The Randomized Linxian Dysplasia Nutrition Intervention Trial After 26 Years of Follow-up: No Effect of Multivitamin Supplementation on Mortality

Although substantial numbers of people worldwide take multivitamin supplements, including an estimated 40% or more of US adults, their effectiveness remains unclear. Recent reports from the Physicians’ Health Study (PHS) II, a randomized trial of daily multivitamins, found fewer total cancers in multivitamin recipients but no effect on overall or cause-specific mortality^1,2 in a Western population that was well nourished. However, few multivitamin trials have been conducted in undernourished populations where the potential for benefit is most likely.

In 1985, we initiated the Linxian Dysplasia Nutrition Intervention Trial (NIT) to evaluate the effect of multivitamin supplements on cancer incidence and mortality in Linxian, China, a region with extremely high rates of esophageal and gastric cardia cancer and multiple vitamin and mineral deficiencies. Individuals with a previous cytological diagnosis of esophageal squamous dysplasia were randomized to receive multivitamin supplementation or placebo for 6 years.³ Results after the 6-year intervention period showed no statistically significant benefit on mortality.⁴ However, an additional 20 years of active follow-up after cessation of the intervention gave us the opportunity to examine the long-term effects of supplementation.

This report updates the results of the Linxian Dysplasia NIT after 26 years of follow-up to provide informative data on the effect of multivitamin supplementation on mortality in an undernourished population. Our findings should be helpful for clinical practice and public health recommendations.

The Linxian Dysplasia NIT was a randomized, double-blind, placebo-controlled trial of multivitamins conducted in 1985 through 1991 in northern China in an undernourished population of 3318 persons aged 40 to 69 years who had received a previous cytologic diagnosis of esophageal squamous dysplasia. Participants were followed up for 20 additional years after cessation of supplementation. The methods³ and results⁴ for the intervention phase of this trial were previously published and are further detailed in eAppendix, eFigure 1, and eTable 1 in the Supplement.

Baseline characteristics are summarized in eTable 2 in the Supplement. Participant characteristics, including age, sex, smoking, drinking, family history of esophageal and gastric cancers, and body mass index, were similar between the supplementation and placebo groups.

A total of 2239 deaths occurred during follow-up (1985-2010), including 42% from cancer, 21% from heart disease, 25% from cerebrovascular disease, and 12% from other causes. Cumulative mortality for all causes, cancer, heart disease, and cerebrovascular disease for all participants is shown in eFigure 2 in the Supplement. Results from Cox models were similar to the cumulative mortality graphs (Table). Overall, multivitamin supplements had no effect on total mortality or mortality from any of the specific causes of death examined (including cancer mortality) among all participants.

When results were examined by subgroups defined by sex and age (Table), heart disease deaths were reduced in supplemented men (hazard ratio [HR],0.73; 95% CI, 0.56-0.96) and cerebrovascular disease deaths were increased in supplemented women (HR, 1.25; 95% CI, 1.00-1.56) (P = .047). Heart disease deaths were also decreased in older supplemented participants (HR, 0.79; 95% CI, 0.64-0.98) and cerebrovascular disease deaths were increased in younger supplemented participants (HR, 1.42; 95% CI, 1.07-1.88).

In the Linxian Dysplasia NIT, after 6 years of supplementation and nearly 20 years of additional follow-up, multivitamin supplementation had no effect on total or cause-specific mortality. Both beneficial and adverse effects on heart disease and stroke mortality were observed among subgroups defined by sex and age.

Most prior micronutrient intervention trials tested only 1 or 2 supplements. Among those that tested 3 or more vitamins and minerals, supplements reduced total mortality in 2 trials.^5,6 However, only 1 previously reported micronutrient trial was truly comparable to the Linxian Dysplasia NIT in terms of testing an existing commercially available multivitamin and mineral supplement formulation: the PHS II supplemented with Centrum Silver (Pfizer Inc) (31 vitamins and minerals), whereas the Linxian Dysplasia NIT supplemented with 2 Centrum tablets (26 vitamins and minerals). Poorly nourished populations should benefit most from multivitamin supplementation, making the present study a strong test of their potential beneficial effects. However, like the well-nourished PHS II population, no benefit of multivitamins for total mortality was observed in our study.

Our results show differences in the effect of supplementation on heart disease and stroke mortality in men and women. Multivitamin trials in well-nourished Western populations have not shown reduced heart disease in supplemented men or women. For stroke, Western trials in women either showed no effect^6-8 or suggested a benefit.⁹ For heart disease in men, the PHS II indicated no effect.² The different cardiovascular disease results observed in the Linxian Dysplasia NIT compared with other multivitamin trials may be due to differences in nutritional status, trial design, or chance.

In conclusion, during 6 years of multivitamin supplementation and 20 years of postintervention follow-up, we observed no effect of multivitamins on total or cause-specific mortality in a nutrient-deficient population. Together with data from previous trials, these results demonstrate little benefit of multivitamin supplementation on mortality in either well- or poorly nourished populations.

Corresponding Author: Philip R. Taylor, MD, Genetic Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, 6120 Executive Blvd, Room 7006, Rockville, MD 20852 (ptaylor@mail.nih.gov).

Published Online: May 27, 2013. doi: 10.1001/jamainternmed.2013.6066

Author Contributions: Drs Qiao and Taylor had access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Taylor.

Acquisition of data: Wang, Fan, Taylor.

Analysis and interpretation of data: Wang, Abnet, Qiao, Taylor.

Drafting of the manuscript: Wang, Abnet, Taylor.

Critical revision of the manuscript for important intellectual content: Wang, Abnet, Fan, Qiao, Taylor.

Statistical analysis: Abnet, Fan, Taylor.

Obtained funding: Taylor.

Administrative, technical, and material support: Abnet, Qiao, Taylor.

Study supervision: Wang, Abnet, Qiao, Taylor.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was sponsored in part by National Cancer Institute contracts N01-SC-91030 and N01-RC-47701 to the Cancer Institute and Hospital, Chinese Academy of Medical Sciences, and by the Intramural Research Program of the Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health.

Additional Contributions: Sanford M. Dawsey, MD, contributed to study concept and design, analysis and interpretation of data, drafting of the manuscript, and critical revision of the manuscript for important intellectual content; and Neal D. Freedman, PhD, contributed to the analysis and interpretation of data, drafting of the manuscript, and critical revision of the manuscript for important intellectual content. We thank the citizens of Linxian who have faithfully participated in these studies over the past 26 years.

Trial Registration: clinicaltrials.gov Identifier: NCT00342654