Abstract
Aims:
Secondary analyses were conducted from a randomized trial of an adaptive behavioral intervention to assess the relationship between protein intake (grams & grams/kg) consumed within 4 hours prior to moderate-to-vigorous physical activity (MVPA) bouts and glycemia during and following MVPA bouts among adolescents with type 1 diabetes (T1D).
Materials & Methods:
Adolescents (n=112) with T1D, 14.5 (13.8, 15.7) years of age and 36.6% overweight/obese, provided measures of glycemia using continuous glucose monitoring (percent of time above range [TAR, >180mg/dL], time-in-range [TIR, 70–180mg/dL], time below range [TBR, <70mg/dL]), self-reported physical activity (Previous Day Physical Activity Recalls), and 24-hour dietary recall data at baseline and 6-months post-intervention. Mixed effects regression models adjusted for design (randomization assignment, study site), demographic, clinical, anthropometric, dietary, physical activity, and timing covariates estimated the association between pre-exercise protein intake on TAR, TIR, & TBR during and following MVPA.
Results:
Pre-exercise protein intakes of 10–19.9g and 20+ grams were associated with an absolute reduction of −4.41% (p=0.04) and −4.83% (p=0.02) TBR during physical activity compared to those who did not consume protein prior to MVPA. Similarly, relative protein intakes of 0.125–0.249g/kg and ≥0.25g/kg were associated with −5.38% (p=0.01) and −4.32% (p=0.03) absolute reductions in TBR during physical activity. We did not observe a significant association between protein intake and measures of glycemia following bouts of MVPA.
Conclusions:
Among adolescents with T1D, a dose of ≥10g or ≥0.125g/kg of protein within 4 hours prior to MVPA may promote reduced time in hypoglycemia during, but not following, physical activity.
INTRODUCTION
For adolescents living with type 1 diabetes (T1D), regular physical activity improves hemoglobin A1c, cardiorespiratory fitness, body mass index, insulin sensitivity, lipid profiles, and perceived well-being.1–3 The risk of hypoglycemia with physical activity, however, is a major barrier to regular participation in physical activity among this population.4,5 Nutritional strategies that may reduce the risk of hypoglycaemia during physical activity would be invaluable in promoting greater levels of physical activity among adolescents with T1D which may then lead to improvements in health and well-being.
Nutrition guidelines for carbohydrate consumption prior to physical activity have been established6; however, less is known about the role of pre-exercise protein intake on glycemia during or following physical activity. Sports nutrition guidelines suggest that adding protein to a carbohydrate containing meal within 4 hours prior to exercise may promote improved recovery from a proceeding exercise bout by promoting increased muscle protein synthesis and reduced muscle protein breakdown during exercise.7–10 Among adolescents and children with T1D, protein ingestion has also been shown to induce a mild, prolonged increase in glycemia lasting up to five hours following a meal, with one study showing higher glucose concentrations as much as 12 hours following fat and protein rich meals among adolescents with T1D.11–14 In theory, this hyperglycemic effect of protein, if consumed prior to exercise as suggested by sports nutrition guidelines, may help to mitigate declining glycemia during and following physical activity among adolescents with T1D.
To the authors’ knowledge, only one study has examined the effect of pre-exercise protein intake on glycemia among adolescents with T1D15. In this study, the authors monitored blood glucose levels during 60 minutes of moderate-intensity cycling exercise following either a protein supplemented breakfast consumed two hours prior to exercise, a standard breakfast with a carbohydrate supplement 15 minutes prior to exercise, or a standardized breakfast with a placebo drink prior to exercise. They found that consumption of a protein-supplemented breakfast two hours prior to exercise was equally effective in reducing the risk of hypoglycaemia during exercise compared to a standardized breakfast which was supplemented with a carbohydrate beverage consumed 15 minutes prior to exercise (n=10).15 To the authors’ knowledge, however, no previous studies have investigated whether pre-exercise protein intake may improve glycemia in the hours following physical activity. This is an important gap in the literature as the risk of hypoglycemia is elevated for up to 24 hours following activity in people with T1D.16 As the effects of protein on glycemia have been shown to persist for at least 5 and possibly as long as 12 hours,11,14 it’s possible that pre-exercise protein may have benefits that persist past the cessation of physical activity.
As such, this study aimed to investigate the relationship between protein intake consumed within 4 hours prior to a bout of moderate-to-vigorous physical activity (MVPA) on the percent of time spent in recommended glucose range (TIR, 70–180mg/dL), percent of time spent above range (TAR, >180mg/dL) and percent time spent below range (TBR, <70mg/dL) during (Aim 1) and until the following morning (Aim 2) among adolescents with T1D. We hypothesized that pre-exercise protein intake will be associated with reduced TBR and improved TIR during and following bouts of MVPA.
