Empirical Claim Critique

MAJOR ARTICLE

Tracking weight change, insulin resistance, stress, and aerobic fitness over 4 years of college

Mari K. Hopper, PhDa and Shana Lynn Moninger, BSNb

aDepartment of Cellular and Integrative Physiology, Indiana University School of Medicine, Evansville, Indiana, USA; bOwensboro Health Regional Hospital, Rockport, Indiana, USA

ARTICLE HISTORY Received 9 December 2015 Revised 29 June 2016 Accepted 11 September 2016

ABSTRACT Objective: To determine if weight gain is accompanied by development of insulin resistance (IR) during 4 years in college. Participants: Two cohorts of college students were enrolled in fall semesters 2009 and 2010 and tracked for 4 years. Methods: Following a 12-hour fast, subjects reported for measurement of body mass index (BMI), perceived stress (PSS), aerobic fitness, and blood glucose, insulin, and lipids. Results: In the first year, 33% of subjects were overweight or obese, and 20% were hyperinsulinemic. Year 4 had 29 remaining subjects with disproportionate attrition of overweight and obese individuals. Just over half the subjects gained weight (WI), whereas nearly 30% lost considerable amounts (WD). WD showed significant decline in fasting insulin, low-density lipoprotein (LDL) cholesterol, and PSS from year 1. WI was primarily highly fit men who did not demonstrate increased IR. Conclusion: WI was not associated with IR over 4 years of college.

KEYWORDS Aerobic fitness; BMI; insulin resistance; stress; weight

The college years are commonly associated with weight gain, improper diet, lack of physical activity, and increased stress.1–6 Each of these conditions places one at greater risk of adverse health outcomes, including weight gain and the development of obesity. Weight gain and being overweight have both been linked to the devel- opment of insulin resistance (IR).7,8 Insulin resistance is a condition in which the body fails to respond normally to insulin and thus produces excessive amounts to com- pensate. Insulin resistance, and its associated state of high insulin production (hyperinsulinemia), has been described as the earliest phase in the progressive develop- ment of type II diabetes.9 Additionally, IR has been implicated in the development of other major diseases, including cardiovascular disease10,11 and cancer.12,13

In 2009, we initiated a longitudinal study with the aim of determining if lifestyle during the college years would be associated with weight gain and the development of IR.14 The study was designed to collect data from college freshmen and then continue to track the students over the course of their 4-year college career. Specifically, we hypothesized that increases in weight would be associ- ated with the development of IR.

College freshmen, aged 18 and 19 years, were enrolled in the study. Assessment of body mass index (BMI),

waist circumference, blood variables including insulin, glucose, and lipids, as well as levels of stress and aerobic fitness were made in the first month of the freshman year. Results from the first year were previously pub- lished and indicated that over a third of the freshmen students entered the study either overweight or obese by BMI standards, and approximately 20% were also hyperinsulinemic.14 To our knowledge, we were the first to report the prevalence of hyperinsulinemia in this population.

Plans were to reassess students near the midpoint (year 2—end of second/sophomore year) and upon com- pletion of college (year 4—end of fourth/senior year). This plan was executed and sharing of data from Years 2 and 4 is the primary objective of this report. Addition- ally, based upon the prevalence of hyperinsulinemia, and observed differences in men and women, we expanded our original aims.

Hyperinsulinemia has been associated with both lower and higher rates of weight gain.15,16 Opposing results in the literature were derived from different ethnic populations, age groups, and samples that included type II diabetics. Therefore, we sought to deter- mine if in this group of similarly aged young adults, none of whom reported significant health conditions, the

CONTACT Mari K. Hopper, PhD [email protected] Department of Cellular and Integrative Physiology, Indiana University School of Medicine, Health Professions Room 3061, 8600 University Boulevard, Evansville IN 47712, USA. © 2017 Taylor & Francis

JOURNAL OF AMERICAN COLLEGE HEALTH 2017, VOL. 65, NO. 2, 81–93 http://dx.doi.org/10.1080/07448481.2016.1238385

hyperinsulinemic subjects would experience outcomes differing from their more insulin-sensitive peers.

Furthermore, hyperinsulinemia is known to progress to prediabetes and type II diabetes.17 This progression has been described in a multistage model where the first stage of type II diabetes development is a compensatory period in which insulin resistance is accompanied by an increased rate of insulin secretion (hyperinsulinemia).9

This is the stage we note several of our freshmen students to be in. The second stage is a period when pancreatic b-cells fail to fully compensate for the IR, and fasting glucose values begin to rise. It was of interest to deter- mine if any of the hyperinsulinemic subjects (who were all maintaining normal fasting glucose levels as fresh- men) would progress to the second stage during their 4 years in college.

Another interesting observation made in the first year was that freshmen women reported less healthy eating patterns, were less likely to engage in regular physical exercise, and were more likely to report an increase in stress since high school.14 Others have also reported dif- ferences between college men and women when consid- ering weight gain,3 diet,3 exercise,3,18 and stress.18,19

Therefore, we set out to determine if responses of men and women would differ over the course of 4 years in college. We hypothesized that due to self-reported life- style, women enrolled in the study would experience greater relative gains in weight and IR when compared with their male counterparts.

Throughout the study, all students were informed of their personal results and educated as to the health implications. It was of value, especially with the over- weight, obese, and hyperinsulinemic subjects, to

determine if education alone resulted in altered student behavior and improved health outcomes over the 4-year college experience.

Preparing students to “live wisely” is stated in the uni- versity mission. University officials make conscious deci- sions promoting student health. All students at this university are required to complete 2 physical education courses, and “healthy” food options have been made available through food services. In light of administrative efforts promoting health, it was of interest to track this group of students to determine if health outcomes, including BMI and IR, improved during a student’s 4 years of college life. It is important to determine if what students experience while enrolled in college influ- ences them to lead a more “healthy” lifestyle, or if pat- terns of behavior during the college years place students at greater risk for adverse health outcomes.

Methods

This study utilized a convenience sample of freshmen students enrolled at a public university of approximately 10,000 students located in the Midwest. Subjects were recruited into 2 cohort groups beginning fall of 2009 and fall 2010 (Table 1). During their first week on campus, freshmen students were recruited from an introductory biology course via announcements and fliers. Volunteers were informed that participation necessitated repeat measurements taken at 2 other intervals during their undergraduate education. Of the initial pool of just over 60 volunteers, 23 men and 25 women were enrolled. Vol- unteers were excluded from participation if they were younger than 18 or older than 19 years, were pregnant,

Table 1. Characteristics of subjects in year 1.

