver enzymes [7, 8]. Changes in heart rate, systolic blood pressure and blood flow are also described in adult amenorrheic athletes [9]. However, the impact of exercise and oligo-amenorrhea in normal-weight adolescent athletes is not known. We have shown that adolescent oligo-amenorrheic athletes (OAA) (even when normal-weight) have reductions in body fat that differentiate them from eumenorrheic athletes (EA) and non athletes [10]. Associations of reduced fat mass with hemodynamic, hematological and biochemical parameters in this population remain to be determined.

Regular exercisers have lower heart rate, systolic blood pressure (SBP) and diastolic blood pressure (DBP) than non-exercisers [11, 12] from enhanced parasympathetic tone [13]. Similar changes are observed with severe energy deprivation, as in AN [8]. However, data are lacking regarding hemodynamic parameters in young adolescent normal-weight OAA compared with EA, particularly in relation to body fat. Also, blood counts may decrease with extreme undernutrition (such as AN) [8, 14]. Studies have not examined hematological parameters in normal-weight adolescent OAA vs. EA, although studies in adult athletes report a high prevalence of dilutional pseudoanemia in this population (reviewed in [15]). Finally, although most athletes do not have significant biochemical abnormalities on prescreening evaluations [16], liver enzymes are sometimes elevated. In fact, liver enzymes are elevated at the extremes of the nutritional spectrum, in fatty liver disease from visceral adiposity [17] and related to malnutrition in AN [7, 18]. Data are lacking regarding associations of chronic exercise and amenorrhea with liver enzymes in relation to body fat (surrogate for energy stores).

Our objective was to examine hemodynamic and hematologic parameters and liver enzymes in normal-weight athletes and non-athletes 14-21 years old, and determine associations of body composition, oligo-amenorrhea and intensity of exercise with these parameters. We hypothesized that oligo-amenorrheic athletes (OAA) would have greater impairment of these parameters than EA and non-athletes related to lower body fat (reflection of energy stores).

Methods

Subjects were recruited from medical clinics, newspapers and colleges. We enrolled 43 OAA, 24 EA and 23 non-athletes between 14-21 years old. Training criteria for athletes were ≥20 miles/week of running or ≥4 hours/week of weight-bearing training for ≥6 months in the past year. Inclusion criteria for non-athletes were <2 hours/week of weight bearing activity. Oligo-amenorrhea was defined as absence of menses for at least 3 months within a ≥6 month period of oligomenorrhea (cycle length >6 weeks) or absence of menarche at >16 years. Three subjects 15.1-16.8 years old had not attained menarche, and other causes of menarchal delay were ruled out. Eumenorrhea was defined as ≥9 menses (cycle length 21-35 days) in the previous year. All subjects had a bone age of ≥14 years and BMI between the 10^th-90^th percentiles. Other causes of amenorrhea such as medications affecting the hypothalamic-pituitary-ovarian axis (including oral contraceptives and hormone replacement), hyperprolactinemia, ovarian failure,polycystic ovary syndrome and thyroid dysfunction were ruled out. The study was approved by the Partners HealthCare Institutional Review Board. Informed consent was obtained from subjects ≥18 years and parents of subjects <18 years. Informed assent was obtained from girls <18 years.

Weight was measured on a single electronic scale with subjects wearing a hospital gown, and height as an average of 3 measurements on a single wall-mounted stadiometer. BMI was calculated as weight (kilograms)/height (meters)². Percent ideal body weight (IBW) was calculated using the BMI percentile method. Bone age was assessed from a left hand radiograph [19]. A detailed exercise history was obtained to calculate activity level (hours/week of athletic activity over 12-months). Resting energy expenditure (REE) was measured by indirect calorimetry (VMAX Encore 29 metabolic cart, Viasys Healthcare, San Diego, CA). Blood pressure was measured using an aneroid sphygmomanometer after adjusting cuff size in a sitting position. Pulse pressure was calculated as the difference between SBP and DBP. Pulse was measured manually and temperature with an infrared tympanic thermometer in a standardized fashion. Complete blood count (CBC) was analyzed in the hospital laboratory (Sysmex® X-series analyzer), and potassium and liver function by LabCorp using standard methods (Roche/Hitachi MODULAR® P/D Analyzers). All subjects were interviewed by the study psychologist for disordered eating behavior, and were administered two questionnaires to assess eating behaviors, the Eating Disorders Inventory-2 and the Three-Factor Eating Questionnaire.

