Can a ‘carbohydrate tolerance questionnaire’ predict outcomes from diets differing in carbohydrate content? A pilot study
Journal of Holistic Performance | ISSN: 2463-7238 | Published: 5 July 2019
Cliff J. d C. Harvey,1 Grant M. Schofield,1 Caryn Zinn,1 Simon J. Thornley1
Background. Clinical trials and experience suggest that there is a wide variation in how people respond to different dietary protocols. Clinical experience suggests that there are common signs of relative carbohydrate ‘tolerance’ that might predict cardiometabolic and anthropometric outcomes resulting from differing diets and the optimal allocation of carbohydrate restriction that might be most suited to the individual.
Objective. We believed that people with a higher carbohydrate intolerance score (CIS) determined from completing a carbohydrate tolerance questionnaire (CTQ) would achieve larger changes in cardiometabolic and anthropometric measures of health from greater magnitudes of carbohydrate restriction.
Methods. Seventy-seven healthy participants were randomised to a very low carbohydrate ketogenic diet (VLCKD), low-carbohydrate diet (LCD), or moderate-low carbohydrate diet (MCD), containing 5%, 15% and 25% total energy from carbohydrate respectively, for 12-weeks. Anthropometric and metabolic health measures were taken at baseline and 12 weeks, and symptoms of carbohydrate withdrawal and mood evaluated by questionnaires. The association between CIS and changes in anthropometric and cardiometabolic markers and mood and symptoms of carbohydrate withdrawal were made by undertaking multiple linear regression. Differences between beta coefficients describing the outcome - CIS relationship by group were determined by an interaction term, testing for significance at a p-value < 0.05.
Results. Baseline carbohydrate tolerance was associated with improvement in serum triglycerides (TG) overall, (Beta = -0.025, p = 0.073) and in the VLCKD group (Beta = -0.034, p = 0.055). The only CIS-outcome relationship to vary significantly between groups was for change in body mass index (BMI); p = 0.007, with higher carbohydrate intolerance inversely associated with the change in BMI in the MCD group (Beta = -0.309, p = 0.032). Higher CIS was also associated with more severe symptoms of carbohydrate withdrawal (Beta = 0.214, p = 0.084) and increased mood disturbance (Beta = 0.044, p = 0.060). There was also a weak association between CIS and mood disturbance in the VLCKD group (Beta = 0.083, p = 0.014).
Conclusions. Our findings demonstrate that those with higher CIS are more likely to benefit from low-carbohydrate diets for the improvement of triglyceride concentrations. Subjects with higher scores are also more likely to experience mood disturbance and symptoms of carbohydrate withdrawal. The questionnaire might be useful for clinicians to allocate those with the highest CIS to a more moderately restricted plan to mitigate symptoms of carbohydrate withdrawal and effects on mood and to offer greater improvements in BMI. However, at this time and contrary to our hypothesis, due to the lack of clear between-group significance, it is unclear whether it can accurately predict the efficacy of dietary allocations for the individual.