HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Taylor, R. W Articles by Mann, J. I Search for Related Content PubMed PubMed Citation Articles by Taylor, R. W Articles by Mann, J. I Agricola Articles by Taylor, R. W Articles by Mann, J. I

Two-year follow-up of an obesity prevention initiative in children: the APPLE project^1,2,3

¹ From the Edgar National Centre for Diabetes Research (RWT, KAM, VLF, and JIM) and the Departments of Human Nutrition (WB and JIM) and Preventive and Social Medicine (SMW), University of Otago, Dunedin, New Zealand

² Supported by the Health Research Council, the National Heart Foundation, The Community Trust of Otago, The University of Otago, and the Otago Diabetes Research Trust.

³ Reprints not available. Address correspondence to RW Taylor, Edgar National Centre for Diabetes Research, C/- Department of Human Nutrition, University of Otago, PO Box 56 Dunedin 9054, New Zealand. E-mail: rachael.taylor{at}otago.ac.nz.

	ABSTRACT

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Background:In a 2-y intervention targeting increased physical activity and healthy eating in primary school children, the adjusted body mass index (BMI) z score was 0.26 units (95% CI: 0.21, 0.32) lower in intervention than in control children. Few obesity prevention initiatives in children have undertaken follow-up analyses.

Objective:The objective was to determine whether differences in BMI persisted 2 y after the cessation of the intervention.

Design:All children who had at least one measurement of height and weight at any time during the study (baseline and years 1 or 2) were invited to participate in follow-up measurements (height and weight).

Results:Five hundred fifty-four of 727 eligible children (76%) participated. Children who refused to participate (n = 14) or had moved from the study area (n = 159) did not differ from the remaining participants in baseline age, sex, or BMI. The mean BMI z score (and 95% CI) remained significantly lower in intervention children at follow-up in the whole group (n = 554; –0.17; –0.25, –0.08) and in the group who underwent at least 1 (n = 389; –0.19; –0.24, –0.13) or 2 (n = 256; –0.21; –0.29, –0.14) full years of intervention. Intervention children were less likely to be overweight, but only in those who were present for the full intervention (n = 256; RR: 0.81; 95% CI: 0.69, 0.94).

Conclusion:Despite the main intervention initiative (school-based activity coordinators charged with the responsibility of enhancing physical activity and promoting healthy eating) being discontinued at the end of the intervention, continued benefits to BMI remained apparent in intervention children 2 y later.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

 
The continuing epidemic of overweight and obesity in children underscores the urgent need for population approaches as well as sustainable programs aimed at reducing the prevalence of overweight and/or the distribution of weight within the population. Recent reviews (1-3) highlight the lack of information regarding the success of population measures. Although successful outcomes from some programs have been reported, little is known about the long-term effectiveness of various approaches. Few obesity prevention initiatives worldwide have undertaken follow-up analyses after termination of the intervention programs (4-10), presumably because of the lack of success at the end of the intervention period. Yet, determining long-term sustainability is critical to discerning the approaches that offer the best hope for the future. Despite considerable differences in the ages of the participants, intervention approaches, and length of follow-up, 3 studies have shown either continued benefits to the body mass index (BMI) in intervention children (6, 7) or enhanced remission of overweight (9). The remaining 3 studies that showed no benefits of intervention at follow-up also observed no benefits immediately after intervention (4, 5, 8). Gutin et al (10) found that the favorable differences in percentage body fat observed during the school year disappeared during the summer break in students involved in an after-school activity program.

APPLE (A Pilot Programme for Lifestyle and Exercise) was a 2-y community-based obesity prevention initiative in primary school children. Intervention components involved the provision of Activity Coordinators based at each intervention school. They were responsible for facilitating activity programs involving parents and other community volunteers and for basic nutrition education focusing on reducing the intake of sugary drinks and increasing the intake of fruit and vegetables (11). Community involvement ensured awareness of the program beyond that achieved during the school day. At 2 y, significant differences were observed between intervention and control children in BMI, waist circumference, and systolic blood pressure (12). The aim of this study was to determine whether the observed benefits observed to BMI were still apparent 2 y after cessation of the intervention.