Waist and hip measurements were taken standing with a plastic tape measure to the nearest 0.1 cm at the end of expiration. Waist-to-hip ratio was calculated as the waist measurement at the iliac crest divided by the greatest hip circumference. We used dual energy X-ray absorptiometry (Hologic Discovery A, Waltham, MA, USA; software version APEX 3.3) to measure percent body fat and regional fat and lean mass. Trunk/extremity fat was calculated by dividing trunk fat by total extremity fat. Coefficients of variation for lean and fat mass are 1.0% and 2.1%.

Statistical Methods

JMP Software (v10: SAS Institute, Inc., Cary, NC) was used for analysis. Results are reported as means±SD. We used ANOVA to determine differences among groups for most parameters, followed by the Tukey-Kramer test to control for multiple comparisons. For non-normally distributed parameters (AST, ALT, alkaline phosphatase, fat mass, and trunk fat) a non-parametric analysis (Kruskal-Wallis) was used followed by the Steel Dwass test to control for multiple comparisons. Subjects were also divided into quartiles based on hours per week of exercise activity, and variables compared across quartiles of activity. We used Spearman correlations to determine associations between variables, and multivariate analysis to control for confounders. Regression modeling was performed to determine predictors of liver enzymes, with a p-value of 0.1 to enter and leave the model. We included hours/week of athletic activity, percent body fat, lean mass amenorrhea and duration as covariates. A p-value of <0.05 was considered significant.

Results

Clinical Characteristics

Groups did not differ for chronologic age, bone age, height, or Tanner stage (Table 1). OAA had lower BMI than EA, although all subjects had a BMI >17.5 (BMI percentile between the 10^th-90^th). OAA had a later menarchal age than non-athletes. Thirty-five percent of OAA reported disordered eating compared to 4.2% of EA and none of the non-athletes (p=0.0008). Activity level was higher in athletes vs. non-athletes per protocol, and did not differ between OAA and EA. REE was highest in EA, even after adjusting for lean mass.

Table 1. Clinical Characteristics of Amenorrheic Athletes, Eumenorrheic Athletes and Non-Athletes^a,^b.

Characteristics	OAA n=43	EA n=24	NA n=23	P	P OAA vs. EA	P OAA vs. NA	P EA vs. NA
Age, y	19.2 (2.1)	18.0 (2.1)	19.1 (1.7)	0.052	-	-	-
Bone age, y	17.5 (1.0)	17.2 (1.2)	17.5 (1.1)	NS	-	-	-
Age of menarche, y	13.7 (1.8)	12.8 (1.4)	12.2 (1.5)	0.0012	0.06	0.001	-
Duration of amenorrhea, m	9.0 (13.2)
Tanner stage (Breast)	4.6 (0.7)	4.7 (0.6)	4.8 (0.6)	NS	-	-	-
Height, cm	165.5 (6.9)	165.8 (8.2)	162.8 (7.2)	NS	-	-	-
Weight, kg	56.3 (7.1)	62.0 (10.7)	57.7 (8.0)	0.03	0.027	-	-
BMI, kg/m2	20.6 (2.1)	22.4 (2.4)	21.7 (2.5)	0.005	0.005	-	-
% Ideal body weight	96.4 (9.7)	107.8 (13.8)	102.6 (13.3)	0.001	0.0008	-	-
Activity level, h/week	9.9 (4.9)	11.0 (4.3)	1.0 (1.4)	<0.0001	-	<0.0001	<0.0001
Resting energy expenditure, cal	1255 (188)	1444 (219)	1255 (193)	0.0007	0.001	-	0.004
Regional Body Composition
Waist circumference, cm	72.6 (6.0)	75.4 (7.3)	74.7 (7.1)	NS	-	-	-
Hip circumference, cm	91.2 (5.2)	94.5 (7.5)	91.9 (6.1)	NS	-	-	-
Percent body fat	21.3 (5.4)	23.5 (3.8)	27.0 (5.2)	0.0001	-	<0.0001	0.04
Total fat mass, kgc	12.4 (3.8)	15.1 (4.4)	16.0 (4.9)	0.006	0.02	0.004	-
Trunk fat, kgc	4.7 (1.8)	5.6 (2.2)	5.7 (2.1)	0.051	0.07	0.028	-
Extremity fat, kg	6.8 (2.1)	8.6 (2.3)	9.4 (3.0)	0.0002	0.004	0.0001	-
Trunk to extremity fat ratio	0.7 (0.2)	0.6 (0.1)	0.6 (0.1)	NS	-	-	-
Total lean mass, kg	43.5 (5.6)	46.3 (7.7)	40.5 (4.4)	0.005	-	-	0.003
Total extremity lean mass, kg	19.2 (2.8)	21.0 (3.9)	18.2 (2.6)	0.01	0.026	-	0.003