MATERIALS & METHODS
Study Design
Data was analysed from the Flexible Lifestyles Empowering Change (FLEX) study17,18 (1UC4DK101132–01), a randomized controlled trial of a behavioral intervention aimed at improving glycemia, psychosocial and metabolic outcomes among adolescents with T1D. The FLEX study was reviewed and approved by institutional review boards at clinical sites in Colorado and Ohio as well as at the coordinating site, the University of North Carolina at Chapel Hill. From 05/01/2014 to 04/04/2016, 258 adolescents between the ages of 13–16 years were enrolled in the FLEX study and randomized to receive either an 18-month adaptive behavioural intervention (n=130) or a usual care control (n=128). Parents provided written informed consent and adolescents provided written assent. The adaptive behavioural intervention incorporated motivational interviewing and problem-solving skills training to promote adherence to T1D self-management skills, insulin dosing, blood glucose testing, diet, and physical activity behaviors. The participants’ diabetes self-management strategies were determined through the motivational interviewing process, with an emphasis on glucose control. The intervention did not systematically incorporate guidance for increasing physical activity. These post-hoc analyses utilize secondary measures from baseline and 6-month visits from the FLEX study to assess the proposed aims. Full details of the design and main results of the FLEX study have been previously published.17,18
Participants
Participants for the FLEX study were enrolled from two sites: Barbara Davis Center for Childhood Diabetes in Colorado and Cincinnati Children’s Hospital and Medical Center in Ohio, coordinated by the University of North Carolina (UNC) at Chapel Hill. Eligible participants were aged 13–16 years at study entry and had HbA1c 8–13% and duration of diabetes >1 year. Youth that were pregnant or had a concurrent severe physical (e.g. Cancer), developmental (e.g. cognitive impairment) or psychiatric (e.g. severe psychopathy) medical condition were excluded from participating in the FLEX study. The distribution of baseline demographic, clinical, glycemic, dietary, and physical activity characteristics among participants included in the final analyses were evaluated and are provided in Table 1. Continuous variables are reported as mean and standard deviation except for non-normally distributed variables, in which median and interquartile range were reported. Categorical variables are described with counts and percentages.
Table 1:
Baseline Characteristics of FLEX Participants Included in Final Analyses (n=112)
| Demographic | Mean ± SD or N (%) |
|---|---|
| Age | 14.5 (13.8, 15.7) |
| Female | 60 (53.6%) |
| Male | 52 (46.4%) |
| Race/Ethnicity | |
| Non-Hispanic White | 90 (80.4%) |
| Non-Hispanic Black | 2 (1.8%) |
| Hispanic | 14 (12.5%) |
| Multiracial/Other | 6 (5.4%) |
| Maximum Education of Parents | n=111 |
| High School or Less | 11 (9.9%) |
| Some College | 30 (27.3%) |
| Four Year College Degree | 49 (44.1%) |
| Graduate Degree | 21 (18.9%) |
| Clinical | |
| Diabetes Duration | 5.5 (3.1, 9.0) |
| Insulin Pump User (n=111) | 81 (73.0%) |
| Previous Day Insulin Dose (units/kg) (n=110) | 1.0 ± 0.3 |
| Anthropometric | |
| Weight (kg) | 58.8 (51.7, 70.0) |
| BMI Z-Score | 0.7 ± 0.9 |
| Estimated Body Fat % | 28.1 (20.0, 33.1) |
| Glycemia | |
| No Personal CGM Use in Past 30 Days (n=102) | 71 (69.6%) |
| Baseline HbA1c (%) | 9.3 (8.5, 9.9) |
| Percent Time in Range (n=106) | 36.5 ± 13.7 |
| Percent Time Below Range (n=106) | 2.1 (0.4, 5.6) |
| Percent Time Above Range (n=106) | 59.5 ± 16.1 |
| Diet | |
| Daily Caloric Intake (kcal) | 1721.0 ± 560.1 |
| Percent of Daily Calories from Protein | 16.0 ± 3.5 |
| Percent of Daily Calories from Carbohydrate | 49.0 ± 7.7 |
| Percent of Daily Calories from Fat | 36.2 ± 6.4 |
| Daily Fiber Intake (grams) | 13.1 (10.1, 18.1) |
| Physical Activity (n = 108) | |
| Meet ADA Guidelines of ≥60mins MVPA/day | 99 (91.7%) |
| Daily Minutes of MVPA | 157.5 (105.0, 225.0) |
| Daily Minutes of Vigorous Physical Activity | 45.0 (0.0, 90.0) |
Continuous variables are reported as mean and standard deviation except for non-normally distributed variables, in which median and interquartile range are reported. Categorical variables are described with counts and percentages.
Measures
Demographics and Health History
Participants completed demographic and health history questionnaires at baseline, 6 months, and 18 months following their baseline visit. From these questionnaires, self-reported age, sex and race/ethnicity were reported as well as insulin regimen, total previous day insulin dose, and T1D duration (years).
Physical Activity
Two previous day physical activity records (PDPAR) were collected in conjunction with the 24-hour dietary recalls over the phone. The previously validated PDPAR19,20 divides the day into half-hour time blocks and queries the dominant activity and the approximate intensity of that activity for that period, categorized as “very light (slow breathing, little or no movement)”, “light (normal breathing, regular movement)”, “medium (increased breathing, moving quickly for short periods of time)” or “hard (hard breathing, moving quickly for 20 minutes or more)” Each activity and perception of effort was matched to a corresponding metabolic equivalent of task (MET) value utilizing the Compendium of Physical Activities21, as detailed by Weston et al.20 From these records, bouts of MVPA were defined as 30 minutes or greater of physical activity at a MET value of greater than or equal to 3 METs.
Continuous Glucose Monitoring
Participants were provided with a blinded Medtronic iPro2 continuous glucose monitor (CGM) with the Enlite sensor for a 7-day period at baseline, 6-, and 18-months following the baseline visit. To enhance compliance and improve quality of CGM data collection, an iPro2 compatible meter (OneTouch Ultra2) was provided to the participant along with test strips (50 strips) for the 7-day CGM study period for calibration for testing 1- and 3 hours after insertion, pre-meal and before bed. Percent of time in recommended glucose range (TIR, 70–180mg/dL), time above range (TAR, >180mg/dL), and time below range (TBR, <70mg/dL) were calculated for the time period during a bout of MVPA (Aim 1) as well as from the end of a bout of MVPA until 6:30 AM the following morning (Aim 2) utilizing consensus report definitions of TIR, TAR, & TBR.22 As the minimal reportable duration of a bout of activity with the PDPAR19,23 is 30 minutes and previous research on the effects of protein intake on glycemia among adolescents with T1D has demonstrated hyperglycemic effects lasting at least 5 hours11–14, these analyses were restricted to observations with at least 30 minutes of CGM data during bouts of MVPA and at least 5 hours of CGM data following those bouts. An example timeline of exposures and outcomes is provided in Figure 1.