	SUBJECTS AND METHODS

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Subjects
APPLE (Australian Clinical Trials Registry Number 12605000578606) was described in detail elsewhere (11, 12). Ethical approval was obtained from the University of Otago Ethics Committee, and all parents or guardians and the older children themselves gave informed consent. Design and reporting procedures followed the Transparent Reporting of Evaluations with Non-Randomized Designs (TREND) statement (13). Control and intervention communities were not randomly selected but were geographically separate in an effort to reduce contamination. All children enrolled at the 4 intervention and 3 control schools as of August 2003, 2004, and 2005 were invited to participate in the intervention. Approximately 90% of eligible children enrolled in the intervention and control schools participated in the baseline measurements: 82.6% were white, 16.5% were Maori, and <1% were Pacific Islanders. Ethnic distribution did not differ by intervention allocation (chi-square test, P = 0.133) or sex (chi-square test, P = 0.108) (12).

Intervention components
APPLE was a multifaceted intervention with the aim of increasing physical activity, increasing the intake of fruit and vegetables, and reducing the intake of sugary drinks. The main intervention initiative in both years was the assignment of Community Activity Coordinators to each intervention school (0.5 FTE per school). They were employed to increase noncurricular activity at recess, lunchtime, and after school. Some time was spent on curricular-based activities. Other activity-related interventions concerned the development of resources facilitating short bursts of activity in class and the increased availability of sports equipment in an effort to encourage "free play." Nutrition-based initiatives were particularly emphasized in the second year of the intervention and included the provision of cooled water filters in each school, science lessons highlighting the adverse health effects of sugary drinks, a community-based healthy eating resource, a novel interactive card game, and the provision of free fruit for 6 mo (12). Control schools received a payment of NZ$500–1000 (depending on school size) for the purchase of school equipment as reimbursement for the time required to conduct measurements on children during school.

Physical measurements
All measurements were made in duplicate during school hours at baseline, 1 y, 2 y (end of intervention), and follow-up. Height was measured with a portable stadiometer (Wedderburn, Dunedin) to the nearest 0.1 cm and weight with an electronic scale (Tanita TI1618) to the nearest 0.1 kg. All measurements were made, while the children were wearing light clothing and no shoes, using the same equipment and procedures as used in the recent New Zealand National Children's Nutrition Survey (14). BMI was calculated as weight (in kg) divided by height squared (in m) and z scores calculated according to 2000 Centers for Disease Control and Prevention reference norms (15). Children with a BMI 85th percentile for age and sex were classified as overweight.

Follow-up
All children with at least one measurement of height and weight (whether at baseline, 1 y, or 2 y) were eligible to participate in the follow-up measurements (n = 381 intervention and 346 control children). Efforts were made to recontact all children living within a 200-km radius of Dunedin (original study site) via existing study addresses, the electoral roll, telephone directories, and information from participating schools. Children from the 7 initial study schools were now attending 44 schools within the local area. Schools often had to be visited several times to ensure that all possible children could be remeasured. Only height and weight were obtained at follow-up. Interviews were held with each intervention school principal to determine whether any components of the intervention had continued once the intervention had ceased and any other changes to the food and activity environment within the school and wider community. Principals also completed a questionnaire regarding policies and procedures for healthy eating and physical activity within each school community setting, information that was also obtained at baseline (16). We also considered the potential influence of any regional or national healthy eating and activity initiatives that were underway during the follow-up period.

