July 3, 2011

Good for Advocates but Bad for Science and a Terrible Model for Public Participation.

On June 21, 201, SCHER (the European Union’s Scientific Committee on Health and Environmental Risks) published the final version of a report entitled, Critical review of any new evidence on the hazard profile, health effects, and human exposure to fluoride and the fluoridating agents of drinking water.

FAN members Paul and Ellen Connett were among a number of citizens and scientists who attended a public consultation on the preliminary version of this report that was held in Brussels on September 17, 2010. Also attending were Professor Vyvyan Howard, Jennifer Luke PhD, DDS (author of a seminal work on fluoride and the pineal gland), Stephen Peckham (Hampshire Against Fluoridation), John Graham and Elizabeth McDonagh (National Pure Water Association of the UK), Robert Pocock (Irish group VOICE) and Rudolf Ziegelbecker, Jr.

All of us attended this meeting (at considerable expense and effort) because we were concerned about the very superficial nature of SCHER’s draft assessment of the health dangers posed by fluoridation. We all had hopes that armed with many references to the scientific literature, which the panel had overlooked, that we could secure a stronger final version of their report. Those hopes hinged on a rigorous scientific performance by this committee. Our hopes were not realized.

Ammunition for opponents of fluoridation

Opponents of fluoridation will be able to find sentences in the SCHER final review that will aid in the effort to fight water fluoridation. For example:

“Fluoride is not an essential element for human growth and development…”

“Systemic exposure to fluoride through drinking water is associated with an increased risk of dental and bone fluorosis in a dose-response manner without a detectable threshold.”

“Scientific evidence for the protective effect of topical fluoride application is strong, while the respective data for systemic application via drinking water are less convincing. No obvious advantage appears in favor of water fluoridation as compared with topical application of fluoride.”

“For younger children (1-6 years of age) the UL (The upper tolerable intake level) was exceeded when consuming more than 1 L of water at 0.8 mg fluoride/L (mandatory fluoridation level in Ireland) and assuming the worst case scenario for other sources. For infants up to 6 months old receiving infant formula, if the water fluoride level is higher than 0.8 mg/L, the intake of fluoride exceeds 0.1 mg/kg/day, and this level is 100 times higher than the level found in breast milk (less than 0.001 mg/kg/day).”

SCHER’s bad science

However, these statements come at the expense of a truly scientific review of both the science of fluoridation’s “effectiveness” and its “safety.” Sadly, it appears that science was not at the top of SCHER’s agenda. If it was, their response to FAN’s (and other citizens’ and scientists’ input) was truly puzzling. Take for example the way the panel dealt with our input on the extensive evidence that fluoride can damage the brain of both animals and humans.

In both oral testimony at the SCHER meeting on Sept 17, 2010, and in follow-up written testimony on Sept 24, 2010, FAN demonstrated that SCHER’s treatment of fluoride’s impact on the brain in their preliminary report was both inaccurate and superficial. However, our input made absolutely no difference and SCHER stuck to its inaccurate and superficial treatment in its final report. Here are several examples:

Animal studies

In SCHER’s preliminary report the panel states:

“There are only limited data on the neurotoxicity of fluoride in experimental animals.”

SCHER cited only two animal studies. On September 24, 2010, FAN pointed out that there have been over 100 animal studies examining the action of fluoride on the brain (119 of these are listed in Appendix 1 in the book I co-authored, The Case Against Fluoride, an electronic copy of which was sent to SCHER). FAN also provided SCHER with pdf files of 12 of the animal studies published since 2004 (SCHER requested studies published since 2005) and FAN also gave direct quotes indicating a plausible mechanism by which fluoride could interfere with brain development, including the following quote from Long et al. 2002:

“In order to investigate the molecular mechanism(s) underlying brain dysfunction
caused by chronic fluorosis, neuronal nicotinic acetylcholine receptors (nAChRs)
in the brain of rats receiving either 30 or 100 ppm fluoride in their drinking water
for 7 months were analyzed in the present study employing ligand binding and
Western blotting… Since nAChRs play major roles in cognitive processes such as
learning and memory, the decrease in the number of nAChRs caused by fluoride
toxicity may be an important factor in the mechanism of brain dysfunction in the
disorder.” (note 30 ppm is equivalent to a human drinking 3-6 ppm or less in water,
as it takes 5-10 times more fluoride to reach the same plasma levels in rats as humans, PC).