Figure 1.
Timeline of Exposures and Outcomes Relative to a Bout of Moderate-to-Vigorous Physical Activity Using Multiple Measurements (Baseline & 6 Months)
Dietary Intake
Two unannounced 24-hour dietary recalls were collected at baseline and 6 months by phone during the 7-day CGM wear time by certified interviewers from the UNC NIH/NIDDK Nutrition Obesity Research Center staff (P30DK056350, MPI Mayer-Davis, Shaikh), using the Nutrient Data System for Research software and the multiple pass interviewing method.24,25 Protein intake consumed within four hours prior to a bout of MVPA was quantified and represents the primary exposure for these post-hoc analyses. For these analyses, observations with pre-exercise protein intake greater than 3 standard deviations above the mean (>71.6g) were excluded as potential outliers. Furthermore, to account for the glycemic effect of carbohydrate and bolus insulin levels, pre-exercise carbohydrate intake (grams) was considered as a potential covariate in the Aim 1 analyses and daily carbohydrate (grams) intake was considered as a potential covariate in the Aim 2 analyses.
Anthropometrics & Body Composition
Height, weight, and natural waist were measured at baseline, 6- and 18-months after their baseline visit utilizing a wall-mounted stadiometer, calibrated electric scale, and a flexible fiberglass or steel tape measure, respectively. Height and weight measurements were also used to determine BMI. These measures were used to estimate percent bodyfat using validated age, race, and sex specific equations.26 Estimated percent body fat was considered as a potential covariate in our statistical models.
Statistical Analysis
All statistical analyses were performed using SAS 9.4 (Cary, NC). Observations with incomplete dietary, physical activity, continuous glucose monitoring or covariate data were excluded in these post-hoc analyses as detailed in Figure 2. Potential sources of selection bias were explored by comparing exposure, covariate, and baseline glycemic data between those with and without adequate data using unadjusted mixed effects models to account for repeated measures. Mixed effects regression models assessed the relationship between protein intake within 4 hours prior to a bout of MVPA and TIR, TBR, TAR during a bout of MVPA (Aim 1) and until 6:30 AM the morning following a bout of MVPA (Aim 2). The estimated effect of protein intake (grams & grams/kg bodyweight) on glycemia was assessed utilizing a categorical variable to account for non-linearity. Categories for grams of protein intake were defined as <10 grams, 10–19.9 grams, and ≥20 grams of protein. Categories for g/kg were defined as <12.5g/kg, 0.125–0.25g/kg, and ≥0.25g/kg bodyweight. In both sets of analyses, non-consumers of protein during the 4 hours preceding the exercise bout were chosen as a reference group. These categories were based on sports nutrition recommendations which suggest a protein intake of 0.25g/kg or an absolute dose of 20–40g as an optimal level to promote positive adaptations to exercise.9,10 To assess whether smaller doses may be effective in promoting improved exercise-related glycemia, we chose to create additional categories above and below half of the dose recommended by sports nutrition guidelines.
Figure 2.
Consort Flow Diagram for Secondary Analyses of the Flexible Lifestyles Empowering Change (FLEX) Randomized Trial
Covariates were introduced into our models in groups: design (randomization assignment, study site), demographics (age, sex, race/ethnicity), clinical (duration of diabetes, insulin regimen, total previous day insulin dose, insulin dose per kilogram), body composition (estimated body fat percentage), physical activity (average bout METs, bout duration (mins), bout volume (MET-minutes), other daily physical activity (MET-minutes)), dietary (pre-exercise carbohydrate intake) and timing variables (hours until midnight). Covariates that produced a ≥10% change in the effect estimate or standard error were included in our final models.
Evaluation of Potential Effect Measure Modification By Exercise Intensity
Previous studies have shown that higher intensity physical activities, such as high intensity interval training or resistance training, may lead to glycemic responses that differ from more moderate intensity activity.27–29 As such, stratified analyses were performed to explore potential effect measure modification by exercise intensity. To do so, we stratified our analyses using a dichotomous variable (1 = vigorous intensity , 0 = moderate intensity). These analytic models adjusted for the same models as our primary models with the exception that bout duration (mins) was substituted for bout volume (met-mins) as analyses were being stratified by intensity which is based on MET values of activity.
RESULTS
Final Sample Size
Of a total of 645 MVPA bouts identified from 135 FLEX participants, 447 bouts from 112 participants were included in our final analytic models as detailed in Figure 2. In sensitivity analyses, there were no significant differences between those with sufficient CGM data and those with missing or insufficient CGM data in pre-exercise or daily nutrient intake, weight, BMI-z score, baseline HbA1c, or any other covariate included in our statistical models.
Baseline Characteristics
Baseline characteristics of participants included in our analyses are provided in Table 1. The median age of participants included in these analyses was 14.5 (IQR: 13.8, 15.7), the median diabetes duration was 5.5 (IQR: 3.1, 9.0) years, and there was a relatively equal inclusion of male and female participants (53.6% female). Furthermore, while 73.0% of participants reported using an insulin pump for their diabetes care, 69.6% reported not having used a personal CGM for their diabetes care in the past 30 days. The participants spent on average 36.5% ± 13.7% TIR, 59.5% ± 16.1% TAR, and 2.1% (IQR: 0.4%, 5.6%) TBR per week at baseline.