Combined(n D 48) Men(n D 23) Women(n D 25) Characteristic M SD M SD M SD

Height (inches) 67.3 0.6 70.5 0.6 64.4z 0.5 Weight (pounds) 156.3 4.8 161.9 6.0 151.2 7.3 BMI 24.2 4.8 22.7 0.6 25.6z 1.2 CDC standards for adults

20–24.99 (Normal) n D 32 n D 18 n D 14 25–29.99 (Overweight) n D 13 n D 5 n D 8 �30 (Obese) n D 3 n D 0 n D 3

Waist circumference (cm)y 79.2 2.2 80.4 2.8 77.3 4.0 Fasting glucose (mg/dL) 85.7 1.3 88.8 1.1 82.8z 2.2 Fasting insulin (mU/mL) 13.9 0.9 13.2 1.4 14.6 1.2 HDL (mg/dL)y 51.4 3.5 45.3 4.2 56.6z 5.2 LDL (mg/dL)y 91.9 7.0 68.9 8.4 111.6z 7.7 Triglycerides (mg/dL)y 64.0 4.1 63.7 4.8 64.6§6.5 6.5 Total cholesterol (mg/dL)y 144.6 8.1 115.5 8.4 169.6z 8.9 Stress—PSS 24.0 1.7 18.8 2.9 28.5z 1.1 VO2max (mL¢kg¡1¢min¡1)y 47.6 1.7 51.8 1.5 42.6 2.1

Note. BMI D body mass index; CDC D Centers for Disease Control and Prevention. yData for cohort 2 only. zp < .05 difference between men and women.

82 M. K. HOPPER AND S. L. MONINGER

were experiencing any type of disease or currently taking any type of prescription medication other than birth control, previously experienced difficulty with venipunc- ture, or were unavailable for testing on weekends during the first month of the fall semester. Participants did not receive compensation of any type. This study was approved by the Institutional Review Board for the Use of Human Subjects in Research, and informed consent was obtained prior to participation.

Baseline measurements were obtained in the first month of the subjects’ freshman year (year 1). Subse- quent assessments were performed near the end of the subjects’ second (year 2) and fourth (year 4) years at the university. Data from the 2 cohorts was combined for each time period (years 1, 2, and 4). For each of the 3 testing periods, subjects reported to the research labora- tory after a 12-hour overnight fast, and weight and height were recorded. Weight was measured to the near- est half pound using a physician beam scale (Healthome- ter, Boca Raton, FL). Scale accuracy was confirmed with external weights, and reliability of the scale was estab- lished. Height was measured to the nearest one-quarter inch using the height rod attached to the scale and con- verted to meters for the calculation of body mass index (BMI) (BMI D weight [kg]/height2 [m2]). All subjects wore shorts and t-shirt with no shoes or belts. As all sub- jects exceeded 22 years of age when project was com- plete, the Centers for Disease Control and Prevention (CDC) definitions for adult BMI were used for classifica- tion: underweight (BMI <18.5 kg/m2); normal weight (BMI 18.5–24.9 kg/m2); overweight (BMI 25–29.9 kg/ m2); and obese (BMI >30 kg/m2).20

Blood was drawn from each study participant using standard venipuncture method. Samples were trans- ferred to serum tubes, centrifuged, serum separated, and stored at ¡80�C. Fasted serum insulin was later deter- mined using radioimmunoassay (Linco, St. Charles, MO), and glucose was analyzed via glucose oxidase enzy- matic method (Yellow Springs Instrument 2300 Stat Plus, Yellow Springs, OH). Normoglycemia was defined as fasting glucose levels between 80 and 100 mg/dL.21

Normal fasting insulin can be difficult to define, as values vary with population and measurement technique. How- ever, in adult populations, fasting insulin values between 5 and 15 mU/mL are generally considered normal val- ues.21,22 Hyperinsulinemia has been defined in a large population based study (Atherosclerosis Risk in Com- munities Study, N D 9,020) as fasting serum insulin �19.1 mU/mL.23 The Mathews HOMA (homeostatic model assessment) model,24 using fasting insulin and glucose values, was used to determine insulin resistance (HOMA D fasting insulin (mU/mL) £ fasting glucose (mmol/L)/22.5). The HOMA model to evaluate insulin

resistance has been directly compared with the euglyce- mic clamp method (considered the gold standard) and was found to be a valid method for quantifying insulin resistance.25,26 As others have reported, fasting insulin levels correlate highly with HOMA values when blood sugar values are normal (r D .99).27,28 Therefore, fasting insulin level is an appropriate measure of insulin resis- tance when blood sugar values are normal.

Self-reported student stress levels were assessed using a 14-item perceived stress scale (PSS). The PSS was previ- ously validated in samples including college students.29

A 5-point response format (0 D never; 1 D almost never; 2 D sometimes; 3 D fairly often; and 4 D very often) was utilized for each question. PSS scores are obtained by reversing scores on the 7 positive items (example: “In the last month, how often have you felt that you were effectively coping with important changes that were occurring in your life?”) and added to scores for the remaining 6 ques- tions (example: “In the last month, how often have you been upset because of something that happened unexpectedly?”).29 High sum scores indicate high levels of perceived stress.

After baseline assessment of cohort 1, additional measurements, including body composition by skinfold, aerobic fitness using a 1-mile walk test, and blood lipids including triglycerides (TG) and cholesterol, were added to all subsequent testing periods for both cohort groups. Therefore, year 1 data reported for skinfolds, walk test, and blood lipids include results obtained only from the second cohort. All subsequent assessments include data collected from both cohort groups.

The Jackson-Pollock 3-site skinfold method for men and women was used to obtain skinfold adiposity (women: triceps, thigh, and suprailliac; men: chest, abdo- men and thigh).30 However, poor interrater reliability in assessing skinfolds prevented meaningful comparison. Therefore, skinfold data are not included in the results.

A 1-mile walk protocol to predict maximal oxygen uptake (VO2max)

31 was used to assess aerobic fitness. Validity for the walk test was previously established by comparing results of a standard treadmill protocol with the submaximal 1-mile walk test in over 300 healthy adult men and women (r D .92).32 For the walk test, an indoor course was measured and subjects were instructed to walk 1 mile as quickly as possible. An investigator monitored the subjects’ progress and noted time of com- pletion to a tenth of a second. Radial pulse was recorded immediately on completion of the timed walk. Pulse (HRpeak), subjects’ weight (W), sex (G; 0 for females, 1 for males), and age (A) were used to calculate predicted VO2max using the following equation: VO2max D 132.853 ¡ (0.0769 £ W) ¡ (0.3877 £ A) C (6.315 £ G) ¡ (3.2649 £ T1) – (0.1565 £ HRpeak).

JOURNAL OF AMERICAN COLLEGE HEALTH 83

According to the manufacturer’s instructions, a Car- dioChek PA and associated lipid panel test strips (Poly- mer Technology Systems, Indianapolis, IN) were used to analyze venous whole blood for triglycerides (TG), total cholesterol, and high-density (HDL) lipoprotein choles- terol. The Friedewald equation was used to estimate low- density lipoprotein (LDL) cholesterol (LDL cholesterol D Total cholesterol ¡ HDL cholesterol ¡ TG). Internal quality control checks were performed before each test- ing period using manufacturer-supplied control materi- als. Comparing 106 patients’ venous blood samples assessed using the Cardiocheck and with standard labo- ratory assay gave correlation coefficients of .86 for total cholesterol (coefficient of variation [CV] � 12%), .74 for HDL cholesterol (CV � 22%), and .98 for TG (CV � 14%).33

Subjects were contacted near the end of their fresh- men and sophomore years to receive study results. A for- mal presentation was made explaining results, implications, and suggested actions to reduce the risk of disease, and students were allowed to ask questions. Sub- jects were weighed, and changes in weight were recorded. Students were asked to verify that measurements were made in the fasted state and were invited to continue involvement in the study.