Statistics
The power calculations for the original study were based on the number of pupils available in the 7 participating schools (250 in each area). An intraclass correlation of 0.05 estimated from an earlier Dunedin study meant that, with 7 schools, the inflation factor was 4.6. The correlation between repeated measures of BMI was used to adjust the error root mean square. Calculations based on these values suggested that our study had the potential to detect an effect size of 0.3 in any of our measures with 80% power at the 5% level of significance. Because the SD of BMI increases with age, we used z scores for BMI (derived from Centers for Disease Control and Prevention tables), which take into account age and sex as the principal outcome measure. Because schools, not students, were the sampling unit, generalized estimating equations with an exchangeable covariance matrix were used to analyze the data (15). Robust SEs were used to estimate the CIs and P values. Generalized estimating equations, using the Poisson distribution and robust SEs, were used to obtain relative risks for the categorical variable overweight obtained by using the Centers for Disease Control and Prevention cutoff of the 85th percentile (15). These results are presented as relative risks. All models were adjusted for age, baseline measures, sex, year of recruitment, time in the study, and change of school. All data were analyzed by using STATA 161 release 10.0 and 162 (StataCorp, Stata Statistical Software, College Station, TX).

Because schools often have very mobile populations and this was a community-based intervention, we have presented data for 1) those children who were present for the whole 2-y intervention (n = 256; Table 1), 2) those children who were present for 1 y of the intervention (n = 389; Table 2), and 3) all children regardless of whether they participated in the actual intervention phase (n = 554; Table 3).

View this table: [in this window] [in a new window]   TABLE 3. Characteristics of the study population at each time point: includes children who were present for 1 y of the intervention and for follow-up1

	RESULTS

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

Characteristics of the subjects who completed the full 2 y of the intervention are presented in Tables 1 and 2 (34% of control and 36% of eligible intervention children; P > 0.05 for difference). The control group contained a greater proportion of girls (P = 0.017), and the total time in the study was significantly shorter, albeit by a small amount (3.6 ± 0.1 compared with 3.8 ± 0.1 y; P < 0.001). At baseline, more control children were overweight, but mean z scores for height, weight, and BMI did not differ. The mean adjusted BMI z score was significantly lower in the intervention than in the control children by 0.30 units (95% CI: 0.24, 0.36) at the end of the intervention and by 0.21 units (0.14, 0.29) at follow-up. The differences in BMI z scores were not due to variations in height z scores among participants, but rather to differences in weight (Table 1). Intervention children were also 19% less likely to be overweight at follow-up after adjustment for confounding variables (Table 1). Rates of remission and the incidence of overweight (BMI 85th) were also calculated for the intervention and control children. Ten (14%) control and 9 (10%) intervention children became overweight during the 2 y, and 14 (25%) control and 12 (30%) intervention children that were overweight at baseline were not overweight at follow-up.

The results for those children who experienced at least one full year of the intervention program and were present at follow-up (53% of eligible intervention and 54% of eligible control children) are shown in Table 3 and Table 4. As was observed in the previous analysis, several characteristics varied between the intervention and control children at baseline (Table 1) and were adjusted for in all analyses. In this group, the mean adjusted BMI z score was significantly lower in the intervention children than in the control children by 0.22 units (95% CI: 0.16, 0.28) at the end of the intervention and by 0.19 units (0.13, 0.24) at follow-up. Follow-up occurred an average of 1.8 y after their last intervention measure (P = 0.06 for difference between intervention and control). Although the likelihood of being overweight was less in intervention children at the end of the intervention (RR: 0.82; 95% CI: 0.70, 0.96), it was no longer significant at follow-up (Table 2). In this group, 18 (17%) control and 13 (10%) intervention children became overweight, and 20 (24%) control and 20 (30%) intervention children became normal weight.

View this table: [in this window] [in a new window]   TABLE 4. Differences (and 95% CIs) between the intervention and control children at the end of the intervention and at follow-up for those children who participated in 1 y of the intervention and follow-up (n = 389)