Imagine our shock when we read in SCHER’s final report (June 21, 2011), that no additional animal studies were cited and that same sentence is repeated:

“There are only limited data on the neurotoxicity of fluoride in experimental animals.”

IQ studies

In SCHER’s preliminary report, the panel states when discussing the Chinese IQ studies:

“All these papers are of a rather simplistic methodological design with no, or at best little, control for confounders, e.g. iodine or lead intake, nutritional status, housing condition, and parents level of education or income.” (emphasis added)

On September 24, 2010, FAN pointed out that one of the Chinese studies (Xiang et al. 2003 a,b) did control for both lead and iodine as well as several other important factors and estimated a threshold for IQ damage at only 1.9 ppm fluoride concentration in the drinking water.

However, in SCHER’s final report the panel neither reviews nor cites the Xiang study or any of the other Chinese, Indian or Iranian studies we had referenced. However, Xiang’s study is listed in the bibliography of studies that the SCHER panel claims to have read. If the panel had read this study as well as our comments it is puzzling how they could repeat this same phrase in their final report:

“All these papers are of a rather simplistic methodological design with no, or at best little, control for confounders, e.g. iodine or lead intake, nutritional status, housing condition, and parents level of education or income.”

SCHER ignores other brain studies cited by FAN

It gets worse. In addition to pointing out the Xiang paper, over 100 animal studies, and other IQ studies, FAN also referenced three fetal brain studies and two behavioral studies. Despite this input, in its final report SCHER made no additions to the two animal studies (Mullenix et al., 1995; Whitford et al., 2009) or the two IQ studies (Wang et al., 2007; Rocha-Amador et al., 2007) it reviewed in the preliminary report. Its final conclusion on neurotoxicity was a complete dismissal of our input on the brain, not to mention the published studies.

SCHER’s conclusion on the brain

In their conclusion to the brain section (excerpted below), I have indicated the words that were removed from the conclusion in the preliminary report in blue in parenthesis and the words added in the final report in bold:

Available human studies do not (allow concluding firmly that fluoride intake hampers children’s neurodevelopment) clearly support the conclusion that fluoride in drinking water impairs children’s neurodevelopment at levels permitted in the EU… The limited animal data can also not support the link between fluoride (noted in the epidemiological studies, at relevant non-toxic doses) exposure and neurotoxicity at relevant non-toxic doses. SCHER agrees that there is not enough evidence to conclude that fluoride in drinking water at concentrations permitted in the EU may impair the IQ of children. SCHER also agrees that a biological plausibility for the link between fluoridated water and IQ has not been established.

The difference between concentration and dose ignored

If anything the word changes represent a toning down of the conclusion presented in the preliminary report. An important change is the insertion of the phrase “at concentrations permitted in the EU.” This despite the fact that FAN had gone to some lengths to underline that there is a difference between concentration and dose, the latter being uncontrolled and potentially highly variable.

For example, a child drinking one liter of optimally fluoridated water at 1 ppm, would get a larger daily dose than a child drinking half a liter of water at 1.9 ppm (i.e. at Xiang’s threshold for lowering IQ). The former would get a daily dose of 1 mg/day the latter a dose of 0.95 mg/day.

Relevance of Xiang study to EU

In terms of the relevance of the Xiang et al.’s finding of a threshold at 1.9 ppm to the EU, we should put this into the context of levels of fluoride permitted in Europe. SCHER indicates that:

“The Council Directive 98/83/EC of 3 November 1998 (Council Directive 98/83/EC) determined a fluoride level (both natural and as a result of fluoridation) for water intended for human consumption of less than 1.5 mg/L.” (p.9)

Even if we restrict ourselves to a comparison of concentrations, 1.9 ppm allows only a minute and meaningless difference from the EU permitted level of 1.5 ppm, especially when, as mentioned above, one cannot control the dose children get because one cannot control how much water they drink, nor the additional dose they get from other sources like toothpaste.