Statistical analysis

Descriptive statistics were used to provide information about the general characteristics of the sample. Numeri- cal values are reported as mean § standard error (SE). In addition, 2-sample t tests assuming unequal variance were used to examine differences between means in men and women. Repeated-measures t test was used to exam- ine differences between testing periods (year 1 vs year 2 and year 1 vs year 4). All significance levels are reported at minimum of p < .05.

Results

Year 1

In the first year of the study (Table 1), 33% of the 48 sub- jects enrolled in the study were defined as either over- weight (8 women and 5 men) or obese (3 women) by BMI standards. Although the men were significantly taller than the women, there were no significant differen- ces in weight or waist circumference. Women’s BMI was significantly higher and fasting glucose lower than men’s. However, all subjects had fasting glucose levels within normal limits. Fasting insulin levels varied from 2.9 to 26.2 mU/mL, with a mean of 13.91 § 6.26 mU/mL, and there were no significant differences in insulin

between men and women. Approximately 20% of sub- jects (7 women and 4 men) entering the study had fast- ing insulin levels considered hyperinsulinemic (>19.1 mU/mL). Total, HDL, and LDL cholesterol were all significantly higher in the women when compared with the men. Two Hispanic females displayed total cho- lesterol levels above 200 mg/dL. Upon entrance to col- lege, women reported significantly higher stress levels than the men, whereas the men demonstrated signifi- cantly higher predicted VO2max. Even though a higher VO2max is expected in men,

34 when compared with values standardized specific to sex and age, the men as a group could be classified as having “excellent” cardiorespiratory fitness, whereas the women were classified as “good.”35

Year 2

At the end of the second year of college, 34 subjects remained enrolled in the study (17 men and 17 women; 70% retention rate). When compared with year 1, there were no significant differences in BMI, fasting glucose, insulin, triglycerides, HDL cholesterol, and LDL choles- terol. As in year 1, men demonstrated significantly higher levels of VO2max and lower levels of stress when compared with women. Due to minimal change from year 1 of the study, data from year 2 are not reported.

Year 4

At the conclusion of the study (Table 2), 29 of the original 48 subjects remained enrolled (16 men and 13 women). Less than 50% of subjects entering the study as overweight or obese were retained, whereas over 70% of normal-weight subjects participated through completion in year 4. Only 2 of the subjects withdrawing from the study remained enrolled at the university at the end of the fourth year.

Average weight gain over 4 years was just over 6 pounds, with men gaining on average 8.6 pounds (5.2% of body weight) and women gaining 3.1 pounds (2.2% of body weight). Accordingly, BMI for the group increased slightly. In the fourth year, 17 subjects were classified by BMI as normal weight (59%), 10 were overweight (34%), and 2 were obese (7%). Four of the male subjects shifted from the normal BMI category to the overweight cate- gory. However, weight increase in men was likely due to an increase in lean muscle mass associated with normal growth and maturation of this age group, as men demon- strated no change in waist circumference or level of aero- bic fitness while tending to increase in height. For women, both weight and waist circumference demon- strated very little change. However, classifications shifted as 1 individual changed from normal weight to over- weight, 1 moved from overweight to obese, and 1 shifted

84 M. K. HOPPER AND S. L. MONINGER

from overweight to normal weight. Of the 3 subjects clas- sified as obese in year 1, only 1 remained enrolled in the study at year 4, and remained classified as obese.

When comparing group means between year 1 and year 4 (Table 2), there were no significant differences in fasting glucose, insulin, total cholesterol, LDL cholesterol, TG, stress, or VO2max. Only HDL cholesterol showed a signifi- cant change. Throughout the study, all subjects were defined as normoglycemic, with no indication of prediabetes based on fasting glucose (fasting glucose >110 mg/dL).36

Neither the men nor the women demonstrated sig- nificant change in any of the variables measured over the course of the 4 years in college (Table 2). In the group of 29 subjects retained throughout the study, women’s body weight was significantly lower than the men’s in both year 1 and year 4. The women also entered the study reporting significantly higher stress level than the men. However, stress levels tended to decline for both men and women from year 1 to year 4, and differences between the 2 groups were not signifi- cant in year 4. Women also demonstrated higher HDL cholesterol levels than the men throughout the study.

VO2max showed a slight, but not significant, decline from year 1 to year 4 for both the men and women. Since VO2max is expressed per kilogram of body weight, change in VO2max may reflect weight gain, with little to no change in cardiorespiratory fitness.

Group means, and even comparisons between men and women, do not adequately describe results. Although the majority of subjects gained weight, others lost considerable amounts, and some showed

little or no change in weight. Weight change from year 1 to year 4 ranged from negative 16 (weight loss) to positive 37 (weight gain) pounds. Therefore, subjects were divided into 3 groups based on change in weight, and further analysis conducted (Table 3; Figures 1 and 2).

Subjects who increased their weight by more than 5 pounds were classified as weight increase (WI; n D 6 women and 9 men); subjects who decreased their weight by more than 5 pounds were classified as weight decrease (WD; n D 5 women and 3 men); and subjects whose weight change was minimal (WM), neither gaining or losing more than 5 pounds (WM; n D 2 women and 4 men). There were no significant points to consider when discussing the WM group; therefore, the WM data are not reported.

In year 1 of the study, there were no differences in BMI between the WI and WD groups. However, the WD group tended to weigh less than the WI. Differ- ences in weight were due to the fact that the WD group was composed of 5 women, only one of whom was defined as overweight in year 1, and 3 of the small- est men enrolled in the study (average weight of 136 pounds).

Weight increase group On average, the WI group (Table 3; Figure 1a), gained 16.5 pounds, a 10.5% increase in body weight from year 1 to year 4. Weight gain ranged from 8 to 37 pounds, with 2 women in the group gaining over 30 pounds. Although not statistically significant, BMI increased an

Table 2. Repeated-measures comparison between years 1 and 4.

