AWA: Academic Writing at Auckland

Anorexia and its effects on human growth and development

Introduction

Eating disorders are a well known problem in western society today that can take a huge toll on family and society as a whole. Anorexia alone is the third most common chronic illness in adolescent girls affecting 0.2-1% of our population typically in the higher socioeconomic status groups (Prabhakaran et al, 2008).  It is common place to hear about Anorexia Nervosa (AN) or Bulimia Nervosa and have it associated with ideas of starvation and self emaciation as well as ideas of body image and perfectionism. But what are less well known are the long term effects on growth, development and reproduction or the links between culture and the development of these diseases. Eating disorders are very much a psychological disease that are thought to be provoked by consumerism, individualism, and the idealism of strict self discipline all values promoted by western society (Lester, 2007). In a study of the eating disorder clinics in Mexico, Lester (2007) showed an association between the acculturation of post industrialist capitalist ideals and a rise in the occurrences of eating disorders. This is an important issue that needs to be addressed because of the global spread in this western way of thinking and the threat of an associated rise in eating disorders which can mean poorer health for individuals. AN is also a good comparison for understanding starvation patterns in nutrient stressed societies as it can provide a possible example of dietary restriction and its effects, isolated from other factors that are generally associated with nutrient deprived families such as poorer quality of food or stresses caused by socioeconomic status etc. However it is not a perfect example there are other stresses associated with AN that can have alternative side effects but AN can still help us to understand bodily development under nutrient restriction in a different light.

Literature review

Eating disorders can have a number of effects on the female body’s functioning. A widely known fact is one that of the main characteristics of AN is a decrease (or cessation during adolescence) of body weight, but it can effect more than this. It can affect the reproductive cycle, changing menstruation patterns. It may also effect neonate development which may have a subsequent impact on the development of the next generation (Kouba et al, 2005). The effects can include a reduction in birth weight, small head circumference (Kouba et al, 2005), possible preterm births, increased risk of a cesarean section, or miscarriage (Bulik 1999 cited in Kouba et al, 2005). These effects are observable in typical starvation settings too. A study by Ellison (2003) looked at energy stores and how these have an effect on reproduction patterns in different populations over varying levels of nutrient abundance. Ellison (2003) theorized energy could be usefully categorized in three ways: energy status which is the amount of stored energy, energy flux which is the amount of energy turnover (i.e. athletes have high flux and anorectics have low flux), and energy balance which is the net residual that can be invested into reproduction. Ellison (2003) tested for variation in ovarian function. He showed how energy flux and energy balance are better associated with this, not energy status that was previously suspected. This is linked with milk production after birth which can cause high energy flux within the body and subsequently lower ovarian function and the probability of getting pregnant (Ellison, 2003). This can also be linked to seasonal variations in resource availability where energy balance decreases in times of lower resources, for example pre-harvest season for the Lese women in the Democratic Republic of the Congo (Ellison, 2003). Among the Lese women a significant pattern of conceptions occur on a yearly basis. Overall Ellison (2003) found that a sustainable energy supply was more important to reproductive success, not a large pre-pregnancy supply. This is how the neonate obtains its energy for growth. If it does not maintain constant nutrient intake and is starved later in the pregnancy, this can lead to a reduction in brain size because this is the period where a lot of the energy is put into brain growth (Ellison, 2003). The neonate can also in extreme cases begin to supply its own fat to support its brain development, and the processes that trigger this, can also trigger early labor so the infant can get out instead of wasting away in the womb (Ellison, 2003).

Along with this AN can also have a long term effect on the subjects own bodily development, particularly if AN began during childhood and adolescence. This can lead to permanent markers in adulthood. As well as the usual decrease in body weight the individual can experience changes in hormone levels that can affect the course of growth within the body (Prabhakaran et al, 2008). Low caloric intake during growth in childhood and adolescence has been linked with a reduction in average adult height, and this is true of males with eating disorders too (Modan-Moses, 2003). This is thought to be because of a reduction in insulin like growth factor-1 which is one of the hormones that modulate growth (Prabhakaran et al, 2008), however this is debated. Other studies have shown disassociations between eating disorders and peak heights in adulthood demonstrating the plasticity of human growth (Modan-Moses, 2003; Prabhakaran et al, 2008).

