Understanding the Role of Volunteerism and Nutrition in the Design of RCTs
A new report finds that older volunteers are in better physical health than younger peers. Purdue University sociologists, Kim and Ferraro, used regression to analyze the effects of volunteering, employment, attending meetings and caregiving on a biomarker of stress and inflammation, C-reactive protein, measured in the blood of 1,790 adults (57-85y) from the National Social Life, Health, and Aging Project. CRP levels were ~15% lower in those who volunteered several times per year (vs those who did not volunteer during the past year).
The Gerontologist paper provides insight into the complexity of conducting randomized controlled trials (RCTs) to test nutrient-disease relationships. It suggests that volunteers may be different from others. Sesso and colleagues reported that physicians participating in the Physicians Health Study II had a significantly lower age-adjusted cardiovascular risk rate than those who did not respond to the invitation. Because the study population were doctors, it may not be surprising that their standardized mortality ratios were significantly lower than those of the general population. But the fact that the doctors who participated (volunteered) had a 4-fold lower age-adjusted cardiovascular risk rate than those who did not respond to the invitation letter is relevant. These studies emphasize the need to reconsider the ‘gold standard’status of RCTs with respect to evidence on nutrient-disease relationships.
There could be no better time to consider RCT design than as the NIH launches the Vitamin D and Type 2 Diabetes (D2d) Study. The NIH goal is to determine if vitamin D supplementation helps prevent or delay type 2 diabetes in adults who have prediabetes. The plan is to recruit 2,500 patients with prediabetes who will be randomized to 4,000 IU vitamin D 3 daily or placebo. Will the results be applicable to the general population? Probably not as it appears from ClinicalTrials.gov (NCT01942694) that assessment of baseline vitamin D3 status isn’t a consideration. Baseline status should be used as a recruitment criteria.
The most recent meta-analysis of 21 prospective studies shows a linear relationship between 20 and 160 nmol/L with a ~4% decrease with each 10 nmol/L increase in 25(OH)D. However, the accompanying editorial by Dr H Resnick says it all, “A notable reduction in diabetes risk appeared at a 25(OH)D threshold of ∼50 nmol/L. It should be emphasized that because most of the studies in this report included participants with 25(OH)D levels <100 nmol/L, it was not possible to reliably assess the relationship of vitamin D to diabetes above this level”. So the epidemiological data linking vitamin D with diabetes risk is observed primarily in those with suboptimal serum 25(OH)D levels (< 50 nmol/L).
While it isn’t known if higher levels of serum 25(OH)D will attenuate diabetes risk, history tells us that nutrients are not drugs. Increasing nutrient intakes affects health under conditions of deficiency and sub-optimal status. Nationally, non-hispanic black (31%) and Mexican-Americans (12%) are more likely to be vitamin D deficient than non-Hispanic whites (3%) (NHANES 2003-2006 data). While randomization to treatment may be blinded, experience from the Physicians Health Study (Sesso et al, 2002) and others, (Kim & Ferraro 2013) indicate that people who participate (volunteer) and give of their time (and body samples) are special. They do not represent the world at large.
Bottom line: Nutrition RCTs should require baseline characterization and be initiated with the volunteers having nutritional status similar to the general population.
Kim S, Ferraro KF. Do productive activities reduce inflammation in later life? Multiple roles, frequency of activities, and c-reactive protein. 2013 The Gerontologist doi:10.1093/geront/gnt090
Sesso HD, Gaziano JM, VanDenburgh M, Hennekens CH, Glynn RJ, Buring JE. Comparison of baseline characteristics and mortality experience of participants and nonparticipants in a randomized clinical trial: the Physicians’ Health Study. 2002 Control Clin Trial 23:686-702
Song Y, Wang L, Pittas AG, Del Gobbo LC, Zhang C, Manson JE, Hu FB. Blood 25-hydroxy vitamin D levels and incident type 2 diabetes: A meta-analysis of prospective studies. 2013 Diab Care doi:10.2337/dc12-0962
Resnick HE. Meta-analysis supports protective role for 25-hydroxy vitamin D in diabetes. 2013 Diab Care doi:10.2337/dc13-ti05