Combined Men Women

Year 1(n D 29) Year 4(n D 29) Year 1(n D 16) Year 4(n D 16) Year 1 (n D 13) Year 4 (n D 13) Characteristic M SD M SD M SD M SD M SD M SD

Height (inches) 68.0 0.8 68.3 0.8 70.9 0.6 71.3 0.7 64.3z 0.8 64.5z 0.8 Weight (pounds) 153.5 4.9 159.6 5.8 163.1 5.8 171.7 7.1 141.6z 7.3 144.7z 7.9 BMI 23.3 0.6 24.1 0.6 22.7 0.6 23.6 0.8 24.0 1.0 24.6 1.0 CDC standards for adults

20–24.99 (Normal) n D 22 n D 17 n D 14 n D 10 n D 8 n D 7 25–29.99 (Overweight) n D 6 n D 10 n D 2 n D 6 n D 4 n D 4 �30 (Obese) n D 1 n D 2 n D 0 n D 0 n D 1 n D 2

Waist circumference (cm)y 79.2 2.2 76.5 2.6 80.4 2.8 76.7 3.4 77.3 4.0 76.3 4.6 Fasting glucose (mg/dL) 87.4 1.6 89.7 1.0 89.3 4.8 90.2 1.5 85.3 3.1 89.1 1.3 Fasting insulin (mU/mL) 13.6 1.2 10.9 0.8 13.4 1.8 10.2 1.1 13.8 1.6 11.8 1.2 HDL (mg/dL)y 49.8 4.0 60.2� 5.7 45.3 3.7 50.9 4.4 57.0 8.2 75.0z 10.5 LDL (mg/dL)y 74.0 8.4 61.4 9.1 65.1 11.8 52.4 7.9 88.2 8.7 67.0 13.9 Triglycerides (mg/dL)y 59.5 4.7 67.2 5.7 60.9 6.5 72.5 7.7 57.2 7.2 58.8 7.6 Total cholesterol (mg/dL)y 124.9 8.8 135.0 9.0 111.6 10.1 132.4 12.5 146.3z 11.5 139.2 13.4 Stress—PSS 23.1 3.2 20.4 2.3 19.1 4.9 17.4 2.8 28.6z 1.9 24.6 3.4 VO2max (mL¢kg¡1¢min¡1)y 50.5 1.9 49.3 2.6 54.4 1.6 53.7 1.4 45.0 2.5 43.2 4.8

Note. BMI D body mass index; CDC D Centers for Disease Control and Prevention. yData for cohort 2 only. �p < .05 difference between year 1 and year 4. zp < .05 difference between men and women.

JOURNAL OF AMERICAN COLLEGE HEALTH 85

average of 2.2 points, with 6 subjects classified as normal weight in year 1 being reclassified as overweight in year 4, and 1 overweight subject reclassified as obese. Despite the significant increase in weight, the only blood parame- ter demonstrating a significant change was TG, which was significantly higher in year 4. Interestingly, this group displayed an increase (not statistically significant) in perceived stress from year 1 to year 4.

Weight decrease group The WD group (Table 3; Figure 1b) lost an average of 9.7 pounds, representing 6.8% of their original body weight. Weight loss ranged from 6 to 16 pounds. The WD group began the study with significantly higher lev- els of LDL cholesterol, total cholesterol, and stress than the WI group. When comparing year 1 with year 4, there were significant declines in fasting insulin, LDL choles- terol, and stress within the WD group.

At the conclusion of the study in year 4 (Table 3; Figure 1), the WI and WD groups displayed nearly iden- tical fasting glucose levels. The WI group weighed signif- icantly more than the WD group and tended to have higher BMIs and VO2max levels. Due to a significant decline in perceived stress in the WD group, stress levels were similar in both groups. It was interesting to note that stress levels tended to increase in the WI group and decrease in the WD group (Figure 2).

Hyperinsulinemia

Due to the relatively high number of hyperinsulinemic subjects entering the study in year 1, it was of interest to track changes in this group. In year 1, 11 subjects were defined as hyperinsulinemic and 5 of these individuals were retained throughout the study. Weight change over the 4 years averaged 5.3 § 3.3 pounds and ranged from a loss of 4.4 pounds to a gain of

30.8 pounds In this small sample, 2 individuals gained weight, 2 lost weight, and 1 maintained their original weight. There were no significant changes in fasting insulin or glucose from year 1 to year 4, and based on

Table 3. Comparison of weight increase and weight decrease groups.

Weight increase(n D 9 men, 6 women) Weight decrease(n D 3 men, 5 women) Year 1 Year 4 Year 1 Year 4

Characteristic M SD M SD M SD M SD

Height (inches) 69.2 1.0 69.7 1.0 64.6 1.5 64.6 1.4 Weight (pounds) 157.0 6.5 173.5� 6.5 141.8 10.3 132.1z 9.7 BMI 22.8 0.5 25.0 0.7 23.9 1.6 22.6 1.5 CDC standards for adults 20–24.99 (Normal) n D 13 n D 7 n D 6 n D 6 25–29.99 (Overweight) n D 2 n D 7 n D 1 n D 1 �30 (Obese) n D 0 n D 1 n D 1 n D 1

Note. BMI D body mass index; CDC D Centers for Disease Control and Prevention. yData for cohort 2 only. �p < .05 difference between year 1 and year 4. zp < .05 difference between WD and WI groups within the same year.

Figure 1. Comparison of mean values for year 1 and year 4 (a) in the weight decrease group (weight loss >5 pounds) and (b) in the weight increase group (weight gain >5 pounds). Compari- sons include fasting blood levels of glucose, triglycerides, total cholesterol, and insulin. �p < .05 difference when comparing year 1 with year 4.

86 M. K. HOPPER AND S. L. MONINGER

fasting glucose level, none of the hyperinsulinemic sub- jects progressed to prediabetes.

Comment

Although it was hypothesized that increases in weight would be associated with the development of IR, there was little change in fasting insulin levels, or any of the other physiological parameters measured. Even when separating subjects who gained more than 5 pounds from those who either lost weight or whose weight did not change, the WI group did not demonstrate an increase in IR (Table 2).

Possible explanations for the unexpected results may include (1) attrition of the most overweight/obese and hyperinsulinemic subjects; (2) WI group consisted primarily of men, and weight gain in men likely differed from weight gain in women; (3) great deal of variability in weight change over the course of 4 years in college; and (4) small sample size limiting statistical power.

Attrition of subjects was anticipated and was nearly identical to institutional rate of attrition for full-time col- lege freshmen over the course of 4 years (42%). What was not anticipated was that a greater percentage of stu- dents in the overweight and obese BMI categories would be lost to attrition when compared with normal-weight individuals. It was interesting to note that both the male and the female with the highest BMIs, and highest fasting insulin levels, did not return to school after their fresh- man year. Primary reason given by both (oral interview) was lack of financial support to continue in school.