Abbreviations: OAA, oligo-amenorrheic athletes; EA, eumenorrheic athletes; NA, non-athletes

^aValues are expressed as mean (SD)

^bp values ≤0.1 are reported

^cNon-parametric testing

Regional Body Composition

Waist and hip circumference did not differ across groups (Table 1). Both athlete groups had lower percent body fat than non-athletes. Total and extremity fat were lower in OAA than EA and non-athletes. Trunk fat was lower in OAA than non-athletes and trended lower than in EA. However, after adjusting for total fat, OAA had higher trunk fat (p=0.006), lower extremity fat (p=0.002) and higher trunk/extremity fat ratio (p=0.004) than the other groups. EA had higher total and extremity lean mass than OAA and non-athletes.

Hemodynamic, Hematological and Biochemical Parameters

Heart rate was lower in athletes than non-athletes (Table 2), but did not differ in OAA versus EA. SBP and pulse pressure were lower in OAA than EA, but did not differ from non-athletes. DBP did not differ among groups. Heart rate was significantly different when athletic activity quartiles were compared (p=0.0001), and girls in each of the upper three quartiles had lower heart rate than those in the first quartile. No relationship was found between activity level and SBP, DBP or pulse pressure. Temperature was lower in OAA than EA and non-athletes, and not affected by date of screening (seasonal variation).

Table 2. Hemodynamic, Hematological and Biochemical Parameters in Amenorrheic Athletes, Eumenorrheic Athletes and Non-Athletes^a.

	OAA n=43	EA n=24	NA n=23	P	P OAA vs. EA	P OAA vs. NA	P EA vs. NA
Hemodynamic Characteristics
BP Systolic, mm Hg	100.4 (8.7)	107.5 (6.7)	103.0 (7.9)	0.003	0.002	-	-
BP Diastolic, mm Hg	66.6 (7.6)	67.2 (7.3)	66.0 (6.9)	NS	-	-	-
Pulse Pressure, mm Hg	33.7 (6.2)	40.3 (6.4)	36.9 (7.5)	0.001	0.0006	-	-
Heart rate, beats/minute	57.5 (9.9)	60.7 (10.1)	71.3 (10.2)	<0.0001	-	<0.0001	0.001
Temperature, degrees F	97.6 (0.9)	98.4 (0.6)	98.4 (0.9)	0.0002	0.002	0.001	-
Hematological Characteristics
WBC, × 103/ul	5.6 (1.4)	6.1 (1.5)	6.3 (1.6)	NS	-	-	-
RBC, × 106/ul	4.2 (0.3)	4.3 (0.3)	4.4 (0.3)	NS	-	-	-
Hemoglobin, g/dl	13.1 (0.9)	13.2 (1.0)	13.1 (0.7)	NS	-	-	-
Hematocrit, %	38.6 (2.9)	39.2 (3.4)	38.8 (2.0)	NS	-	-	-
MCV, fL	91.1 (5.3)	90.7 (6.2)	89.2 (3.2)	NS	-	-	-
MCH, pg	30.8 (1.9)	30.6 (2.2)	30.1 (1.2)	NS	-	-	-
MCHC, g/dl	33.9 (0.8)	33.8 (1.3)	33.8 (0.9)	NS	-	-	-
RDW, %	13.5 (1.0)	13.2 (0.7)	13.4 (0.7)	NS	-	-	-
Biochemical Characteristics
Potassium, mmol/L	4.1 (0.4)	4.3 (0.4)	4.2 (0.4)	NS	-	-	-
Total Protein, mg/dl	7.1 (0.5)	7.2 (0.4)	7.3 (0.3)	NS	-	-	-
Albumin, mg/dl	4.6 (0.3)	4.5 (0.2)	4.5 (0.2)	NS	-	-	-
Total Bilirubin, mg/dl	0.4 (0.3)	0.4 (0.2)	0.4 (0.2)	NS	-	-	-
Direct Bilirubin, mg/dl	0.1 (0.1)	0.1 (0.1)	0.1 (0.04)	NS	-	-	-
Alkaline Phosphatase, U/Lb	70.1 (23.4)	72.8 (17.3)	63.0 (16.6)	NS	-	-	-
AST, U/Lb	26.5 (9.0)	26.7 (10.3)	18.6 (4.4)	0.0001	-	<0.0001	0.0008
ALT, U/Lb	23.7 (12.6)	16.7 (7.7)	14.0 (4.8)	<0.0001	0.0006	<0.0001	-