Effects of Absolute & Relative Protein Intake Within 4 Hours Prior to a Bout of MVPA on Glycemia During Physical Activity
The median protein intake within 4 hours prior to a bout of MVPA was 14.0 (IQR: 5.0, 26.3) grams or 0.23 (IQR: 0.08, 0.41) grams/kg of bodyweight. The mean TIR, TAR, and TBR during MVPA bouts was 34.50% ± 41.64%, 62.37% ± 43.53%, & 3.13% ± 13.99%, respectively. We observed that protein intakes of 10–19.9g and ≥20g compared to no protein intake were associated with a −4.41% (95% CI: −8.57%, −0.25%) TBR and −4.83% (95% CI: −9.00%, −0.66%) TBR, respectively (Table 2). Similarly, protein intakes of 0.125 – 0.249g/kg and ≥0.25g/kg compared to no protein intake were associated with −5.38% (95% CI: −9.63%, −1.13%) TBR and −4.32% (−8.27%, −0.38%) TBR, respectively (Table 3). No associations were observed between any category of absolute or relative protein intake and TIR or TAR during bouts of MVPA.
Table 2.
Results of Mixed Effects Regression Models Estimating the Association Between Protein Intake (grams) Consumed Within 4 Hours Prior to a Bout of Moderate-to-Vigorous Physical Activity and Glycemia During Physical Activity (n=112, bouts=447)
| Category of Protein Intake | TIR | TBR | TAR | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Estimate | P-Value | 95% CI | Estimate | P-Value | 95% CI | Estimate | P-Value | 95% CI | |
| Unadjusted Models | |||||||||
| < 10 grams protein (bouts=99) | 0.80% | 0.90 | (−11.80%, 13.39%) | −1.06% | 0.62 | (−5.33%, 3.21%) | 0.26% | 0.97 | (−12.82%, 13.34%) |
| 10 – 19.9 grams protein (bouts=122) | −1.45% | 0.81 | (−13.59%, 10.69%) | −5.08% | 0.02 | (−9.19%, −0.97%) | 6.15% | 0.34 | (−6.47%, 18.76%) |
| ≥ 20 grams protein (bouts=159) | 2.36% | 0.69 | (−9.30%, 14.02%) | −5.86% | <0.01 | (−9.80%, −1.91%) | 3.56% | 0.56 | (−8.56%, 15.67%) |
| Fully Adjusted Models* | |||||||||
| < 10 grams protein (bouts=99) | 2.58% | 0.69 | (−10.22%, 15.39%) | −0.46% | 0.33 | (−6.84%, 2.30%) | −2.03% | 0.76 | (−15.30%, 11.23%) |
| 10 – 19.9 grams protein (bouts=122) | 0.22% | 0.97 | (−12.10%, 12.54%) | −4.41% | 0.04 | (−8.57%, −0.25%) | 3.85% | 0.55 | (−8.93%, 16.62%) |
| ≥ 20 grams protein (bouts=159) | 1.19% | 0.85 | (−11.17%, 13.54%) | −4.83% | 0.02 | (−9.00%, −0.66%) | 3.74% | 0.56 | (−9.07%, 16.53%) |
Estimates are adjusted for Estimates were adjusted for intervention group, study site, age, sex, insulin regimen, previous day insulin dose (units/kg), estimated body fat percentage, bout MET/mins, pre-exercise carbohydrate intake, and hours until midnight.
Reference Group = No Protein Intake Within 4 Hours Prior to MVPA Bouts (bouts=67)
Table 3.
Results of Mixed Effects Regression Models Estimating the Association Between Relative Protein Intake (grams/kg) Consumed Within 4 Hours Prior to a Bout of Moderate-to-Vigorous Physical Activity and Glycemia During Physical Activity (n=112, bouts=447)
| Category of Protein Intake | TIR | TBR | TAR | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Estimate | P-Value | 95% CI | Estimate | P-Value | 95% CI | Estimate | P-Value | 95% CI | |
| Unadjusted Models | |||||||||
| <12.5grams/kg (bouts=72) | 4.33% | 0.53 | (−9.15%, 17.81%) | 0.91% | 0.69 | (−3.63%, 5.45%) | −5.28% | 0.46 | (−19.23%, 8.67%) |
| 0.125 – 0.249grams/kg (bouts=102) | −0.90% | 0.89 | (−13.45%, 11.65%) | −5.69% | 0.01 | (−9.92%, −1.47%) | 6.57% | 0.32 | (−6.42%, 19.58%) |
| ≥0.25grams/kg (bouts=206) | 0.12% | 0.98 | (−11.14%, 11.38%) | −5.56% | <0.01 | (−9.35%, −1.77%) | 5.35% | 0.37 | (−6.32%, 17.02%) |
| Fully Adjusted Models* | |||||||||
| <12.5grams/kg (bouts=72) | 6.27% | 0.37 | (−7.41%, 19.95%) | 1.60% | 0.49 | (−2.98%, 6.19%) | −7.80% | 0.28 | (−21.93%, 6.33%) |
| 0.125 – 0.249grams/kg (bouts=102) | 0.70% | 0.91 | (−11.98%, 13.38%) | −5.38% | 0.01 | (−9.63%, −1.13%) | 4.72% | 0.48 | (−8.39%, 17.82%) |
| ≥0.25grams/kg (bouts=206) | −0.71% | 0.91 | (−12.48%, 11.07%) | −4.32% | 0.03 | (−8.27%, −0.38%) | 4.94% | 0.42 | (−7.23%, 17.10%) |
Estimates are adjusted for Estimates were adjusted for intervention group, study site, age, sex, insulin regimen, previous day insulin dose (units/kg), estimated body fat percentage, bout MET/mins, pre-exercise carbohydrate intake, and hours until midnight.