Determining the reasons students leave college was not a defined objective of this study. Reasons for leaving col- lege are individual and complex. However, it appears that these data support the observation that low socio- economic status effects both health and ability to sustain pursuit of higher education.37,38 When tracking obese and normal-weight adolescents, others have observed that the prevalence of college graduation in normal- weight individuals to be more than twice that of obese individuals. Furthermore, after adjustment for sociode- mographic variables (age, sex, race, height, parental income-to-poverty ratio, parental education, aptitude test scores), obese adolescents were significantly less likely to have attained a college degree compared with normal-weight peers.37–39 Findings are of concern on this and college campuses everywhere, as being over- weight or obese places a student at greater risk for not obtaining a college education.38

Based on previous reports, we sought to determine if the hyperinsulinemic subjects would experience out- comes that differed from their more insulin-sensitive peers. Of those identified as hyperinsulinemic in year 1, some gained weight whereas others lost weight. Within this small sample, there was no pattern of weight change associated with initial fasting insulin level. Fur- thermore, no comment can be made as to whether ele- vated insulin leads to weight gain, or if weight gain leads to an increases in insulin.40,41 In this study, fasting insulin levels in year 1 did not significantly differ between those who went on to either gain or lose weight (no difference in WI and WD groups in year 1 insulin levels). Furthermore, when using fasting glucose levels between 100 and 125 mg/dL as the criteria, none of the hyperinsulinemic subjects progressed to become prediabetic.42

Another aim of the study was to determine if responses of men and women would differ over the course of 4 years in college. Although the original hypothesis was that women would experience greater rel- ative gains in weight and IR when compared with their male counterparts, results are contrary. In the group of subjects retained over the course of the study, the men gained more weight than the women, both as total pounds, and when expressed as a percentage of their original weight (men increased weight 5.5% above origi- nal weight, whereas women increased 2.2%). Although BMI classification changed from normal weight to over- weight for several of the men, this change in classifica- tion was not likely due to a change in body adiposity. Men displayed little change in waist circumference, retained high levels of aerobic fitness (as indicated by VO2max), and tended to increase in height, all suggesting that weight gain was likely due to gains in lean muscle

Figure 2. Relationship between change in weight and perceived stress levels (PSS) in years 1 and 4. Weight increase group (n D 15) includes subjects gaining >5 pounds. Weight decrease group (n D 8) includes subjects losing >5 pounds. �p < .05 difference when comparing year 1 with year 4 within the same group. yp < .05 difference when comparing weight increase with weight decrease group within the same year.

JOURNAL OF AMERICAN COLLEGE HEALTH 87

mass as part of the normal maturation of males during this developmental period.43 Furthermore, despite the increase in weight, the men did not demonstrate signifi- cant change in any of the health-related variables assessed (Table 2).

Even in the 2 women who gained over 30 pounds, fasting insulin was not significantly altered, but glucose did trend higher. These findings do suggest loss of insu- lin sensitivity with excessive weight gain. We acknowl- edge that findings are limited by small sample size, and future studies enrolling greater numbers of subjects are necessary to sort out the differences in college students who gain weight due to an increase in lean muscle mass from those who gain moderate and extreme amounts of adiposity. Additionally, future studies may be strength- ened by including the administration of an oral glucose tolerance test (glucose challenge) in order to more fully assess the individuals level of IR.42

Another difference between men and women is their perception of stress. When surveyed in the first year of the study, nearly all subjects indicated that they worried over school, job, income, or other problems. Stress stores ranged from 10 to 34 on a perceived stress scale of 0 to 56 and were similar to stress levels reported by others in college-aged students.29 However, twice as many women as men indicated that their stress level had already increased since high school. Throughout the 4 years of college, the men appeared to experience less stress than women (Table 2). Higher stress levels in college women have been reported by others, and those stressors included academic, family, financial, social, and daily hassles.44 These data indicate a need to assist students, especially women, in identifying and developing coping skills for handling stress while in college.

Higher stress levels,45,46 and ethnicity,47 are likely con- tributing to higher cholesterol levels in the women. In a review addressing blood lipids, the authors report that both acute and chronic stress (experienced over a few weeks to years) have been associated with elevations in blood lipid levels.46 The authors conclude that it is likely the severity and consequence of the stressor inter- acting with the individual’s perception of the stressor that impacts lipid metabolism. Although incompletely understood, the mechanism(s) by which psychological stress produces elevations in lipids likely includes neuro- endocrine responses (involving both the sympathetic nervous system and the hypothalamic-pituitary-adrenal axis involving cortisol) and changes in health behaviors, including diet, exercise, and disruption in sleep patterns.

In addition to stress, differences in blood lipid levels may be related to ethnicity. An atherogenic profile including higher LDL cholesterol and TG and lower

HDL cholesterol has been identified in Hispanic popula- tions when compared with Caucasians.47 This study enrolled 2 Hispanic women. Despite neither being obese, both exhibited cholesterol levels above 200 mg/dL in year 1. Both women were directed to see their physician for further consultation. In year 4 of the study, one of the Hispanic women had lost weight and reduced her cholesterol levels below 200 mg/dL, whereas the other gained weight and maintained levels above 200 mg/dL.

In the first-year assessment, all the men described themselves as “in the best shape of their life,” whereas only 50% of the women described themselves this way. Predicted values for VO2max in the fourth year place the men in the “excellent” category for aerobic fitness (VO2max >50), with 15 out of 16 subjects achieving scores for VO2max categorized as good or better.

35 The women as a group scored lower and would be classified as “good,” with only 5 of 13 subjects achieving a VO2max value that rate as good or better.35 Low level of aerobic fitness is of particular concern, as low fitness has been associated with a 2- to 3-fold increased risk of death independent of cardiovascular risk factors.48,49

In this study, the women not only displayed lower lev- els of aerobic fitness, they also had higher BMIs, blood lipids, and stress levels, and a greater number of women were hyperinsulinemic. The observations in women raise the question: are college men “healthier” overall than women, or did this study enroll a sample of young men not entirely indicative of the general population? Similar to our findings, others have reported higher levels of physical activity and physical fitness in college males compared with females. If the observed differences between men and women in this sample are true, the data indicate the importance of getting college women to be active and achieve higher levels of physical fitness, improved measures of insulin sensitivity, and lower stress levels.

In addition to differences between men and women, differences were noted when comparing those who gained weight with those who lost weight. In the group that lost weight, fasting insulin levels in year 4 were sig- nificantly lower when compared with year 1. Others have shown improvements in insulin sensitivity associ- ated with similar amounts of weight loss.50,51

Although others have reported a decline in insulin sensitivity associated with weight gain,52,53 neither fast- ing glucose nor insulin differed between years 1 and 4 in the WI group. Our results may differ due to the specific sample of subjects demonstrating weight gain. As previ- ously discussed, most of the weight gainers in this study were men, who likely increased their lean muscle mass. An increase in lean muscle mass increases the body’s ability to manage glucose,54 and physical activity

88 M. K. HOPPER AND S. L. MONINGER

(as evidenced by high levels of aerobic fitness) has been shown to improve insulin sensitivity.55,56

Although slight differences in insulin were observed between the WI and WD groups, all glucose values assessed were well within the normal range for all sub- jects, at all testing periods. Throughout the course of 4 years in college, fasting glucose levels appeared to be well regulated, even in subjects who gained considerable weight and or were hyperinsulinemic. None of the sub- jects were identified as prediabetic based on fasting glu- cose. At this stage, the hyperinsulinemic subjects were able to regulate fasting blood glucose level. It would be of interest in future studies to determine if hyperinsuline- mic, but otherwise healthy, young men and women would fully regulate blood glucose levels during an oral glucose tolerance test.