Root and Powers (1983) conducted a qualitative study looking at three patients who began AN between 9-12 years of age and were long time sufferers of AN (between 3-8 years). They found a large decrease in linear growth and particularly in the female subject they found advanced bone maturation without the expected linear height spurt. Nussbaum et al (1985) found in a larger study of girls with AN that 76% were below the 50^th percentile for height and most were significantly shorter than parent mid height. However they did not find a difference between those that developed AN before or after menarche suggesting it is not the reduced caloric intake that affects height. If it was, those patients pre-menarche who still had a relatively larger amount of growth left should have shown greater deficits in height. They suggested that it is possibly something else to do with AN that is effecting height. On the other hand some studies have reported a sparing of height in AN patients. Pfeiffer et al (1986) looked at AN patients and their height percentiles. These were typically low but they found they increased over time with proper nourishment to their expected values. However there were some that did not meet their potential. These patients also tended to be relatively small adults too. Rozé et al (2007) found a sparing of height too. They compared pre-pubertal and post-pubertal patients and found no associations with greater height deficits in either group. They also found that target heights were typically reached. Lastly Prabhakaran et al (2008) conducted a much more in depth analysis and found a reduction in Insulin like growth factor-1 in AN patients which would lead us to expect a reduction in adult height. However they also showed how a reduction in caloric intake can slow down skeletal growth as well which allows for a longer period of time for growth hormones to act allowing the individual to reach peak height even if it is at a slower rate (Prabhakaran et al, 2008). They found that the body does this by delaying the growth plate closure as a result of hypogonadism, and by also having increased levels of nadir growth hormone that can act independently of the Insulin like growth factor-1 which is reduced by the metabolic stress caused by the eating disorder (Prabhakaran et al, 2008).  However this prolonged growth effect appears to only be true for those who undergo short periods of reduced caloric intake (Prabhakaran et al, 2008). Their research appears to display plasticity mechanisms in place that can cope with short periods of severe reductions of nutrient intake (like that seen in populations under nutrient stress such as drought etc.).

No studies have yet looked at the duration and severity of AN and its effects on linear growth explicitly. A few have found small associations e.g. Rozé et al (2007) found a correlation between lengths of hospitalization for AN and reduction in adult height but there was not enough data to prove significance. If we take a different view and we look at starvation patterns and its effects on height we find that short term environmental insults like reduced nutrient intake can slow growth in a developing child. But if conditions improve a catch up period where growth rate is above typical can restore the child to, or at least near, his or her regular course of growth (Eveleth and Tanner, 1990). In other circumstances growth can just be slowed by delaying the bone maturation which allows more time for it to occur so expected adult height is still reached, it’s just reached at a later age (Eveleth and Tanner, 1990). However in more chronic starvation settings permanent height stunting can occur especially if the environmental insult occurs before puberty (Eveleth and Tanner, 1990).

It is possible that AN could follow the same patterns of effects on the body during development in that acute starvation (mild AN) could have little or no effect on linear growth due to human developmental plasticity. Growth could be slowed down but occur over a longer time. Prabhakaran et al (2008) found this already. We could then speculate the effects could be different for chronic starvation (severe AN) where more prominent stunting is observed in younger patients when the boundaries of our plasticity are pushed. Theoretically we could find results more similar to that of Root and Powers (1983); however this has not been looked at in depth yet.

The level of severity of AN could explain the discrepancy in Rozé et al (2007) and Nussbaum et al (1985)’s data. This may be the reason why they did not find a difference in height between girls that developed AN before and after menarche. If they had distinguished between mild and sever cases their results may have shown more of an effect in those chronic, pre-menarche onset patients. It is also possible that this could explain some of the patients results in Pfeiffer et al (1986)’s study. Those that did not increase their height percentile and were relatively small at their adult measurement may have been patients with a more chronic form of AN. This is where I will focus my research. I propose to look at the effects of acute vs. chronic Anorexia Nervosa on adult heights of girls that develop the disease before and after menarche. I hypothesize that acute AN will not show much, if any, effect on adult height in the pre or post menarche group, but chronic AN will cause stunting of height particularly in the pre menarche group and possibly in the post menarche group.