Reference Group = No Protein Intake Within 4 Hours Prior to MVPA Bouts (bouts=67)
Effects of Absolute & Relative Protein Intake Within 4 Hours Prior to a Bout of MVPA on Glycemia From Cessation of Physical Activity Until 6:30 AM the Following Morning
The mean duration of time from the end of MVPA bouts until 6:30 AM the following morning was 15.56 ± 3.89 hours and the mean TIR, TAR, and TBR during this time was 40.80% ± 24.84%, 55.58% ± 28.00%, & 3.6% ± 8.33%, respectively. No association was observed between absolute (Table 4) or relative (Table 5) categories of protein intake and TIR, TAR, or TBR from the cessation of physical activity until 6:30 AM the following morning. Estimated associations ranged from −0.24% - 1.72% (p>0.52), −1.16% - 0.30% (p>0.33), −1.47% - 0.50% (p>0.68) for TIR, TBR and TAR, respectively across absolute and relative intakes of protein pre-exercise.
Table 4.
Results of Mixed Effects Regression Models Estimating the Association Between Protein Intake (grams) Within 4 Hours Prior to a Bout of Moderate-to-Vigorous Physical Activity and Glycemia Following Cessation of Physical Activity until 6:30 AM the Following Morning (n=112, bouts=447)
| Category of Protein Intake | TIR | TBR | TAR | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Estimate | P-Value | 95% CI | Estimate | P-Value | 95% CI | Estimate | P-Value | 95% CI | |
| Unadjusted Models | |||||||||
| <10 grams protein (bouts=99) | 1.20% | 0.74 | (−6.00%, 8.39%) | −0.29% | 0.81 | (−2.68%, 2.10%) | −0.71% | 0.85 | (−8.30%, 6.89%) |
| 10 – 19.9 grams protein (bouts=122) | 1.04% | 0.77 | (−5.90%,7.98%) | −1.13% | 0.33 | (−3.44%, 1.17%) | 0.30% | 0.94 | (−7.03%, 7.63%) |
| ≥ 20 grams protein (bouts=159) | 1.45% | 0.67 | (5.22%, 8.12%) | 0.28% | 0.8 | (−1.93%, 2.49%) | −1.41% | 0.69 | (−8.45%, 5.63%) |
| Fully Adjusted Models* | |||||||||
| < 10 grams protein (bouts=99) | 1.56% | 0.67 | (−5.75%, 8.87%) | −0.50% | 0.68 | (−2.92%, 1.93%) | −0.92% | 0.81 | (−8.64%, 6.80%) |
| 10 – 19.9 grams protein (bouts=122) | 1.72% | 0.63 | (−5.31%, 8.74%) | −1.16% | 0.33 | (−3.49%, 1.17%) | −0.37% | 0.92 | (−7.79%, 7.06%) |
| ≥ 20 grams protein (bouts=159) | 0.54% | 0.88 | (−6.32%, 7.40%) | 0.30% | 0.80 | (−1.98%, 2.57%) | −0.52% | 0.89 | (−7.76%, 6.72%) |
Estimates are adjusted for intervention group, study site, age, sex, insulin regimen, previous day insulin dose (units/kg), estimated body fat percentage, bout MET/mins, daily carbohydrate intake, post-activity protein intake and hours until midnight.
Reference Group = No Protein Intake Within 4 Hours Prior to MVPA Bouts (bouts=67)
Table 5.
Results of Mixed Effects Regression Models Estimating the Association Between Relative Protein Intake (grams/kg) Within 4 Hours Prior to a Bout of Moderate-to-Vigorous Physical Activity and Glycemia Following Cessation of Physical Activity until 6:30 AM the Following Morning (n=112, bouts=447)
| Category of Protein Intake | TIR | TBR | TAR | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Estimate | P-Value | 95% CI | Estimate | P-Value | 95% CI | Estimate | P-Value | 95% CI | |
| Unadjusted Models | |||||||||
| <12.5grams/kg (bouts=72) | 0.91% | 0.81 | (−6.76%, 8.59%) | −0.69% | 0.59 | (−3.25%, 1.86%) | 0.02% | 1.00 | (−8.08%, 8.12%) |
| 0.125 – 0.249grams/kg (bouts=102) | −0.85% | 0.81 | (−8.01%, 6.32%) | −0.22% | 0.86 | (−2.61%, 2.17%) | 1.27% | 0.74 | (−6.30%, 8.83%) |
| ≥0.25grams/kg (bouts=206) | 2.49% | 0.45 | (−3.95%, 8.93%) | −0.24% | 0.83 | (−2.38%, 1.91%) | −1.97% | 0.57 | (−8.78%, 4.83%) |
| Fully Adjusted Models* | |||||||||
| <12.5grams/kg (bouts=72) | 1.11% | 0.78 | (−6.69%, 8.91%) | −0.86% | 0.51 | (−3.46%, 1.73%) | −0.05% | 0.99 | (−8.29%, 8.18%) |
| 0.125 – 0.249grams/kg (bouts=102) | −0.24% | 0.95 | (−7.47%, 7.00%) | −0.07% | 0.96 | (−2.47%, 2.34%) | 0.50% | 0.9 | (−7.14%, 8.14%) |
| ≥0.25grams/kg (bouts=206) | 2.13% | 0.52 | (−4.48%, 8.74%) | −0.42% | 0.71 | (−2.62%, 1.78%) | −1.47% | 0.68 | (−8.45%, 5.51%) |
Estimates are adjusted for intervention group, study site, age, sex, insulin regimen, previous day insulin dose (units/kg), estimated body fat percentage, bout MET/mins, daily carbohydrate intake, post-activity protein intake and hours until midnight.