Comparison of WI and WD groups also suggests that stress levels may be related to change in body weight. The WI group had lower stress levels than WD in year 1. However, the WI group demonstrated an increase in both weight and stress when comparing year 1 with year 4. In contrast, the WD group lowered both weight and stress level from year 1 to year 4. It has been well documented that stress can alter patterns of food intake, leading to both hyperphagia and hypopha- gia.57 Therefore, one must ask, did higher stress levels in the WD group lead to changes in behavior resulting in weight loss, or did weight loss contribute significantly to the reduction in perceived stress in year 4? In the same line of reasoning, in the WI group, did an increase in stress level lead to behaviors resulting in weight gain, or did an increase in body weight add an additional stressor? At this time, we have no data indicating that high stress levels in the WD group led to lifestyle choices resulting in weight loss, or, alternatively, that weight gain in the WI group contributed to increased perceptions of stress. The relationship between stress and change in body weight, especially in college stu- dents, is of interest and merits further investigation. It may be that managing one’s weight during the college years contributes to management of stress, or, alterna- tively, that managing stress level may be important in maintaining one’s weight.

Although most students gained weight, some students lost weight over the course of 4 years in college. Weight loss, at least superficially, seems desirable. These data must be interpreted with caution, as there was no way of knowing if students practiced habits conducive to good health while losing the weight. One might ask: “Was weight loss attributed to better dietary practice, and inclusion of physical activity, or could the loss be attrib- uted to negative factors, including lack of financial resources and time required to attain groceries and

prepare meals, change in mood including depression, or intake of appetite-depressing drugs and laxatives?” Fur- thermore, only 2 of the 8 individuals in the WD group entered college as overweight or obese. The motivation for weight loss does not appear to be related to one’s knowledge of personal BMI.

Incoming first-year subjects displayed normal glyce- mia, with approximately 20% of subjects being hyperin- sulinemic.14 It is well known that hyperinsulinemia is a reversible condition when individuals participate in appropriate exercise and diet leading to reduced levels of insulin resistance.55,58,59 Due to the prevalence of hyper- insulinemia in this group (along with number of subjects in the overweight and obese categories), it seemed ethical to personally share individual data with all subjects in year 1. The principal investigator met with subjects, explained results, made general suggestions for change, and advised hyperinsulinemic subjects to further consult with their physician.

It was observed that sharing of data with all subjects had little impact on health-related variables. When com- paring year 1 with year 4, only 1 subject achieved a “lower” BMI classification, changing from overweight to normal weight. Over the same span, 7 individuals “increased” in BMI category, changing from normal to overweight, or overweight to obese. These findings indi- cate that testing, informing, and tracking college students was not sufficient in producing change. It is likely that more formal and targeted interventions would be neces- sary to impact change in this group.

Limitations and future directions

Several limitations are present in the current study. Plans were to recruit, schedule, and complete trials for all sub- jects within the first month of college. Therefore, time constraints limited the number of subjects that could be enrolled. Although trends were observed, the sample size was small and therefore lacked sufficient statistical power. All results are considered preliminary and merit further investigation. The sample was also primarily Caucasian. Differences due to ethnicity likely exist; there- fore, future studies should enroll not only a greater num- ber of students, but also those with greater ethnic diversity.

Students were made aware that their weight and other physiological parameters were related to their health. Additionally, students knew they would be monitored and reassessed at the end of their second and fourth years of college. Perhaps early education and monitoring impacted student behavior during the college years and contributed to the avoidance of significant weight gain in some subjects. Although possible, this seems unlikely, as

JOURNAL OF AMERICAN COLLEGE HEALTH 89

mean weight change for the group was similar to change reported by others. This idea merits further investigation.

Although subjects were instructed to maintain normal diet and activity patterns in the week prior to testing, we did not control for these variables (no record was kept). Diet60 and activity levels55 may acutely influence insulin sensitivity and blood lipid levels, and future studies may attempt to control for these variables.

It has been reported that subjects with a more periph- eral distribution of body fat maintain higher levels of insulin sensitivity when compared with subjects with a more central distribution of fat.61,62 Therefore, in addi- tion to BMI, waist circumference was added to this study. Attempts to assess percentage of body fat more accu- rately using skinfold calipers were made, but interrater reliability was an issue and therefore data have not been reported.

It would be of value to conduct future studies that investigate factors that contribute to both weight loss and weight gain during college. Data reported here sug- gest that stress level may be associated with student behaviors that contribute to weight management and further investigation is merited.

It would also be of interest to assess similarly aged adults who choose not to attend college over the same time period. Such data would determine if being enrolled in col- lege is associated with differences in lifestyle that are either more or less beneficial when considering one’s health.

Conclusions

Similar to the findings of others,63,64 the majority of stu- dents in this study gained weight. These findings are to be interpreted with caution, as weight gain included a likely increase in lean body mass in most males. Addi- tionally, weight change during college was highly vari- able, as some students demonstrate little fluctuation in weight whereas others lost considerable amounts.

For those who gain weight and increase adiposity dur- ing the college years, there are no easy answers to avoid- ing this pattern contributing to the development of obesity. Development of obesity is multifactorial, with contributions from genetic, environmental, behavioral, and socioeconomic factors. However, if institutions of higher education are to successfully address their mission to prepare students to live wisely, weight gain and the development of obesity is an issue that must be specifi- cally and intentionally addressed.

For those who lose weight, it is important to address that not all weight loss is considered healthy. Inaccurate body weight perception, and inappropriate weight loss strategies including use of laxatives and diet pills, has been documented among college students.65 It would seem

responsible for college administrators to consider imple- mentation of programs that provide education and access to weight management programs emphasizing appropriate physical activity, healthy eating, and healthy body image.

Ultimately, understanding the myriad of factors (genetic, environmental, behavioral, and socioeconomic) that impact one’s health is a personal responsibility. It is up to each individual to become aware of the multitude of factors affecting health and make responsible deci- sions based upon this information. Data from this study indicate that imparting personal health information, requiring physical education, and offering healthier food options may not improve student health outcomes. It may be time for colleges to look for more creative ways to meet students’ health education and programming needs, such as support for healthy weight loss or gain, stress management, and programs aimed for the specific needs of women, minority students, and those who enter college with compromised health. Additionally, it would be advisable for college students to become aware of exercising their “voice” in effort to impact health policy in their schools, communities, and government and to work diligently to make change possible in their own lives, and the lives of others. On this campus, incoming freshmen students are required to complete a 100 level Introduction to College course. This course may provide the ideal opportunity to assess student health, provide information as to the complexity of health outcomes, encourage students to take personal responsibility, and support students in using their voice to impact change.

In addition to personal responsibility, it is also up the institution to act responsibly. It is likely that the built campus environment plays a role in the health of the stu- dent. Increasingly, “obesogenic environments” are thought to play an important role in the development of obesity, as they provide a setting that promotes a reduc- tion in energy expenditure and an increase in energy con- sumption.66 On this campus, both the dining and recreation facilities are located centrally on the main cam- pus. Perhaps one of many possible actions would be to decentralize food and fitness facilities and provide services in multiple locations, close to where students spend the majority of their time. Having a variety of healthy food choices and exercise facilities close-by may encourage stu- dents to be active and make better dietary choices.