Abbreviations: OAA, oligo-amenorrheic athletes; EA, eumenorrheic athletes; NA, non-athletes

^aValues are expressed as mean (SD)

^bNon-parametric testing

Groups did not differ for hematological parameters, albumin and total protein. However, three OAA and 2 EA had hematocrit <34.8 % (mean – 2SD for controls). Aspartate aminotransferase (AST) was elevated in both groups of athletes compared to non-athletes, and alanine aminotransferase (ALT) higher in OAA versus EA and non-athletes. 9.3, 8.7 and 0% of OAA, EA and non-athletes respectively had AST levels above the normal range, and 18.6, 0 and 0 % of OAA, EA and non-athletes respectively had ALT levels above the normal range. After adjusting for hours/week of athletic activity, the difference in AST across groups was not significant; however ALT remained higher in OAA compared to the other groups. Total bilirubin, direct bilirubin, and alkaline phosphatase did not differ among groups.

Determinants of Hemodynamic, Hematological and Biochemical Parameters

SBP and DBP were positively associated with BMI (Table 3). Heart rate was positively associated with percent body fat, total, trunk and extremity fat, and inversely with lean mass and athletic activity. Duration of amenorrhea inversely predicted SBP, pulse pressure, heart rate and temperature.

Table 3. Correlations of Body Composition with Hemodynamic Parameters and Liver Enzymes^a,^b.

	Systolic BP		Diastolic BP		Pulse pressure		Pulse		Temperature		AST		ALT
	r	p	r	p	r	p	r	p	r	p	r	p	r	p
Menarchal age, y	0.02	-	0.0	-	−0.06	-	−0.16	-	−0.16	-	0.28	0.009	0.24	0.02
Duration of amenorrhea, m	−0.25	0.02	0.04	-	-0.39	0.0003	-0.23	0.04	−0.27	0.01	0.24	0.03	0.38	0.0003
Athletic activity, h/week	0.11	-	0.07	-	0.07	-	−0.36	0.0006	−0.18	-	0.38	0.0002	0.26	0.01
BMI, kg/m2	0.30	0.005	0.32	0.003	0.06	-	0.17	-	0.16	-	−0.07	-	−0.15	-
Percent body fat	0.04	-	0.17	-	−0.09	-	0.37	0.0002	0.15	-	−0.44	<0.0001	−0.33	0.001
Total fat, kg	0.10	-	0.19	-	−0.06	-	0.28	0.008	0.11	-	−0.35	0.0008	−0.31	0.003
Trunk fat, kg	0.09	-	0.21	0.046	−0.08	-	0.24	0.02	0.05	-	−0.31	0.003	−0.24	0.02
Extremity fat, kg	0.13	-	0.17	-	−0.01	-	0.33	0.002	0.19	0.08	−0.36	0.0005	−0.36	0.0005
Total lean mass, kg	0.06	-	0.05	-	−0.03	-	-0.32	0.003	-0.12	-	0.28	0.008	0.19	0.07
Total extremity lean mass, kg	0.10	-	0.10	-	−0.03	-	−0.25	0.02	-0.09	-	0.25	0.02	0.14	-

Abbreviations: BMI, body mass index; BP, blood pressure

^ap values ≤0.1 are reported

^bSpearman correlations are reported

AST and ALT were associated inversely with fat stores (percent body fat, total, trunk and extremity fat), and positively with athletic activity level, with AST having a stronger correlation than ALT. We found a positive association between AST and total and extremity lean mass. In addition, menarchal age and amenorrhea duration were positive predictors of liver enzymes. WBC counts were positively associated with BMI (r= 0.23, p=0.03), whereas RBC counts were inversely associated with duration of amenorrhea (r=-0.29, p=0.006).

