Krill oil supplementation
We sought to determine if krill oil supplementation could maintain HDL-C in individuals following a Daniel Fast dietary plan for 21 days. This was not the case, as HDL-C decreased similarly regardless of whether krill oil or a placebo was consumed. Krill oil supplementation also did not affect any other outcome measure. Our findings therefore strongly contradict the work of Bunea et al. , who reported that krill oil supplementation (1–3 g/day for 90 days) increased HDL-C (43-60%) and decreased LDL-C (32-39%). On the other hand, our findings support the results of recent investigations [9, 12, 13] that found no effect for krill oil supplementation (2 g/day for 28 days, 1 g/day for 42 days, and 3 g/day for 49 days, respectively) on blood lipids. The baseline LDL-C values in the Bunea et al. study (165–183 mg/dL)  were considerably larger than the corresponding values of the other investigations mentioned in this paragraph (including the present study) [9, 12, 13]. This may explain why only Bunea and colleagues  found that krill oil supplementation reduced LDL-C. However, baseline HDL-C concentrations were similar across studies (also including the present study) [8, 9, 12, 13], leaving little scientific explanation for the striking improvement in this cholesterol noted by Bunea et al. . Due to the inability of krill oil supplementation to affect any outcome measure in the present study, values for subjects in both the krill oil and placebo conditions were collapsed and are discussed as such from this point forward.
Following a Daniel Fast dietary plan resulted in clinically significant reductions in LDL-C, the LDL:HDL ratio, blood glucose, blood insulin, HOMA-IR, systolic BP, and body weight. This study demonstrates that meaningful improvements to cardiovascular and glycemic parameters can be obtained in response to dietary changes in only 21 days.
Plant-based diets have consistently been shown to reduce LDL-C . The main dietary changes responsible for this are likely reductions in total fat, saturated fat, and dietary cholesterol intake, as well as an increase in fiber intake ; each of these dietary changes was observed in this study. HDL-C decreased in response to the dietary plan, and decreases in this cholesterol have been observed in other studies examining vegan diets [3, 6]. A recent investigation found that substituting monounsaturated fat for carbohydrate within a vegetarian diet increased HDL-C by 12.5% . A similar approach should be considered for future investigations of the Daniel Fast.
The reductions in blood glucose, blood insulin, and HOMA-IR in the present study were not observed in our previous investigation of the Daniel Fast . Smaller baseline values for each of the variables in the previous investigation  likely explain these discrepancies. A trial examining a vegan diet in nondiabetic, overweight and obese, postmenopausal women noted similar improvements in blood glucose and blood insulin to those of the present investigation . Vegan diets have also been shown to improve glycemic control in individuals with type 2 diabetes [3, 5, 16].
The reduction in systolic BP may have been mediated by an increase in nitric oxide, a signaling molecule that promotes vasodilation . Support for this claim comes from our observed 45% increase in the surrogate marker for nitric oxide, NOx, which is likely a consequence of increased vegetable consumption (leafy green vegetables in particular) .
There were some limitations associated with this study. First, the study lacked a control group of subjects maintaining their usual diet during the 21-day intervention phase. Many prior investigations have demonstrated that our measured outcome variables experience minimal to no change in 21 days in the absence of modification in dietary habits, exercise habits, or medication usage. Nonetheless, such changes are possible, and our failure to include a control group clouds our ability to determine how much of the changes in our measured outcome variables were due to dietary changes. Second, as mentioned above, some subjects were enrolled in the study despite using medications that may have affected outcome measures. Finally, the inclusion criteria were broad, making it difficult to generalize the results of this investigation.
A key strength of this study is that the observed changes were of sizeable magnitudes despite occurring in individuals whose average baseline values for BMI (25.9 kg/m2), LDL-C (100.6 mg/dL), blood glucose (101.4 mg/dL), and systolic BP (110.7 mm Hg) did not differ greatly from healthy norms (< 25.0 kg/m2, < 100 mg/dL, < 100 mg/dL, and < 120 mm Hg respectively). Insofar as healthy individuals have less “room for improvement”, this suggests that individuals with less-desirable values for these variables could potentially derive greater clinical benefit from this dietary plan. As mentioned above, this was observed when the results of this study were compared to those of our previous investigation of the Daniel Fast  (i.e., the greater improvements in glycemic variables noted in this study were likely due to less-healthy baseline values). Therefore, it is important that future trials examining the Daniel Fast include subject populations that would likely experience the greatest clinical benefit from this dietary plan, such as individuals with type 2 diabetes, hyperlipidemia, hypertension, and/or overweight/obesity.