Due to the myriad of health complications that accompany IR and obesity, these conditions have gar- nered much attention and will continue to do so. The college years, and the campus environment, may provide the ideal opportunity to educate and encourage students to take personal responsibility for their choices, and to recognize and promote an environment in which healthy choices are readily available.

90 M. K. HOPPER AND S. L. MONINGER

Acknowledgments

The authors thank Jessica Richards and Gary White for assis- tance in preparing tables and figures. The authors also thank the college students who volunteered as lab assistants and subjects.

Conflict of interest disclosure

The authors have no conflicts of interest to report. The authors confirm that the research presented in this article met the ethi- cal guidelines, including adherence to the legal requirements, of the United States and received approval from the Institu- tional Review Board of the University of Southern Indiana.

Funding

The study was supported by the Pott College of Science and Engineering SERGA and Student Endeavor Grants.

References

1. Racette SB, Deusinger SS, Strube MJ, Highstein GR, Deusinger RH. Weight changes, exercise, and dietary patterns during freshman and sophomore years of college. J Am Coll Health. 2005;53:245–251.

2. Nelson TF, Gortmaker SL, Subramanian S, Cheung L, Wechsler H. Disparities in overweight and obesity among US college students. Am J Health Behav. 2007;31:363–373.

3. Huang TT, Harris KJ, Lee RE, Nazir N, Born W, Kaur H. Assessing overweight, obesity, diet, and physical activity in college students. J Am Coll Health. 2003;52:83–86.

4. Grunbaum JA, Kann L, Kinchen S, et al. Youth risk behav- ior surveillance—United States, 2003. MMWR Surveill Summ. 2004;53:1–96.

5. Anding JD, Suminski RR, Boss L. Dietary intake, body mass index, exercise, and alcohol: are college women fol- lowing the dietary guidelines for Americans? J Am Coll Health. 2001;49:167–171.

6. Roberts AS, Campa MJ, Gottlin EB, et al. Identification of potential prognostic biomarkers in patients with untreated, advanced pancreatic cancer from a phase 3 trial (Cancer and Leukemia Group B 80303). Cancer. 2012;118:571–578.

7. Shalitin S, Abrahami M, Lilos P, Phillip M. Insulin resis- tance and impaired glucose tolerance in obese children and adolescents referred to a tertiary-care center in Israel. Int J Obes. 2005;29:571–578.

8. Kahn SE, Hull RL, Utzschneider KM. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature. 2006;444:840–846.

9. Weir GC, Bonner-Weir S. Five stages of evolving beta-cell dysfunction during progression to diabetes. Diabetes. 2004;53(Suppl 3):S16–S21.

10. Reaven GM. Insulin resistance: the link between obesity and cardiovascular disease. Endocrinol Metab Clin N Am. 2008;37:581–601, vii–viii.

11. Ingelsson E, Sundstrom J, Arnlov J, Zethelius B, Lind L. Insulin resistance and risk of congestive heart failure. JAMA. 2005;294:334–341.

12. Komninou D, Ayonote A, Richie JP Jr, Rigas B. Insulin resistance and its contribution to colon carcinogenesis. Exp Biol Med (Maywood). 2003;228:396–405.

13. Bruning PF, Bonfr�er JM, van Noord PA, Hart AA, de Jong-Bakker M, Nooijen WJ. Insulin resistance and breast-cancer risk. Int J Cancer. 1992;52:511–516.

14. Hopper MK, Brown GW, Funke KA, Pike Brown LR. Prevalence of hyperinsulinemia associated with body mass index, genetic predisposition, and lifestyle in college fresh- men students. J Am Coll Health. 2012;60:27–36.

15. Hoag S, Marshall J, Jones R, Hamman R. High fasting insulin levels associated with lower rates of weight gain in persons with normal glucose tolerance: the San Luis Valley Diabetes Study. Int J Obes Relat Metab Disord. 1995;19:175–180.

16. Folsom A, Vitelli L, Lewis C, Schreiner P, Watson R, Wagenknecht L. Is fasting insulin concentration inversely associated with rate of weight gain? Contrasting findings from the CARDIA and ARIC study cohorts. Int J Obes Relat Metab Disord. 1998;22:48–54.

17. Tab�ak AG, Jokela M, Akbaraly TN, Brunner EJ, Kivim€aki M, Witte DR. Trajectories of glycaemia, insulin sensitivity, and insulin secretion before diagnosis of type 2 diabetes: an analysis from the Whitehall II study. Lancet. 2009;373:2215–2221.

18. Economos CD, Hildebrandt ML, Hyatt RR. College fresh- man stress and weight change: differences by gender. Am J Health Behav. 2008;32:16–25.

19. Cluskey M, Grobe D. College weight gain and behavior transitions: male and female differences. J Am Diet Assoc. 2009;109:325–329.

20. Overweight and obesity: defining overweight and obesity. Centers for Disease Control and Prevention (CDC) Web site. http://www.cdc.gov/healthyweight/assessing/bmi/. Accessed May 25, 2016.

21. Joslin EP, Kahn CR, eds. Joslin’s Diabetes Mellitus. 14th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2005.

22. Li C, Ford ES, McGuire LC, Mokdad AH, Little RR, Reaven GM. Trends in hyperinsulinemia among nondiabetic adults in the U.S. Diabetes Care. 2006;29: 2396–2402.

23. Carnethon MR, Fortmann SP, Palaniappan L, Duncan BB, Schmidt MI, Chambless LE. Risk factors for progression to incident hyperinsulinemia: the Atherosclerosis Risk in Communities Study, 1987–1998. Am J Epidemiol. 2003;158:1058–1067.

24. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabe- tologia. 1985;28:412–419.

25. Gungor N, Saad R, Janosky J, Arslanian S. Validation of surrogate estimates of insulin sensitivity and insulin secretion in children and adolescents. J Pediatr. 2004;144: 47–55.

26. Mather KJ, Hunt AE, Steinberg HO, et al. Repeatability characteristics of simple indices of insulin resistance: implications for research applications. J Clin Endocrinol Metab. 2001;86:5457–5464.

JOURNAL OF AMERICAN COLLEGE HEALTH 91

27. McAuley KA, Williams SM, Mann JI, et al. Diagnosing insulin resistance in the general population. Diabetes Care. 2001;24:460–464.

28. Katz A, Nambi SS, Mather K, et al. Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab. 2000;85:2402–2410.

29. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983;24:385–396.

30. Jackson AS, Pollock ML, Gettman LR. Intertester reliabil- ity of selected skinfold and circumference measurements and percent fat estimates. Res Q Am Alliance Health Phys Educ Rec. 1978;49:546–551.

31. Weiglein L, Herrick J, Kirk S, Kirk EP. The 1-mile walk test is a valid predictor of VO2max and is a reliable alterna- tive fitness test to the 1.5-mile run in US Air Force males. Mil Med. 2011;176:669–673.

32. Kline GM, Porcari JP, Hintermeister R, et al. Estimation of VO2max from a one-mile track walk, gender, age, and body weight. Med Sci Sports Exerc. 1987;19:253–259.