Reference Group = No Protein Intake Within 4 Hours Prior to MVPA Bouts (bouts=67)
Effect Measure Modification of the Estimated Effect of Pre-exercise Protein Intake on TBR During Bouts of MVPA By Exercise Intensity
Results of analyses stratified by exercise intensity for the estimated effects of relative and absolute pre-exercise protein intake on glycemia during and following exercise are reported in Supplementary Tables 1–4. Stratified analyses found that protein intakes of 10–19.9g and ≥20g were associated with an −6.69% (95% CI: −11.59%, −1.78%) and −8.15% (95% CI: −13.01%, 3.29%) reduction in TBR for moderate intensity bouts of activity, respectively, but not for vigorous intensity bouts, 1.78% (95% CI: −6.60%, 10.16%) and 2.18% (95% CI: −6.43%, 10.79%), respectively. Similarly, protein intakes of 0.125–0.25g/kg and ≥0.25g/kg was associated with an −8.23% (95% CI: −13.15%, −3.29%) and −7.06% (95% CI: −11.68%, −2.46%) reductions in TBR during exercise for moderate intensity bouts, respectively, but not vigorous intensity, 1.97% (95% CI: −6.82%, 10.75%) and 1.76% (95% CI: −6.31%, 9.91%). No associations were observed between pre-exercise protein intake and glycemia following bouts of physical activity for either moderate or vigorous activity.
DISCUSSION
This study utilized existing data from the FLEX trial to explore a unique intersection between diabetes care and sports nutrition by evaluating the role of pre-exercise protein intake on glycemia during and following exercise among adolescents with T1D. It was hypothesized that elevated protein intake within the 4 hours prior to MVPA bouts would be associated with improved TIR and reduced TBR during and following exercise. The results of this study demonstrated that consumption of at least 10g or 0.125g/kg bodyweight was associated with reduced TBR during MVPA, indicating improved safety for adolescents with T1D. No association was observed between pre-exercise protein intake and TIR or TAR during exercise. Similarly, no association was observed between pre-exercise protein intake and glycemia following exercise. Additionally, in stratified analyses, we observed that the benefits of protein intake on glycemia were observed only during moderate-intensity bouts of physical activity which may reflect differing glycemic trajectories following more high-intensity physical activity, however, more research is needed to clarify the role of exercise intensity has on the effect of pre-exercise protein intake on exercise-related glycemia.27–29
These findings are in agreement with the findings of Dube et al. who observed that, among adolescents with T1D, consuming a protein supplemented breakfast two hours prior to exercise was equally effective at preventing hypoglycemia during exercise compared to a standard breakfast that was followed by consumption of a carbohydrate beverage 15 minutes prior to exercise.15 While the size of the reduction in TBR may appear relatively small (4.32% – 5.69% or ~ 2.59 – 3.41 minutes), it is important to note that the mean TBR during physical activity among participants in this study was 3.13% ± 14.0% and guidelines recommend that TBR be minimized among adolescents with T1D.22 As such these findings represent a clinically significant decrease in TBR during physical activity. While previous studies have shown that the hyperglycemic effect of protein intake among adolescents with T1D may persist for 5 hours or longer,11,13,14 we did not observe an association between pre-exercise protein intake and glycemia following that MVPA. It is also important to note that, in healthy populations, consuming protein prior to exercise has been suggested to have potential benefits for promoting recovery or reducing fatigue during exercise among healthy populations.8,30–32 While such effects haven’t been tested among people with T1D, it’s possible that consuming protein prior to exercise may be a promising strategy to assist people with T1D in improving both the safety and benefits of exercise.
Challenges and Opportunities
As with all studies, this study has several limitations. First, self-reported measures of dietary intake are prone to under-reporting due to recall and social desirability biases33, however, the use of a multiple pass method for 24-hours dietary intake data, as was used in the FLEX study, has been shown to minimize the effects of these biases in dietary intake data.25,34 Furthermore, MVPA is often over-reported among adolescents compared to accelerometry35 which may have influenced the number of bouts that we identified. The PDPAR instrument that we utilized in the FLEX study, however, has been validated among adolescents against accelerometers for both relative energy expenditure (r=0.77, p<0.01) and identification of MVPA bouts on a previous day (0.63, p<0.01).20,36 Furthermore, the use of interviewers to administer recalls of physical activity has been shown provide a more reliable measurement of MVPA compared to self-administered methods.37 Additionally, the lack of time-stamped insulin dosing data for these analyses limits are ability to understand the role of insulin-dosing behaviours on the observed associations. By controlling for carbohydrate intake in these analyses we hoped to partially account for bolus insulin levels, which are determined by carbohydrate intake, however, we cannot account for basal insulin dosing and potential insulin dosing strategies which may have been implemented to reduce the risk of exercise-related hypoglycemia.
Finally, approximately 26% of identified bouts of MVPA were missing adequate CGM data which may be a source of selection bias in our analyses. In exploration of differences between those with and without adequate CGM data we did not observe any significant differences between the groups by any variable included in our analyses which may indicate the amount of selection bias present in these analyses is minimal. The availability of time-stamped CGM, dietary intake and physical activity measures, however, provided a unique opportunity to observe a temporal relationship between pre-exercise protein intake and glycemia during and following physical activity, which begin to address an important gap in the literature and start bridging sports nutrition and diabetes care guidelines.
Future Directions
Randomized controlled trials are needed to establish whether a causal relationship exists between pre-exercise protein intake and glycemia during exercise and the hours thereafter among adolescents and adults with T1D. As fear of hypoglycemia is major barrier to regular participation in physical activity among people with T1D, these trials are essential to continuing to address these important gaps in our understanding of the role of peri-exercise protein intake on exercise-related glycemia and to inform dietary guidelines to support safe participation in exercise for those living with T1D. Additionally, while safe participation in exercise is the primary concern for people living with T1D, there are numerous reasons for which a person may decide to participate in exercise that we should aim to support when forming nutritional guidelines. As such, future work should continue to strive to bridge sports nutrition and diabetes care guidelines to help identify nutritional strategies which may promote both enhanced glycemic management and positive adaptive benefits with exercise among people living with T1D.