Need for a safety margin

However, it is far worse than that. It is anticipated that will be a large range of sensitivity to a given dose of any toxic substance among individuals across a large population. This is what FAN had sent to SCHER on this matter on April 26, 2009:

In your evaluations on the fluoridation of drinking water we ask that you apply a margin of safety
analysis to those adverse end points associated with high dose levels, to determine a dose which
is protective for everyone – not just the average citizen. Normally toxicologists use a factor of 10
when extrapolating from an adverse effect, in order to protect for the usual range of sensitivity to
a toxic substance in a human population (intra-species variation). Some advocate a larger safety
factor than 10 for the protection of infants because of their extra vulnerability to toxic substances,
especially those substances known to interfere with development.

Because of this variation, we had urged the SCHER panel to apply a standard safety margin to the doses at which harm had been found in the Xiang study (or other IQ studies) so that a safe dose could be determined to protect ALL children in a large population. Normally a safety factor of 10 is used for this purpose when extrapolating from the harmful doses found in small population studies.

Despite these inputs, in its final conclusions SCHER states,

“Fluoride intake from drinking water at the level occurring in the EU does not appear to hamper children’s neurodevelopment and IQ levels” (emphasis added).

This statement gives the impression that such studies have been conducted in the EU and they have not. This conclusion is entirely based on a) its dismissal of the methodology of the 23 IQ studies conducted in India, Iran and Mexico and b) its contention that the levels of fluoride in these studies is much higher than the EU, which is not the case for the Xiang study.

Specifically, when SCHER states “that there is not enough evidence to conclude that fluoride in drinking water at concentrations permitted in the EU may impair the IQ of children” a) it is glossing over the very small difference between the EU permitted level of 1.5 ppm and Xiang’s threshold value of 1.9 ppm (as discussed above) and b) it has completely ignored the need to apply a safety margin (as discussed above).

Nor did SCHER recommend that fluoridating countries in Europe investigate the matter for themselves. The absence of studies should not be taken as absence of harm.

Other health issues ignored or downplayed

There are many other examples where SCHER ignored FAN’s (and others’) input on health issues including:

1. The scientifically unrefuted evidence that many individuals appear to be highly sensitive to fluoride (Elizabeth McDonagh, sent a copy of Dr. Bruce Spittle’s book Fluoride Fatigue to the SCHER panel).
2. The first symptoms of skeletal fluorosis are identical to arthritis (stiffness of the joints and pain in both the joints and bone) and thus it is premature to claim that there is no evidence that skeletal fluorosis is occurring in fluoridated communities in Europe. Especially since no one is investigating this possible connection in fluoridated communities.
3. Our rejection of their claim that fluoride was not an endocrine disrupter. In the preliminary report SCHER states, “fluoride is not considered to be an endocrine disruptor (ATSDR 2003).” We pointed out that the NRC (2006) panel spent a whole chapter examining this matter, and concluded that fluoride is an endocrine disruptor. Moreover, SCHER heard this information first-hand from Kathleen Thiessen, PhD, the lead author of that chapter. However, in the final SCHER report the same statement appears unchanged, “fluoride is not considered to be an endocrine disruptor (ATSDR 2003).”

A copy of our full September 24, 2010, submission can be accessed at http://www.fluoridealert.org/re/connett.sept.24.2010.pdf

The question of risk in the context of the unique practice of water fluoridation

This is what Professor Spedding Micklem (co-author of The Case Against Fluoride) wrote about the SCHER dismissal of scientific input as well as his comments on risk in the context of the unique practice of water fluoridation:

“I wasn’t at the Brussels meeting, but all my written comments, both specific and general, appear to have been disregarded. Unfortunately I’m in good company. What a waste of everyone’s time. There seem to be few changes or additions to the preliminary version, and those mostly for the worse. They further depleted the balance in the osteosarcoma discussion by referring to an abstract of unpublished work. I wonder who suggested that.

Their treatment of risk seems incompatible with their own guidelines. As I pointed out in my submission:

The Rules of Procedure for Scientific Committees, Annex V.B.19 enjoins that risks be set in the appropriate context. Artificial water fluoridation is unique in administering medication indiscriminately to large populations, in contrast to other drugs, which are prescribed for individuals with individual consent. Fluoridation therefore provides a unique context in which to consider risk. Failure to take this fully into account seems to be a major shortcoming of the Preliminary Opinion. In effect, small risks incurred by large populations can produce significant incidences of harm. It is therefore necessary to look very carefully at the evidence for risk and, where the evidence is weak or equivocal, the precautionary principle should be invoked. In the Opinion, however, risk is not considered in this way; on the contrary, the approach, particularly in relation to the thyroid, the brain and IQ, is remarkably casual. Although application of the precautionary principle may sometimes need to be tempered by some overriding benefit of the procedure under consideration, no such need arises in the case of fluoridation since the Committee justifiably concludes that fluoridation offers little benefit over topical applications. Thus, for example, the IQ studies from China and elsewhere have to be taken very seriously even when they do not exclude a sufficient range of possible confounding factors. It is not legitimate to downplay them, still less to suggest that they do not merit further investigation. Similarly, the evidence for an association of fluoride with osteosarcoma should carry decisive weight even though some studies fail to show an association. Nor can fluorosis resulting from a lifetime’s skeletal accumulation of fluoride be overlooked, even if it does not proceed to the severe crippling phase. Arguably, there should be a moratorium on fluoridation until these and other risk factors have been properly studied.”

On the benefits side of the ledger, SCHER failed to acknowledge the important submission by Rudolf Ziegelbecker Jr. that a delayed eruption of the teeth caused by fluoride ingestion, for which there is some evidence (Komarek et al., 2005), would eliminate all the benefits claimed by fluoridation promoters.

Nor did SCHER acknowledge the strongly expressed opinion of Robert Pocock that a prevalence of 4% of moderate dental fluorosis (where 100% of the enamel is damaged) in fluoridated areas was morally an unacceptable trade-off for a very small (if any) decline in tooth decay. Moreover, there is some evidence that this condition can cause psychological damage in young teenagers, which should put moderate dental fluorosis in the category of a health effect. The unacceptability of dental fluorosis was further emphasized to the committee by Doug Cross (UK Councils Against Fluoridation) in a written submission.

On the subject of dental fluorosis, Elizabeth McDonagh points out a subtle twist in the definition of dental fluorosis that appears in Appendix 1 of the SCHER report. Elizabeth cites SCHER’s definition of dental fluorosis:

“Appendix I Classification of dental fluorosis

The dictionary definition of fluorosis is ‘an abnormal condition (as mottled enamel of human teeth) caused by fluorine or its compounds’ or ‘a pathological condition resulting from an excessive intake of fluoride (usually from drinking water).’ This is a very simplistic definition since mottling of the enamel of teeth is common and may have many causes including caries, childhood infections, developmental abnormalities and trauma.”

The last sentence looks as if it has been inserted by the pro-fluoridation lobby. Elizabeth explains the sleight of hand operating here.

“We are talking about the dictionary definition of ‘dental fluorosis’ not a dictionary definition of ‘mottled teeth’. The definition is far from simplistic, indeed it is very specific. Dental fluorosis is always caused by excess of fluoride. It would not be called fluorosis otherwise. Clearly there are other causes of mottled and stained teeth but no other causes of dental fluorosis than exposure to fluoride. All the other causes of mottling could be described as amelogenesis imperfecta and classified under that term. Dental fluorosis would have its place within such a classification. Those who promote fluoridation just cite these other causes of damaged enamel to ‘muddy the waters’. Such a comment had no place in the SCHER report.”

Dr. Jennifer Luke has pointed out some elementary mistakes in SCHER’s final review, which, while not earth shattering, are disturbing indications of how little background knowledge this panel had on fluoride. For example, on page 9, SCHER claims that fluorine is the seventh most abundant element in the earth’s crust. A simple internet search (search “earth’s crust elements”) reveal at least eight sources that indicate that fluorine is not in the top 10 of the most abundant elements. Some place it at 13th others at 17th in relative abundance. Luke also points to a significant omission when, on page 13, SCHER discussed accumulation of fluoride in the teeth, bones and kidney but failed to point out that the highest concentrations occur in the calcified parts of the pineal gland (Luke, 1997, 2001).

SCHER’s snub of public participation

When SCHER issued this report on June 21, 2011, it was a sad day for science in Europe. However, it was an even sadder day for the notion that the public has a role to play in these kinds of deliberations. SCHER’s treatment of the public’s solicited input on this issue is very discouraging. We expected better.