Table 5

Table 6

	DISCUSSION

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

We report here a follow-up of the APPLE study, which has shown the continuing benefit of a relatively inexpensive program aimed at reducing the risk of excessive weight gain in children nearly 2 y after the completion of a formal intervention phase. School-based coordinators (one at each intervention school) had been responsible during a 2-y period for a program designed to enhance extracurricular physical activity (during and after the end of the school day) and to promote healthy eating to pupils, teachers, parents, and the local community. The initial reduction in adjusted BMI z score (0.22–0.30 units) in the intervention children relative to the control children was largely sustained despite the termination of employment of the coordinators (follow-up differences of –0.17 to –0.21 BMI z score units). In children who were present for the full 2 y of intervention, the prevalence of overweight was also significantly reduced by 19%. Whereas the largest effect was shown in those children who were actually present to receive the intervention in its entirety, this analysis may overestimate the true effect of the intervention at a wider community level. Thus, we also presented analyses for children who were present for 1 y of the intervention and in all children, including those who may have been recruited at the end of the intervention. It is of interest and encouraging that there was little difference in outcome when considering those who had had the opportunity of experiencing 1 to 2 y of the intervention program and those with only a shorter period of participation. The relative biological importance of a difference in BMI z score of 0.2 is difficult to describe. Such an effect translates into differences in BMI of 0.4 in 7-y-olds and of 0.5 in 11-y-olds of median height (12). In comparison, mean BMI increased by approximately twice this amount in US children from 1963–1970 (National Health Examination Survey II/III) to 1988–1994 (National Health and Nutrition Examination Survey III) (26).

We believe that the involvement of both the school and wider communities and the introduction of concurrent national initiatives are the most likely reasons for the sustained improvement. Schools had made changes to some of the policies and procedures targeting healthy eating and activity. In particular, reductions in access or types of foods available for purchase were most common. One school was working with the local fast food store to meet new Ministry of Health guidelines aimed at providing healthier options for foods commonly consumed by children at school (www.moh.govt.nz, Food and Beverage Classification system). Another had discontinued a barbeque held each week and were finding alternative businesses to provide bought lunches for children. Another had run a very successful healthy food stall at a local community activity event and edible gardens were still ongoing. One school had introduced an overall nutrition policy that had not had one before, and 2 schools had removed access to all fried and takeaway foods. Several activity initiatives introduced as part of the intervention were still apparent. One school had found the funding to retain their coordinator (at a reduced level of 0.3 FTE) because of the perceived importance of this role. Friendship buses were still strong in some schools, which gave senior students the responsibility of ensuring that younger children were included in activities at breaks and lunchtimes. Parental involvement in sports and activities remained strong, and several activities introduced or enhanced by the coordinators continued, including triathalons and cross-country training, line dancing, and active transport to school. One school reported being more aware of the importance of keeping children active during breaks and had reintroduced tackle rugby for boys and supervised swimming during lunch breaks to increase activity. Various government initiatives were also introduced during the tenure of the APPLE project, principally under the umbrella of the Healthy Eating, Healthy Action campaign, which commenced in 2003. In late 2006, Mission-On was launched, a broad-based package of 10 initiatives that aim to improve the well-being of young people, particularly their nutrition and physical activity levels (http://mission-on.tki.org.nz/). It is difficult to evaluate how such initiatives may have influenced our findings, although there is no reason to believe that such national initiatives would have affected control and intervention children differently.

A limited number of studies have examined the extent to which the effort of intensive intervention persists after the conclusion of the active intervention phase. Some have reported few longer-term benefits (4, 5, 8, 27), and others (6, 7, 9, 10) suggest a varying persistent effect. Given the appreciably different experimental designs, it is difficult to disentangle consistent threads that might predict positive outcomes. Nevertheless, in as much as it is possible to generalize from a relatively small number of disparate studies, it appears that 2 attributes characterize those studies in which some persisting benefit has been observed. For the 4 previously published successful reports (6, 7, 9, 10) and in APPLE (11, 12), the intervention was started in young (prepubertal) children and there appeared to be at least some involvement of parents and or the community as a whole. On the other hand, in the studies carried out in older children or when there was little evidence of involvement outside the school environment, changes in measures of overweight or obesity were less likely to persist (4, 5, 8, 27). Thus, it appears that age and parental and community involvement may be particularly relevant.