Supplementary Material
Acknowledgements:
We are indebted to all the adolescents and their families whose participation in the FLEX study made this work possible. We would also like to thank all the students, staff, and faculty who contributed their time and expertise to make the FLEX study possible.
Funding sources:
The FLEX study was funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) of the National Institutes of Health (NIH) under award number UC4DK101132.
Footnotes
Conflict of interest: All authors report no conflicts of interest.
Protocol: The FLEX study is registered on clinicaltrials.gov, NCT01286350, and a detailed description of the design17, main results18,38, and secondary analyses39–42of the FLEX study have been previously published.
REFERENCES
- 1.Yardley JE, Hay J, Abou-Setta AM, Marks SD, McGavock J. A systematic review and meta-analysis of exercise interventions in adults with type 1 diabetes. Diabetes Res Clin Pract. 2014;106(3):393–400. [DOI] [PubMed] [Google Scholar]
- 2.Aljawarneh YM, Wardell DW, Wood GL, Rozmus CL. A Systematic Review of Physical Activity and Exercise on Physiological and Biochemical Outcomes in Children and Adolescents With Type 1 Diabetes. J Nurs Scholarsh. 2019;51(3):337–345. [DOI] [PubMed] [Google Scholar]
- 3.Quirk H, Blake H, Tennyson R, Randell TL, Glazebrook C. Physical activity interventions in children and young people with Type 1 diabetes mellitus: a systematic review with meta-analysis. Diabetic Medicine. 2014;31(10):1163–1173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kahkoska AR, Watts ME, Driscoll KA, et al. Understanding antagonism and synergism: A qualitative assessment of weight management in youth with Type 1 diabetes mellitus. Obesity Medicine. 2018;9:21–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Livny R, Said W, Shilo S, et al. Identifying sources of support and barriers to physical activity in pediatric type 1 diabetes. Pediatric Diabetes. 2020;21(1):128–134. [DOI] [PubMed] [Google Scholar]
- 6.Riddell MC, Gallen IW, Smart CE, et al. Exercise management in type 1 diabetes: a consensus statement. The Lancet Diabetes & Endocrinology. 2017;5(5):377–390. [DOI] [PubMed] [Google Scholar]
- 7.Thomas DT, Erdman KA, Burke LM. Position of the Academy of Nutrition and Dietetics, Dietitians of Canada, and the American College of Sports Medicine: Nutrition and Athletic Performance. Journal of the Academy of Nutrition and Dietetics. 2016;116(3):501–528. [DOI] [PubMed] [Google Scholar]
- 8.Schoenfeld BJ, Aragon A, Wilborn C, Urbina SL, Hayward SE, Krieger J. Pre- versus post-exercise protein intake has similar effects on muscular adaptations. PeerJ. 2017;5:e2825–e2825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jäger R, Kerksick CM, Campbell BI, et al. International Society of Sports Nutrition Position Stand: protein and exercise. Journal of the International Society of Sports Nutrition. 2017;14(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kerksick CM, Arent S, Schoenfeld BJ, et al. International society of sports nutrition position stand: nutrient timing. Journal of the International Society of Sports Nutrition. 2017;14(1):33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Smart CEM, Evans M, O’Connell SM, et al. Both Dietary Protein and Fat Increase Postprandial Glucose Excursions in Children With Type 1 Diabetes, and the Effect Is Additive. Diabetes Care. 2013;36(12):3897–3902. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Paterson MA, Smart CE, Lopez PE, et al. Influence of dietary protein on postprandial blood glucose levels in individuals with Type 1 diabetes mellitus using intensive insulin therapy. Diabet Med. 2016;33(5):592–598. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Paterson MA, Smart CEM, Lopez PE, et al. Increasing the protein quantity in a meal results in dose-dependent effects on postprandial glucose levels in individuals with Type 1 diabetes mellitus. Diabetic Medicine. 2017;34(6):851–854. [DOI] [PubMed] [Google Scholar]
- 14.Neu A, Behret F, Braun R, et al. Higher glucose concentrations following protein- and fat-rich meals - the Tuebingen Grill Study: a pilot study in adolescents with type 1 diabetes. Pediatr Diabetes. 2015;16(8):587–591. [DOI] [PubMed] [Google Scholar]
- 15.Dube M-C, Lavoie C, Galibois I, Weisnagel J. Nutritional Strategies to Prevent Hypoglycemia at Exercise in Diabetic Adolescents. Medicine and science in sports and exercise. 2012;44(8):1427–1432. [DOI] [PubMed] [Google Scholar]
- 16.Gomez AM, Gomez C, Aschner P, et al. Effects of performing morning versus afternoon exercise on glycemic control and hypoglycemia frequency in type 1 diabetes patients on sensor-augmented insulin pump therapy. J Diabetes Sci Technol. 2015;9(3):619–624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kichler JC, Seid M, Crandell J, et al. The Flexible Lifestyle Empowering Change (FLEX) intervention for self-management in adolescents with type 1 diabetes: Trial design and baseline characteristics. Contemporary Clinical Trials. 2018;66:64–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mayer-Davis EJ, Maahs DM, Seid M, et al. Efficacy of the Flexible Lifestyles Empowering Change intervention on metabolic and psychosocial outcomes in adolescents with type 1 diabetes (FLEX): a randomised controlled trial. Lancet Child Adolesc Health. 2018;2(9):635–646. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Keadle S, Lyden K, Hickey A, et al. Validation of a previous day recall for measuring the location and purpose of active and sedentary behaviors compared to direct observation. International Journal of Behavioral Nutrition and Physical Activity. 2014;11(1):12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Weston AT, Petosa R, Pate RR. Validation of an instrument for measurement of physical activity in youth. Medicine and science in sports and exercise. 1997;29:138–143. [DOI] [PubMed] [Google Scholar]