Reported differences among groups remained significant after controlling for athletic activity, percent body fat, lean mass and amenorrhea duration. Results of regression modeling are shown in Table 4. Significant predictors of AST were subject group (EA and OAA vs. non-athletes) and percent body fat, contributing to 24% of the variability. The only predictor for ALT was subject group (OAA vs. EA and non-athletes), explaining 16% of the variability.

Discussion

With increasing participation of young women in athletics, there is increasing awareness of the female athlete triad, a key component of which is suboptimal energy availability, leading to menstrual dysfunction. All cellular processes including growth, thermoregulation and reproduction require energy, and in energy deficit states, available energy is directed towards crucial life-sustaining functions such as temperature and glucose homeostasis. This is consistent with extreme energy depletion (as in AN) causing alterations in hemodynamic, hematological, and biochemical parameters [8]. We report differences in body composition, hemodynamic, and biochemical parameters in normal-weight adolescent OAA compared with EA and non-athletes.

Regional Body Composition

EA had lower percent body fat and higher lean mass than non-athletes, consistent with known effects of endurance training on body composition [20]. Disordered eating is common in athletes [21], and with excessive exercise increases the risk for low energy availability. Thirty-five percent of our normal-weight OAA reported disordered eating, which may explain the lower BMI and %IBW in OAA than EA (though OAA did not differ from non-athletes). Moreover, OAA had lower total and regional fat than the other groups, indicative of an overall reduction in energy stores (despite normal weight).

Interestingly, after controlling for total fat, OAA had higher trunk fat and trunk/extremity fat than EA and non-athletes, suggestive of a central fat redistribution in normal-weight OAA. This is in contrast to low-weight AN girls who have reductions in trunk fat and trunk/extremity fat relative to total fat [8, 22]. Thus severe energy depletion (as in AN) that causes severe weight loss likely has a different effect on fat distribution than less severe energy depletion where weight is preserved (as in normal-weight OAA). In a previous study, we reported that OAA have higher cortisol than EA and non-athletes [23]. Because increased cortisol favors central adiposity [24, 25], it is possible that trunk fat redistribution in OAA is at least partially mediated by increased cortisol. This is consistent with the increased trunk fat reported following weight gain in women with AN with higher baseline cortisol [26-28]. The absence of truncal adiposity in active AN (despite high cortisol) is likely from direct lipolytic effects of increased growth hormone (GH) (from low IGF-1 and reduced negative feedback) [27, 29]. Normal-weight OAA do not have the marked reductions in IGF-1 seen in AN [8, 30], thus GH is likely not as high as in AN. We did not assess cortisol associations with body composition, and this warrants further investigation.

Hemodynamic Parameters

Long-term endurance exercise decreases heart rate and blood pressure [11, 31] from intrinsic cardiac remodeling and increased parasympathetic tone [32]. Our findings are consistent in that EA and OAA had lower heart rate than non-athletes. Heart rate was inversely associated with activity [33] and positively with body fat, which may indicate an adaptation to conserve energy rather than a physiologic response to exercise.

Exercise has the beneficial effect of lowering blood pressure by reducing vascular resistance and decreasing sympathetic activity [11, 12]. Studies have reported lower SBP in adult OAA compared to EA [9, 34]. In our study, OAA had lower SBP than EA, however, this was not associated with activity. A difference in SBP among normal-weight young athletes with similar athletic activity (OAA and EA) suggests that factors other than exercise may drive the hemodynamic response in OAA. Heart rate and blood pressure are predicted by body fat in AN, a state of extreme energy deficit [8]. However, we found no association between fat mass and SBP in normal-weight athletes and non-athletes. We did find an inverse association between SBP and duration of amenorrhea, consistent with reported associations of hypoestrogenism with lower blood pressure and heart rate in exercisers[9]. Thus, low SBP in OAA may be mediated by menstrual dysfunction and hypoestrogenism rather than by exercise or nutritional deficit.