The APPLE project has limitations, which are inevitably associated with control and intervention communities based on geographic location rather than on random selection. However, we considered that relative avoidance of contamination was an important prerequisite to achieving meaningful results and believe that many consequential concerns have been overcome by following the design and reporting procedures recommended in the TREND statement (13). Strict compliance with the classic prerequisites of a randomized controlled trial is nearly impossible in a community-based intervention and during follow-up. We believe that reasonably high rates of follow up have been achieved for a study of this type and that this provides further evidence for the reliability of our findings. In addition, similar findings were observed in all 3 analyses, which provide further support for a real intervention benefit. APPLE was designed as a demonstration project, that is, with sufficient power to determine whether differences in the major outcome variables were significant (mean BMI z score), but not necessarily large enough to complete various subgroup analyses. For example, we were unable to calculate remission and incidence of overweight because of the sample size. Data were also analyzed after controlling for sex, rather than conducting separate analyses in boys and girls to increase the power. However, sex was not a significant predictor of difference in BMI z score in any of the analyses presented (data not shown). We did not obtain measures of pubertal staging or socioeconomic status (apart from school decile ratings), and follow-up measurements of waist circumference, diet, and physical activity were not completed in an effort to reduce respondent burden and school involvement.

There is no doubt that local and national governments and the international food industry are critically important to attempts to reverse the escalating rates of overweight and obesity in young persons. However, our findings suggest that local initiatives may also be worthwhile and may be most appropriately directed at the prepubertal age group, with parental and community involvement being an important component of the intervention.

	ACKNOWLEDGMENTS

 
We thank the many children and families who participated in APPLE and especially in this follow-up study.

The authors' responsibilities were as follows—RWT, KAM, and JIM (Principal Investigators): participated in the study conceptualization and ongoing project management; SMW: completed all statistical analyses; and WB (Project Coordinator) and VF: completed the follow-up measures. All authors contributed to the writing of the manuscript. None of the authors had any conflicts of interest.