- 21.Ainsworth BE, Haskell WL, Herrmann SD, et al. 2011 Compendium of Physical Activities: a second update of codes and MET values. Med Sci Sports Exerc. 2011;43(8):1575–1581. [DOI] [PubMed] [Google Scholar]
- 22.Battelino T, Danne T, Bergenstal RM, et al. Clinical Targets for Continuous Glucose Monitoring Data Interpretation: Recommendations From the International Consensus on Time in Range. Diabetes Care. 2019;42(8):1593–1603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Clark BK, Pavey TG, Lim RF, Gomersall SR, Brown WJ. Past-day recall of sedentary time: Validity of a self-reported measure of sedentary time in a university population. J Sci Med Sport. 2016;19(3):237–241. [DOI] [PubMed] [Google Scholar]
- 24.Beaton GH, Milner J, Corey P, et al. Sources of variance in 24-hour dietary recall data: implications for nutrition study design and interpretation. The American journal of clinical nutrition. 1979;32(12):2546–2559. [DOI] [PubMed] [Google Scholar]
- 25.Conway JM, Ingwersen LA, Moshfegh AJ. Accuracy of dietary recall using the USDA five-step multiple-pass method in men: An observational validation study. Journal of the American Dietetic Association. 2004;104(4):595–603. [DOI] [PubMed] [Google Scholar]
- 26.Stevens J, Ou FS, Cai J, Heymsfield SB, Truesdale KP. Prediction of percent body fat measurements in Americans 8 years and older. International Journal of Obesity. 2016;40(4):587–594. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.García-García F, Kumareswaran K, Hovorka R, Hernando ME. Quantifying the Acute Changes in Glucose with Exercise in Type 1 Diabetes: A Systematic Review and Meta-Analysis. Sports Medicine. 2015;45(4):587–599. [DOI] [PubMed] [Google Scholar]
- 28.Yardley JE, Kenny GP, Perkins BA, et al. Resistance versus aerobic exercise: acute effects on glycemia in type 1 diabetes. Diabetes Care. 2013;36(3):537–542. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Guelfi KJ, Ratnam N, Smythe GA, Jones TW, Fournier PA. Effect of intermittent high-intensity compared with continuous moderate exercise on glucose production and utilization in individuals with type 1 diabetes. Am J Physiol Endocrinol Metab. 2007;292(3):E865–870. [DOI] [PubMed] [Google Scholar]
- 30.Baty JJ, Hwang H, Ding Z, et al. The effect of a carbohydrate and protein supplement on resistance exercise performance, hormonal response, and muscle damage. J Strength Cond Res. 2007;21(2):321–329. [DOI] [PubMed] [Google Scholar]
- 31.Saunders MJ, Kane MD, Todd MK. Effects of a carbohydrate-protein beverage on cycling endurance and muscle damage. Medicine & Science in Sports & Exercise. 2004;36(7):1233–1238. [DOI] [PubMed] [Google Scholar]
- 32.Saunders MJ, Luden ND, Herrick JE. Consumption of an oral carbohydrate-protein gel improves cycling endurance and prevents postexercise muscle damage. Journal of Strength and Conditioning Research. 2007;21(3):678. [DOI] [PubMed] [Google Scholar]
- 33.Vance VA, Woodruff SJ, McCargar LJ, Husted J, Hanning RM. Self-reported dietary energy intake of normal weight, overweight and obese adolescents. Public Health Nutrition. 2009;12(2):222–227. [DOI] [PubMed] [Google Scholar]
- 34.Moshfegh AJ, Rhodes DG, Baer DJ, et al. The US Department of Agriculture Automated Multiple-Pass Method reduces bias in the collection of energy intakes. The American Journal of Clinical Nutrition. 2008;88(2):324–332. [DOI] [PubMed] [Google Scholar]
- 35.LeBlanc AGW, Janssen I. Difference Between Self-Reported and Accelerometer Measured Moderate-to-Vigorous Physical Activity in Youth. Pediatric Exercise Science. 2010;22(4):523–534. [DOI] [PubMed] [Google Scholar]
- 36.Anderson CB, Hagströmer M, Yngve A. Validation of the PDPAR as an adolescent diary: effect of accelerometer cut points. Med Sci Sports Exerc. 2005;37(7):1224–1230. [DOI] [PubMed] [Google Scholar]
- 37.Sallis JF. Self-Report Measures of Children’s Physical Activity. Journal of School Health. 1991;61(5):215–219. [DOI] [PubMed] [Google Scholar]
- 38.Mayer-Davis EJ, Seid M, Crandell J, et al. Flexible Lifestyles for Youth (FL3X) behavioural intervention for at-risk adolescents with Type 1 diabetes: a randomized pilot and feasibility trial. Diabetic Medicine. 2015;32(6):829–833. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Igudesman D, Crandell J, Zhong VW, et al. Dietary intake on days with and without hypoglycemia in youth with type 1 diabetes: The Flexible Lifestyle Empowering Change trial. Pediatric Diabetes. 2020;21(8):1475–1484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Kahkoska AR, Crandell J, Driscoll KA, et al. Dysglycemia among youth with type 1 diabetes and suboptimal glycemic control in the Flexible Lifestyle Empowering Change trial. Pediatr Diabetes. 2019;20(2):180–188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Sarteau AC, Kahkoska AR, Crandell J, et al. More hypoglycemia not associated with increasing estimated adiposity in youth with type 1 diabetes. Pediatric Research. 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Zhong VW, Crandell JL, Shay CM, et al. Dietary intake and risk of non-severe hypoglycemia in adolescents with type 1 diabetes. J Diabetes Complications. 2017;31(8):1340–1347. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.