33. Pl€uddemann A, Thompson M, Price CP, Wolstenholme J, Heneghan C. Point-of-care testing for the analysis of lipid panels: primary care diagnostic technology update. Br J Gen Pract. 2012;62:e224–e226.

34. Sparling PB. A meta-analysis of studies comparing maxi- mal oxygen uptake in men and women. Res Q Exerc Sport. 1980;51:542–552.

35. Heyward VH. Advanced Fitness Assessment and Exercise Prescription. 5th ed. Champaign, IL: Human Kinetics; 2006.

36. Group NDD (National Diabetes Data Group). Classifica- tion and diagnosis of diabetes mellitus and other catego- ries of glucose intolerance. Diabetes. 1979;28:1039–1057.

37. Hruby A, Hu FB. The epidemiology of obesity: a big pic- ture. PharmacoEconomics. 2015;33:673–689.

38. Fowler-Brown AG, Ngo LH, Phillips RS, Wee CC. Adoles- cent obesity and future college degree attainment. Obesity. 2010;18:1235–1241.

39. Crandall CS. Do parents discriminate against their heavy- weight daughters? Pers Soc Psychol Bull. 1995;21:724–735.

40. Srinivasan SR, Myers L, Berenson GS. Temporal associa- tion between obesity and hyperinsulinemia in children, adolescents, and young adults: the Bogalusa Heart Study. Metabolism. 1999;48:928–934.

41. Cusin I, Rohner-Jeanrenaud F, Terrettaz J, Jeanrenaud B. Hyperinsulinemia and its impact on obesity and insulin resistance. Int J Obes Relat Metab Disord. 1992;16(Suppl 4):S1–S11.

42. Unwin N, Shaw J, Zimmet P, Alberti K. Impaired glucose tolerance and impaired fasting glycaemia: the current sta- tus on definition and intervention. Diabet Med. 2002;19:708–723.

43. Marshall W, Tanner J. Growth and physiological develop- ment during adolescence. Annu Rev Med. 1968;19:283–300.

44. Brougham RR, Zail CM, Mendoza CM, Miller JR. Stress, sex differences, and coping strategies among college stu- dents. Curr Psychol. 2009;28:85–97.

45. Muldoon MF, Herbert TB, Patterson SM, Kameneva M, Raible R, Manuck SB. Effects of acute psychological stress on serum lipid levels, hemoconcentration, and blood vis- cosity. Arch Intern Med. 1995;155:615–620.

46. Niaura R, Stoney CM, Herbert PN. Lipids in psychological research: the last decade. Biol Psychol. 1992;34:1–43.

47. Aguilar-Salinas CA, Olaiz G, Valles V, et al. High preva- lence of low HDL cholesterol concentrations and mixed hyperlipidemia in a Mexican nationwide survey. J Lipid Res. 2001;42:1298–1307.

48. Mora S, Redberg RF, Cui Y, et al. Ability of exercise testing to predict cardiovascular and all-cause death in asymp- tomatic women: a 20-year follow-up of the lipid research clinics prevalence study. JAMA. 2003;290:1600–1607.

49. Aktas MK, Ozduran V, Pothier CE, Lang R, Lauer MS. Global risk scores and exercise testing for predicting all- cause mortality in a preventive medicine program. JAMA. 2004;292:1462–1468.

50. Rocchini AP, Katch V, Schork A, Kelch RP. Insulin and blood pressure during weight loss in obese adolescents. Hypertension. 1987;10:267–273.

51. Goodpaster BH, Kelley DE, Wing RR, Meier A, Thaete FL. Effects of weight loss on regional fat distribution and insu- lin sensitivity in obesity. Diabetes. 1999;48:839–847.

52. Ondrak KS, McMurray RG, Battaglini CL, Evenson KR, Harrell JS. The relationship between changes in weight sta- tus and insulin resistance in youth. Int J Pediatr Endocri- nol. 2009;2009:862061.

53. Erdmann J, Kallabis B, Oppel U, Sypchenko O, Wagenp- feil S, Schusdziarra V. Development of hyperinsulinemia and insulin resistance during the early stage of weight gain. Am J Physiol Endocrinol Metab. 2008;294:E568– E575.

54. Yki-J€arvinen H, Koivisto VA. Effects of body composition on insulin sensitivity. Diabetes. 1983;32:965–969.

55. Kirwan JP, Solomon TP, Wojta DM, Staten MA, Holloszy JO. Effects of 7 days of exercise training on insulin sensi- tivity and responsiveness in type 2 diabetes mellitus. Am J Physiol Endocrinol Metab. 2009;297:E151–E156.

56. Holloszy JO, Schultz J, Kusnierkiewicz J, Hagberg JM, Ehsani AA. Effects of exercise on glucose tolerance and insulin resistance. Brief review and some preliminary results. Acta Med Scand Suppl. 1986;711:55–65.

57. Rutledge T, Linden W. To eat or not to eat: affective and physiological mechanisms in the stress–eating relation- ship. J Behav Med. 1998;21:221–240.

58. Goodyear LJ, Kahn BB. Exercise, glucose transport, and insulin sensitivity. Annu Rev Med. 1998;49:235–261.

59. Parillo M, Rivellese A, Ciardullo A, et al. A high-monoun- saturated-fat/low-carbohydrate diet improves peripheral insulin sensitivity in non-insulin-dependent diabetic patients. Metabolism. 1992;41:1373–1378.

60. Chen M, Bergman RN, Porte D Jr. Insulin resistance and beta-cell dysfunction in aging: the importance of dietary carbohydrate. J Clin Endocrinol Metab. 1988;67:951–957.

61. Brotons C, de la Figuera M, Franch J, et al. Prediction of glucose and insulin resistance disorders by using anthro- pometric parameters of central adiposity: PRED-IR study. Med Clin (Barc). 2008;131:366–370.

62. Sidney S, Lewis CE, Hill JO, et al. Association of total and central adiposity measures with fasting insulin in a biracial population of young adults with normal glucose tolerance: the CARDIA study. Obes Res. 1999;7:265–272.

63. Holm-Denoma JM, Joiner TE Jr, Vohs KD, Heatherton TF. The “freshman fifteen” (the“ freshman five” actually):

92 M. K. HOPPER AND S. L. MONINGER

predictors and possible explanations. Health Psychol. 2008;27:S3–S9.

64. Hoffman DJ, Policastro P, Quick V, Lee S-K. Changes in body weight and fat mass of men and women in the first year of college: a study of the “freshman 15”. J Am Coll Health. 2006;55:41–46.

65. Wharton CM, Adams T, Hampl JS. Weight loss practices and body weight perceptions among US college students. J Am Coll Health. 2008;56:579–584.

66. Papas MA, Alberg AJ, Ewing R, Helzlsouer KJ, Gary TL, Klassen AC. The built environment and obesity. Epidemiol Rev. 2007;29:129–143.

JOURNAL OF AMERICAN COLLEGE HEALTH 93

Copyright of Journal of American College Health is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.