Search

A large number of adults and children consume soy in various forms, but little information is available regarding potential neurological side effects. Prior work indicates an association between the consumption of soy-based diets and seizure prevalence in mouse models of neurological disease and in children with autism. Herein, we sought to evaluate potential associations between the consumption of soy-based formula during infancy and disease comorbidities in persons with fragile X syndrome (FXS), while controlling for potentially confounding issues, through a retrospective case-control survey study of participants with FXS enrolled in the Fragile X Online Registry with Accessible Research Database (FORWARD). There was a 25% usage rate of soy-based infant formula in the study population. We found significant associations between the consumption of soy-based infant formula and the comorbidity of autism, gastrointestinal problems (GI) and allergies. Specifically, there was a 1.5-fold higher prevalence of autism, 1.9-fold GI problems and 1.7-fold allergies in participants reporting the use of soy-based infant formula. The major reason for starting soy-based infant formula was GI problems. The average age of seizure and allergy onset occurred long after the use of soy-based infant formula. We conclude that early-life feeding with soy-based infant formula is associated with the development of several disease comorbidities in FXS.

Best practices in mail survey design advise personalizing correspondence, but most research predates the use of address-based sampling (ABS) in which a householder’s name either cannot be matched to an address or may be matched incorrectly. Further, recent evidence casts doubt on the effectiveness of personalization. The current study examines the impact of using a personalized versus a generic salutation on response rates and data quality in an ABS mail survey of the general population. A sample of 2,000 household addresses in Wisconsin was randomly selected from the U.S. Postal Service Delivery Sequence File. For households in the personalized salutation group, all materials used the surname of the household; cases in the generic salutation group referred to the state conducting the survey. While personalization was not related to response rates, it was associated with higher levels of item nonresponse. Findings support current recommendations not to use names in ABS mail surveys.

As researchers increasingly use mail survey methods to collect health and other data, research focused on optimizing features of survey design and administration in cost-effective ways is needed. We conducted two studies to examine the effects of sequential prepaid incentives and envelope messaging on response rates, item nonresponse, and costs in two mail surveys using address-based samples. Sample members in study 1 were randomly assigned to groups that received a preincentive of $2 versus $5, a second incentive of $0 versus $2 (sent only to nonresponders), and an envelope bearing no message or a monetary-focused message. Results indicated that the $5 preincentive significantly increased response rates, but neither the second incentive nor the message had an effect. Informed by the results of study 1, study 2 delivered a $2 preincentive to all sample members, but increased the amount of the second incentive to $5. Study 2 also tested a health-focused envelope message. While envelope messaging again failed to motivate participation, the administration of a second incentive, valued at a larger amount than the preincentive, significantly increased response rates. With regard to their effects on item nonresponse, while none of the variables was associated with item-missing levels in study 2, a significant three-way interaction was found in study 1 such that missing data was higher in the presence of a monetary message when a $2 preincentive was paired with a $2 second incentive. Cost analyses indicated that while larger incentives increased response rates, they also increased costs