Unfortunately, SCHER is not alone in asking for public input and treating it with disdain when it is provided. This kind of response has occurred time and time again. It happened with the Fluoridation Forum in Ireland (2000); the US Agency for Toxic Substances and Disease Registry (ATSDR, 2002); Australia’s National Health and Medical Research Council (NHMRC, 2007); the UK’s South Central Strategic Health Authority (SHA, 2009); Health Canada (2011) and is probably going to happen with NZ’s Ministry of Health (MOH), the US Department of Health and Human Services (DHHS) and the Office of Water at the US Environmental Protection Agency (OW).

The latter two agencies (HHS and EPA OW) are collaborating on the determination of a new safe drinking water standard for fluoride, but have sent signals that they will determine this value without threatening the fluoridation program. In other words from the very start of their exercise they are moving backwards from a desired conclusion (i.e. fluoridation is desirable and safe) rather than moving forwards from the best science available on fluoride’s potential harm. This point is most clearly manifested by the fact that they have ignored the potential for fluoride to lower IQ, insisting that severe dental fluorosis is the most sensitive endpoint of fluoride’s toxicity, which is nonsense. A quick review of the IQ studies shows that some children in the study groups with lowered IQ did not have severe dental fluorosis.

The two exceptions to this sorry list of agencies and review panels are the UK York Review (2000) and the US National Research Council (NRC, 2006) report, which used a science-based and balanced approach to the evidence provided by both pro- and anti-fluoridation sources.

SCHER’s Agenda?

It is difficult to fathom SCHER’s agenda, as it clearly was not scientific. It is possible that it offered little support for the benefits of fluoridation so as not to encourage the vast majority of non-fluoridating European countries to start, but by toning down the dangers they avoided the embarrassment of telling the few EU countries that fluoridate (Ireland, England and Spain) to stop this practice.

Professor Vyvyan Howard summed up the feelings of those who attended the SCHER meeting in Brussels very well, when, after examining the final report, he wrote: “ The SCHER document is political. There has clearly been political pressure to soften the message and thus support the status quo.”

A setback but not the end of the story

While recognizing that SCHER’s report gives little succor to those who would wish to expand fluoridation to other countries and communities, it represents another serious setback for those of us who would like to see the issue of fluoridation’s safety and effectiveness resolved scientifically rather than by the arbitrary use of authority.

But like a harpoon sinking into a whale’s flesh the evidence of fluoride’s harm only goes one way. With each passing month more and more research is coming out that supports our position that fluoride at low levels damages the brain and other tissues. For example:

1. Earlier this year Ding et al. (2011) published a preliminary study showing an association between lowered IQ and very low levels of fluoride in water (0.3-3 ppm).
2. A review entitled “Effects of the fluoride on the central nervous system” has just been published in Mexico by Valdez-Jiménez et al. (2011). The thoroughness with which these authors reviewed the matter and the seriousness with which they treated it is in sharp contrast to SCHER’s superficial and cavalier treatment. The authors conclude that children should avoid consuming fluoridated toothpaste and water at levels higher than 0.7 ppm.
3. Scientists from a different EPA branch (National Health and Environmental Effects Research Laboratory) classified fluoride as a “chemical having substantial evidence of developmental neurotoxicity.” (Mundy, et al.)

This news just in.

An abstract of a new study by Poureslami et al. reported a lowering of children’s IQ associated with exposure to fluoride at 2.4 ppm. This was published in the International Journal of Pediatric Dentistry (page 47). This special issue consists of abstracts of papers presented at the 23rd Congress of the International Association of Paediatric Dentistry, held in Athens, Greece, 15-18 June 2011. Poureslami et al.’s abstract reads (abstract on page 47)

High fluoride exposure in drinking water: Effect on children‘s IQ, one new report

Aim: We investigated the effect of chronic high fluoride (F) exposure on children‘s intelligence.

Design: In this cross sectional study two urban communities with similar socioeconomic and cultural status but with different levels of F in drinking water, in the Kerman province, were studied: Koohbanan city (F 2.38 mg/L), Baft city (F 0.41 mg/L). Study samples consisted of 119 children 6–9 years old: 59 children from Koohbanan and 60 children from Baft. The Raven‘s test used to determine the effect of F exposures on children‘s IQ.

Results: In the low F area (control group) the mean IQ score of children was 97.80 ± 15.95 that decreased to 91.37 ± 15.63 for the high F group (Koohbanan‘s children), which was significantly different from the control group (P < 0.05).