To our knowledge, this is the first study to assess associations of oligo-amenorrhea and lower fat stores with body temperature in OAA. Klentrou et al. showed that thermoregulation was affected by hormonal status and fat mass in premenarcheal and eumenorrheic young girls [35]. We found lower temperatures in OAA than EA and non-athletes associated with amenorrhea duration, suggesting efforts towards energy conservation in OAA.

Resting Energy Expenditure

EA had higher REE than OAA and non-athletes, even after controlling for lean mass. In non-athletes, an inverse association is expected between REE and parasympathetic activity [36], whereas in athletes, a paradoxical positive association is reported, implying higher REE at lower heart rate and blood pressure [13]. In our study, EA demonstrated athletic adaptive changes (higher REE with lower heart rate), indicating optimum exercise benefits. In contrast, despite normal-weight, similar activity and enhanced parasympathetic activity, OAA had lower REE than EA, likely an additional adaptive measure to conserve energy and prevent weight loss.

Hematological Parameters

Information about hematologic changes in adolescent regular training athletes is limited. One study reported abnormal findings during routine screening in 51% of athletes, such as reduced hematocrit and increased WBC count [37]. In fact, dilutional pseudoanemia is a common finding in adult elite endurance athletes consequent to plasma volume expansion [15]. We did not find differences in hematological parameters among groups, which may reflect the younger age of our subjects or lower intensity of training. This differ from AN girls, who have alterations in many hematological parameters associated with lower BMI [8]. In our study, OAA had WBC counts within the low normal range positively associated with BMI. There is evidence of transitory leukocytosis and decreased hematocrit following intense short-term exercise; however, long-term alterations are not reported [38, 39]. We conclude that chronic exercise and oligo-amenorrhea in young athletes do not affect hematological parameters when weight is preserved.

Biochemical Parameters

We also assessed associations of endurance training, hypogonadism and low body fat with liver enzymes. Aminotransferases are markers of liver injury; ALT elevation is more specific to hepatic cell damage, whereas AST may increase with muscular injury as in myocardial infarction or myolysis [40]. One study reported increased AST (27%) and bilirubin (12%) during routine biochemical screening in elite athletes [16]. Intense short-term exercise can cause AST elevation (sparing ALT) [41], however effects of chronic exercise on liver enzymes remain unclear. In addition, elevated liver enzymes are reported in AN consequent to malnutrition [7, 18], and inverse associations of ALT and GGT with BMI and body fat [42]. We thus expected elevated AST in athletes related to activity, and elevated ALT in OAA from energy depletion (indicated by lower body fat).

Consistent with expectations, AST was elevated in both groups of athletes compared to non-athletes, and ALT (more predictive of hepatocyte injury) was higher in OAA than EA and non-athletes. In fact, 18.6% of OAA had ALT levels above normal. Differences between groups persisted after controlling for confounders. Both transaminases were inversely associated with body fat in normal-weight subjects, consistent with data in AN [42]. In a regression model, body fat and being an athlete (OAA and EA vs. non-athletes) were predictors of AST, while being amenorrheic (OAA vs. EA and non-athletes) was the only predictor of ALT. AST had a stronger association with activity level, and ALT with amenorrhea duration. We thus speculate that ALT elevations are related to the nutritional and hormonal state of OAA (low fat stores and amenorrhea duration), whereas AST elevations are from muscular workload (athletic activity). This indicates that ALT elevations are more concerning in athletes than AST elevations, and suggest poorer nutrition and risk for hypogonadism.

In conclusion, athletic activity is associated with lower body fat and heart rate, and elevated AST. Oligo-amenorrhea is associated with reductions in body fat, SBP and body temperature, and elevations in ALT. Our data suggest that body weight provides limited information about nutritional status in athletes. OAA may have altered metabolic and endocrine function despite being normal-weight, and need careful monitoring of hemodynamic parameters and liver enzymes. A trend over time of decreasing heart rate and increasing liver enzymes may indicate reductions in body fat (and thus energy stores), and should trigger discussions of caloric intake to keep pace with exercise energy expenditure.