This CER study was initially designed to compare changes in BMD in a bone health plan that incorporated the three components recommended in the SG's "call to action" (improved nutrition, increased health literacy, and increased physical activity) with expected changes in BMD as reported in non-intervention studies. In addition to the primary outcome measure of changes in BMD, measures of safety, volunteer biases, dropout effects, and effects of compliance were also examined. Upon completion of this initial study, a second study was commissioned to compare the effects of following the same bone-health plan, but with a different version of the calcium bone-health supplement.
Although the sequential design, as opposed to an RCT, posed difficulties in interpreting the data, the results suggests that following the AC-2 plan led to significantly greater increases in BMD than expected and than the AC-1 plan. This conclusion is based on between-group comparisons of the MAPC in: (A) AC-1 as compared to (B) AC-2, an untreated age- and gender-adjusted expected change control group, and (C) between compliant sub-groups in AC-1 and AC-2. Support for The Plan's efficacy is also provided by the within-group comparisons of changes from baseline and between compliant and partially-compliant sub-groups, a finding consistent with an exhaustive meta analysis of 23 trials (n = 41,419) of the effects of supplementation on BMD . These researchers concluded that poor compliance is the major obstacle to obtaining full benefit of supplementation and that compliant subjects doubled their risk reduction, suggesting high compliance is needed to demonstrate the therapeutic efficacy of supplementation. Support for the safety of the AC-1 Plan is provided by the absence of adverse events or changes from baseline in the QOL, daily tracking reports and the 43-chemistry blood panel.
One apparent difficulty in interpreting the findings is the loss of 1.02% of MAPC in the partially-compliant sub-group following the AC-1 plan (bar-graph #3, Figure 2). Although the bar-graph suggests that this sub-group lost more BMD than expected (bar-graph #1), the difference was not statistically significant (p = 0.29) suggesting that changes in the AC-1 plan were no different than expected. The most parsimonious explanation for the absence of any change in BMD for the partially compliant sub-group taking AC-1 is that this bone-health plan had no effect on BMD when subjects only partially adhered to The Plan, but did facilitate change among more compliant subjects. Conversely, when following The Plan with AC-2, even the partially-compliant subjects increased their BMD and subjects classified as compliant had greater increases than partially-compliant subjects taking AC-2. Thus, the data suggest that there may be a threshold below which no changes in BMD occur and above which changes do occur. The threshold appears to be between partially compliant and compliant subjects taking AC-1.
Study Weaknesses and Mitigating Factors
Although these data support a comparative effectiveness interpretation of the superiority of the AC-2 plan, absent a placebo-controlled arm, one could conclude that the increased MAPC from baseline and over-expected was attributable to using a sequential design resulting in unequal subject groups, a placebo effect or invalid expected change data. With regard to the equivalence of the groups, while it was impossible to rule out all potentially confounding variables, increased confidence in the similarity of the two groups was obtained from comparisons of a number of baseline measures associated with changes in BMD. There were no statistical differences between the groups on age, gender and BMD. Nor were there any significant baseline and pre- post-study differences on body composition variables that have been reported to affect BMD  (weight, lean mass, % fat, and BMI). Additionally, no differences were found between the groups on the QOL, lipid panel, C-reactive protein, serum calcium or thyroid levels. Of the 43 blood chemistries measured in both groups, only two (platelets and alkaline phosphatase) differed between the groups (p < 0.01). In neither group were there differences between volunteers and non-volunteers, nor between those subjects who completed versus those who dropped out. Taken together, these similarities provide considerable evidence that the two groups were reasonably equivalent at baseline.
With regard to placebo effects, it seems implausible to suggest that the reported changes in BMD were the result of placebo effects, particularly in view of a number of studies comparing changes in BMD that showed virtually no change in the placebo arms. For example, three randomized double-blinded placebo-controlled studies measuring the effects of strontium ranelate [27–29] found a progressive and linear decrease in BMD in each of three years with a 1% decline after 12 months--a decline virtually identical to the two age-gender adjusted expected changes used in this study.
In contrast to the absence of studies on placebo effects on BMD, a number of studies have suggested that consumption of the multiple nutrients in AC-1 and AC-2 could facilitate increases in BMD. Although we could find no studies on the effects on BMD of strontium citrate used in both formulas (as opposed to the plethora of studies on strontium ranelate), considerable evidence is available on the bone-health effects of the other nutrients in the AC formula--supplemental magnesium, vitamin K-2, [30, 31] and calcium and Vitamin D3 . Vitamin C has also been reported as an essential nutrient for collagen formation and normal bone development, particularly in older men and women [33, 34]. Further support for the increased BMD may be because AC is a plant-sourced supplement and some studies have suggested plant-sourced minerals may be more easily absorbed than non-plant-sourced calcium and minerals [35–39] suggesting that the body was able to use less than 10 percent of the synthetic minerals contained in the most popular brands of multivitamins as opposed to over 80 percent of minerals derived from plant sources. Other studies have also reported positive associations between fruit and vegetable consumption and BMD in elderly adults [40, 41], adolescents  and children .
With regard to the validity of the expected change data, the studies cited above suggest that the annual expected decrease in BMD is closer to -1.0% as opposed to the -0.67% and -0.63% used in this study. Additionally, although some studies have reported that supplementation with vitamin D3 and calcium had no effect on the decline of age-related BMD [44, 45], the general consensus is that supplementation does result in a lower rate of annual bone loss . These data would suggest that even with supplementation, the expected annualized change in BMD is between the -1.0% and -0.2%, particularly since there is no compelling evidence that supplementation leads to an increase in BMD.
No attempt was made to partition the effects of the three components of The Plan, since the goal of the study was to examine the effectiveness of the plans, not the individual components in the plans. However, the increased MAPC found in AC-2, as compared to AC-1, suggests that the modifications made to the nutritional profile of AC-2, while holding all other components constant, provided additional benefits over and above the benefits provided by the other components of The Plan.
One strength of this study is that it was conducted in "real world" conditions which maximized the inclusion criteria and minimized the exclusion criteria by enrolling adults of all age, gender and ethnicity, which increases confidence that the results could be generalized to populations that are most likely to use the product. Other strengths include the well established reliability and validity of DXA measurements of BMD, the analysis and absence of evidence of volunteer and attrition biases, baseline similarities between the two treatment groups, consistency of the expected within-groups differences in compliance, and the experience levels of the testing and research technicians.
With regard to safety, the use of pre- and post-study QOL inventories and independently-measured blood chemistries completed by both groups at baseline and end-of-study contribute to the safety of the study, as does the absence of reported adverse effects on these measures and on the daily tracking forms.