Conclusion: Based on the findings, the chronic exposure to high levels of F can be one of the factors that influence intellectual development.

When published this study will become the 25th study that has reported a lowering of IQ associated with moderate levels of fluoride. In this case the authors report an average reduction of three IQ points associated with exposure to water fluoride levels of 2.4 ppm.

Like Xiang et al. (2003a,b) and Ding et al. (2011), the fluoride levels in the Poureslami et al. study offer no adequate margin of safety to protect children drinking artificially fluoridated water from this serious end point.

Time will tell

It may take more time, but with the involvement of concerned and informed citizens and independent scientists the truth will come out.

Bureaucrats in government agencies need to learn that it is more important to protect the public’s health (especially vulnerable infants) than to protect fluoridation schemes. In this matter informed citizens are not the enemy but allies in this exercise.

FAN’s task is to educate citizens and lawmakers about fluoridation’s dangers. How many parents, and how many decision makers, will continue to support a practice that offers only minute, if any, benefits to children’s teeth while it has the potential to threaten their mental development?

References

ATSDR (Agency for Toxic Substances & Disease Registry). 2003. Toxicological profile for fluorides, hydrogen fluoride, and fluorine. Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service. Available at http://www.fluoridealert.org/re/atsdr-2003.pdf

Ding Y, Yanhuigao, Sun H, Han H, Wang W, Ji X, Liu X, Sun D. 2011. The relationships between low levels of urine fluoride on children’s intelligence, dental fluorosis in endemic fluorosis areas in Hulunbuir, Inner Mongolia, China. Journal of Hazardous Materials Feb 28;186(2-3):1942-6.

Long YG, Wang YN, Chen J, Jiang SF, Nordberg A, Guan ZZ. 2002. Chronic fluoride toxicity decreases the number of nicotinic acetylcholine receptors in rat brain. Neurotoxicology and Teratology 24:751-7.

Luke J. 1997. The effect of fluoride on the physiology of the pineal gland. Ph.D. Thesis. University of Surrey, Guildford. Full thesis (long) at http://www.fluoridealert.org/luke-1997.pdf

Luke J. 2001. Fluoride deposition in the aged human pineal gland. Caries Res. 35(2):125-128 Mar-Apr.

Mullenix P, Denbesten PK, Schunior A, Kernan WJ. 1995. Neurotoxicity of Sodium Fluoride in Rats. Neurotoxicology and Teratology. 17(2):169-177. Mar-Apr.

NRC (National Research Council of the National Academies). 2006. Fluoride in drinking water: a scientific review of EPA’s standards. National Academies Press, Washington D.C. Available at http://www.nap.edu/catalog.php?record_id=11571

Rocha-Amador D, et al. 2007. Decreased intelligence in children and exposure to fluoride and arsenic in drinking water. Cadernos de Saude Publica 23 Suppl 4:S579-87. Available at http://www.scielosp.org/scielo.php?script=sci_arttext&pid=S0102-311X2007001600018&lng=en&nrm=iso&tlng=en

UK York Review. McDonagh M,, et al. 2000. A systematic review of public water fluoridation. NHS Center for Reviews and Dissemination,. University of York, September 2000. Available at http://www.bmj.com/cgi/content/full/321/7265/855 – see also http://www.fluoridealert.org/york.htm

Valdez-Jiménez L, et al. 2011. Effects of the fluoride on the central nervous system. Neurologia Jun;26(5):297-300. Abstract in English, full article in Spanish.

Wang SX, et al. 2007. Arsenic and fluoride exposure in drinking water: children’s IQ and growth in Shanyin County, Shanxi Province, China. Environmental Health Perspectives 115(4):643-647. Available at http://ehp03.niehs.nih.gov/article/info:doi/10.1289/ehp.9270

Whitford G.M. et al. 2009. Appetitive-based learning in rats: Lack of effect of chronic exposure to fluoride. Neurotoxicology and Teratology 31(4):210–15.

Xiang Q, et al. 2003a. Effect of fluoride in drinking water on children’s intelligence. Fluoride 36(2):84-94. Available at http://www.fluoridealert.org/scher/xiang-2003a.pdf

Xiang Q, et al. 2003b. Blood lead of children in Wamiao-Xinhuai intelligence study. Fluoride 36(3):198-199. Available at http://www.fluoridealert.org/scher/xiang-2003b.pdf