Low thiamine status in adults following low-carbohydrate / ketogenic diets: a cross-sectional comparative study of micronutrient intake and status.

ABSTRACT

BACKGROUND: Low-carbohydrate diets (LCD) are popular for weight loss but lack evidence about micronutrient sufficiency in real-life use. This study assessed the intake and biochemical status of selected micronutrients in people voluntarily following LCDs. METHODS: A cross-sectional study was conducted (2018-20) among 98 adults recruited as self-reporting either LCD (n = 49) or diets not restricting carbohydrates (controls; n = 49). Diets were assessed using the 130-item EPIC-Norfolk food-frequency questionnaire. Red-blood-cell thiamine diphosphate (TDP) was measured for thiamine status using HPLC. Plasma magnesium, zinc, copper, and selenium were measured using inductively coupled plasma mass spectrometry. Between-group biomarker comparisons were conducted using ANCOVA and adjusted for age, sex, body mass index (BMI), and diabetes status. RESULTS: LCD-followers (26% male, median age 36 years, median BMI 24.2 kg/m2) reported adhering to LCDs for a median duration of 9 months (IQR 4-36). The most followed LCD type was 'their own variations of LCD' (30%), followed by ketogenic (23%), 'palaeolithic' (15%), and Atkins diets (8%). Among controls, 41% were male (median age 27 years, median BMI 23 kg/m2). Median macronutrient intakes for LCD vs control groups were carbohydrate 16%Energy (E) vs. 50%E; protein 25%E vs. 19%E; and fat 55%E vs 34%E (saturated fat 18%E vs. 11%E). Two-thirds of LCD followers (32/49) and half of the controls (24/49) reported some use of dietary supplements (p = 0.19). Among LCD-followers, assessing from food data only, 21 (43%) failed to meet the reference nutrient intake (RNI) for thiamine (vs.14% controls, p = 0.002). When thiamine from supplementation (single- or multivitamin) was included, there appeared to be no difference in thiamine intake between groups. Still, red-blood-cell TDP was lower in LCD-followers than controls (407 ± 91 vs. 633 ± 234 ng/gHb, p < 0.001). Three LCD-followers were thiamine-deficient (RBC thiamine < 275 ng/gHb) vs. one control. There were no significant differences in dietary intakes or plasma concentrations of magnesium, zinc, copper, and selenium between groups. CONCLUSIONS: Following LCDs is associated with lower thiamine intake and TDP status than diets without carbohydrate restriction, incompletely corrected by supplement use. These data, coupled with a lack of RCT evidence on body weight control, do not support recommending LCDs for weight management without appropriate guidance and diet supplementation.