	REFERENCES

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Summerbell CD, Waters E, Edmunds LD, Kelly S, Brown T, Campbell KJ. Interventions for preventing obesity in children. Cochrane Database Syst Rev 2005; Issue 3. CD001871. DOI: 10.1002/14651858.CD001871.pub2.
2. Flodmark CE, Marcus C, Britton M. Interventions to prevent obesity in children and adolescents: a systematic literature review. Int J Obes (Lond) 2006;30:579–89.[CrossRef][Medline]
3. Lissau I. Prevention of overweight in the school arena. Acta Paediatr Suppl 2007;96:12–8.[Medline]
4. Nader PR, Stone EJ, Lytle LA, et al. Three-year maintenance of improved diet and physical activity: the CATCH cohort Child and Adolescent Trial for Cardiovascular Health. Arch Pediatr Adolesc Med 1999;153:695–704.[Abstract/Free Full Text]
5. Neumark-Sztainer D, Story M, Hannan PJ, Rex J. New Moves: a school-based obesity prevention program for adolescent girls. Prev Med 2003;37:41–51.[CrossRef][Medline]
6. Fitzgibbon ML, Stolley MR, Schiffer L, Van Horn LV, KauferChristoffel K, Dyer A. Two-year follow-up results for Hip-Hop to Health Jr: a randomised controlled trial for overweight prevention in preschool minority children. J Pediatr 2005;146:618–25.[CrossRef][Medline]
7. Kafatos A, Manios Y, Moschandreas J, and the Preventive Medicine & Nutrition Clinic University of Crete Research Team. Health and nutrition education in primary schools in Crete: follow-up changes in body mass index and overweight status. Eur J Clin Nutr 2005;59:1090–2.[CrossRef][Medline]
8. Reilly JJ, Kelly L, Montgomery C, et al. Physical activity to prevent obesity in young children: cluster randomised controlled trial. BMJ 2006;333:1041–5.[Abstract/Free Full Text]
9. Danielzik S, Pust S, Müller MJ. School-based interventions to prevent overweight and obesity in prepubertal children: process and 4-years outcome evaluation of the Kiel Obesity Prevention Study (KOPS). Acta Paediatr Suppl 2007;96:19–25.[Medline]
10. Gutin B, Yin Z, Johnson M, Barbeau P. Preliminary findings of the effect of a 3-year after-school physical activity intervention on fitness and body fat: the Medical College of Georgia Fitkid Project. Int J Pediatr Obes 2008;3:3–9.[Medline]
11. Taylor RW, McAuley KA, Williams SM, Barbezat W, Nielsen G, Mann JI. Reducing weight gain in children through enhancing physical activity and nutrition: the APPLE project. Int J Pediatr Obes 2006;1:146–52.[Medline]
12. Taylor RW, McAuley KA, Barbezat W, Strong A, Williams SM, Mann JI. APPLE project: 2-y findings of a community-based obesity prevention program in primary school-age children. Am J Clin Nutr 2007;86:735–42.[Abstract/Free Full Text]
13. Des Jarlais DC, Lyles C, Crepaz N, and the TREND group. Improving the reporting quality of nonrandomized evaluations of behavioral and public health interventions: the TREND statement. Am J Public Health 2004;94:361–6.[Abstract/Free Full Text]
14. Parnell W, Scragg R, Wilson N, Schaaf D, Fitzgerald E. NZ Food NZ Children: Key results of the 2002 National Children's Nutrition Survey. Wellington, New Zealand: Ministry of Health, 2003:1–267.
15. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC growth charts for the United States: methods and development. Data from the National Health Survey. Vital Health Stat 11 2002;246:1–190.
16. Williden M, Taylor RW, McAuley KA, Simpson JC, Oakley M, Mann JI. The APPLE project: an investigation of the barriers and promoters of healthy eating and physical activity in children aged 5-12 years. Health Educ J 2006;65:135–48.[Abstract/Free Full Text]
17. Swinburn B, Egger G, Raza F. Dissecting obesogenic environments: the development and application of a framework for identifying and prioritizing environmental interventions for obesity. Prev Med 1999;29:563–70.[CrossRef][Medline]
18. Swinburn B, Egger G. Preventive strategies against weight gain and obesity. Obes Rev 2002;3:289–301.[CrossRef][Medline]
19. Jain A, Sherman SN, Chamberlin LA, Carter Y, Powers SW, Whitaker RC. Why don't low-income mothers worry about their preschoolers being overweight? Pediatrics 2001;107:1138–46.[Abstract/Free Full Text]
20. Borys JM, Lommez A, Guinot S, Raffin S, and the FLVS Study Group. Long-term outcomes of a successful childhood obesity prevention program in France. Obes Rev 2006;7:28.
21. Williamson DA, Copeland AL, Anton SD, et al. Wise Mind project: a school-based environmental approach for preventing weight gain in children. Obesity (Silver Spring) 2007;15:906–17.[CrossRef][Medline]
22. Salmon J, Ball K, Hume C, Booth M, Crawford D. Outcomes of a group-randomized trial to prevent excess weight gain, reduce screen behaviours and promote physical activity in 10-year-old children: Switch-Play. Int J Obes (Lond) 2008;32:601–12.[Medline]
23. Martínez Vizcaíno V, Salcedo Aguilar F, Franquelo Gutiérrez R, et al. Assessment of an after-school physical activity program to prevent obesity among 9- to 10-year-old children: a cluster randomized trial. Int J Obes (Lond) 2008;32:12–22.[Medline]
24. Foster GD, Sherman S, Borradaile KE, et al. A policy-based school intervention to prevent overweight and obesity. Pediatrics 2008;121:e794–802.[Abstract/Free Full Text]
25. Plachta-Danielzik S, Pust S, Asbeck I, et al. Four-year follow-up of school-based intervention on overweight children: the KOPS study. Obesity (Silver Spring) 2007;15:3159–69.[Medline]
26. Flegal KM, Troiano RP. Changes in the distribution of body mass index of adults and children in the US population. Int J Obes Relat Metab Disord 2000;24:807–18.[CrossRef][Medline]
27. James J, Thomas P, Kerr D. Preventing childhood obesity: two year follow-up results from the Christchurch obesity prevention programme in schools (CHOPPS). BMJ 2007;335:762.[Abstract/Free Full Text]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Taylor, R. W Articles by Mann, J. I Search for Related Content PubMed PubMed Citation Articles by Taylor, R. W Articles by Mann, J. I Agricola Articles by Taylor, R. W Articles by Mann, J. I