Participants and methods

Participants

I will conduct a cross sectional survey study of women with Anorexia Nervosa currently living in New Zealand. I aim to get around 100 participants. Participants chosen will be New Zealand women from any ethnic group that have lived in New Zealand most of their lives. They will be from moderate to high socioeconomic status families and be between 20-30yrs old. They must have been clinically diagnosed with Anorexia at some age before 20. I will use CREDS (central region eating disorder services) in Johnsonville, Wellington to contact eligible women via an email invite to participate. Those not clinically diagnosed or those who had other diseases at the same time will be disqualified along with those with bulimia nervosa (subjects with mild reports of purging episodes may still be included). Participants will also be unrelated to each other; only one subject from each family may participate to maintain independence of the data.

Method

Participation will involve approximately a half hour visit by me at the participant’s convenience (but where possible a morning visit to control for height shrinkage over the day will be encouraged). During the visit participant’s height will be measured and they will be required to fill in a survey of their disease history which can be complete at the time or emailed to me at a later date. Medical records of all participants will be requested for a more detailed and accurate account of the disease history but disclosure is at the participants consent and refusal is not discouraged. Participants will also be required to give parent contacts so that parents can be met with too for parental height measurements, for working out parent mid height.

The survey will ask the following: Participants current age, Age at onset (to the nearest month), menarche status at onset (pre/post), span of the illness (approximately how many years), regularity of the illness (i.e. were there recovery periods and relapses), hospitalizations (were there any, if so how many?), and any accounts of amenorrhea.

Data analysis

Participants will be firstly categorized into two groups, those pre-menarche and those post-menarche. For the first analysis these data sets will be treated separately. Within each group four categories of AN severity will be distinguished (mild, moderate, severe, and chronic) based on their AN history records. Along with this, height relative to the expected parental mid height will be attained (i.e. those that reach expected height will be zero, those below the expected height will receive a negative value equal to how much shorter they are then the parent mid height, and those above expected height will have a positive number equal to how much taller they are then the parent mid height). The four severity categories and the difference in adult height from the expected value will be compared in an ANOVA test (P<.05) to see if there are differences between the mean heights of the women over the various severities of the disease. For the second part I will do four two sample t-tests (P<.05) and compare each condition in the pre-menarche group with the post-menarche group. For example I will compare mild AN mean adult height pre and post menarche, moderate AN mean adult height pre and post menarche etc. I will use this to see if there is a difference in absolute growth for women affected earlier or later in development. If there is no difference between pre and post menarche this could indicate that AN is not effecting the typical developmental growth curves of people, it is effecting (or being effected by) something else that is causing this correlation between presence of AN and reduced height.

Resources

Ellison P.. 2003. Energetics and Reproductive Effort. American Journal of human Biology. 15:342–351. 

Eveleth P., Tanner J.. 1990. World wide variation in human growth (2^nd ed.). Cambridge university press, New York. 

Kouba S., Hallstrom T., Lindholm C., Linde´n Hirschberg A.. 2005. Pregnancy and Neonatal Outcomes in Women with Eating Disorders. The American College of Obstetricians and Gynecologists. 105 (2): 255-260. 

Lester R.. 2007. Critical therapeutics: Cultural politics and clinical reality in two eating disorder treatment centers. Medical Anthropology Quarterly. 21 (4): 369-387. 

Modan-Moses D., Yaroslavsky A., Novikov I., Segev S., Toledano A., Miterany E., Stein D..2003. Stunting of Growth as a Major Feature of Anorexia Nervosa in Male Adolescents Pediatrics. 111:270-276. 

Nussbaum M., Baird D., Sonnenblicki M., cowan K., Shenker R.. 1985. Short Stature in Anorexia Nervosa Patients. Journal of Adolescent Health Care. 6:453-455. 

Pfeiffer R., Lucas A., Ilstrup D.. 1986. Effects of Anorexia Nervosa on linear growth. Clinical Pediatrics. 25(1): 7-12. 

Prabhakaran R., Misra M., Miller K., Kruczek K., Sundaralingam S., Herzog D., Katzman D., Klibanski A.. 2008. Determinants of Height in Adolescent Girls With Anorexia Nervosa. Pediatrics. 121(6): 1517-1523.

Root A., Powers P.. 1983. Anorexia Nervosa in Adolescents Presenting as Growth Retardation. Journal of Adolescent Health Care. 4:25-30. 

Rozé C., Doyen C., Le Heuzey M., Armoogum P., Mouren M., Léger J.. 2007. Predictors of late menarche and adult height in children with anorexia nervosa. Clinical Endocrinology. 67: 462–467.