Purpose: The purpose of this study was to compare female runners with a history of stress fractures to matched runners who have no history of stress fractures to determine if differences existed in adolescent dairy intake, current dietary calcium intake, total energy, bone mineral density, and menstrual irregularities.
Methods: 27 female runners (aged 18-40) who had at least one stress fracture that had been diagnosed by a doctor were matched to a control sample of 32 female runners who were similar in age, weight, and fat free mass (FFM). Subjects visited the lab in the late follicular phase of their menstrual cycle and consented to a fasting blood draw for assessment of serum estradiol, serum intact parathyroid hormone (iPTH), and serum 25-hydroxyvitamin D (25-(OH)D). Bone measurements were performed by dual-energy x-ray absorptionmetry (DXA). Subjects answered a questionnaire on stress fracture history, training, menstrual status, and lifetime dairy intake, and they also completed a 3-day food record.
Results: Subjects did not differ in age, weight, height, FFM, body mass index, or training volume. Menstrual history and current menstrual characteristics were not significantly different between groups, nor were bone measurements at any sites. A greater number of stress fracture subjects were using calcium and/or vitamin D supplements compared to control subjects (p<0.05), and there was a trend for greater dietary calcium intake in control subjects (p=0.077). Previous and current servings of dairy intake did not differ between groups. However, in multiple linear regression, weekly servings of milk during middle school were a significant predictor (p<0.05) of BMD in the total femur, femoral neck, Ward’s triangle, and femoral shaft. Number of menstrual cycles in the previous year was a significant positive predictor of femoral neck BMD (p<0.05). Caffeine intake was significantly negatively correlated with BMD of the total femur, femoral neck, trochanter, and femoral shaft. Some training variables were significantly negatively associated with bone, including running pace and total femur, days run per week and 33% radius, and also hours spent exercising per week were negatively correlated with total body BMD, total body BMC, and lumbar spine BMD. Logistic regression was used to identify risk factors for stress fractures, from which two models were created. The first model included the number of years subjects had been running, regular or irregular menstrual history, total body BMD, current dietary calcium, serum 25-(OH)D, and body mass index. The second model included the percent of time subjects trained on hard ground, current dietary calcium, regular or irregular menstrual history, total body BMD, serum 25-(OH)D, and body mass index .
Conclusions: Although no significant differences in BMD of various skeletal sites emerged between female runners with and without stress fractures, higher dairy intake during adolescent years, especially middle school milk servings, was highly positively correlated with BMD. Additionally, caffeine intake and training variables were negatively associated with BMD. These results show that adolescent dairy intake contributes to accumulation of bone mass, although improved BMD alone may not be protective against the development of stress fractures later in life. Combined risk factors for stress fractures that emerged in logistic regression included longer running history, training predominantly on a hard surface, irregular menstrual history, lower dietary calcium intake, reduced serum 25-(OH)D, and lower total body BMD.