CHAPTER TWO: REVIEW OF THE LITERATURE

Outline of Literature Review

This thesis explores a vital area yet emergent = area of inquiry. While much has be= en done to explore social determinants of health, the body of literature on transgender health is considerably = less developed. Many have decried = the lack of research in this area, especially in light of the disproportionate burden of morbidity and mortality demonstrated in early studies. In 2002, Boehmer attempted to quantify the disproportionate lack of LGBT research by reviewing all Medline articles dealing with these topics. In the past 20 years, only 0.1% of Medline articles dealt with issues related to lesbian, gay, bisexual, or transgender health. Of this set, articles addressing transgender health (including word variants such as transexual, intersex, or transvestite) were the most underrepresented. Only 346 of the 3777 LGBT articles= , or 0.0000% of all Medline articles addressed transgender health³². Three factors contribute to this dearth. First, the phenomena = of being transgender is low-incidence, resulting in a relatively small sample = population. Second, researchers have difficulty accessing this hidden population. Third, bias has directly affected the ability of researchers to get funding for this research area, and likely impacts personal choices about whether to study it as well. = In light of this acknowledged dearth, extra effort has been taken to augment published literature with locally published community needs assessments in = this review.

The review begins with a very brief overview of social determinants of health to provide a necessary foundation for the wor= k. This review will be based around t= he framework for social determinants proposed by the World Health Organization (WHO). Following this, the literature on transgender health outcomes will be presented. After these two fundamental areas = are covered, the literature on social determinants of transgender health will be presented, again organized as per the WHO framework.

Social Determinants of Health

The emergence of the importance of social determinants of health is intertwined with two major epidemiological shifts= in health during the 1900s. In t= he year 1900, eleven major infectious conditions accounted for 40% of all population mortality³³. Infectious viruses were the single largest cause of death, and as a result, a major health threat on both a pu= blic and a personal level. Epidemi= cs that affected full communities were all too common. In comparison, the three major chr= onic conditions: heart disease, cancer, and stroke; were responsible for less th= an 20% of the full mortality burden. The early 1900s marked an enormous shift in public health. During this era, many of the virus= es that had routinely taken thousands of lives in each new generation lost the= ir ability to sweep through populations with such vigor. Scarlet Fever, Typhoid, Measles, Whooping Cough, Diphtheria and others slowly faded from being household nam= es and fears. This precipitous d= rop in viral outbreaks was roughly coincident with the development of effective vaccinations or treatments for the diseases. This fact fueled theories that med= ical advancements were spurring declines in mortality. But as early as the 1950s, dissent= ers started to challenge that view³⁴. By the 1970s, evidence had accumul= ated and that widespread view started to change. As McKinlay and McKinlay aptly poi= nt out, the history of health improvements in the 1900s is relatively independ= ent of medical interventions³³. McKinlay refers to McKeown’s theories that nutrition, rising standard of living, and better hygiene may = all be key factors in this change^35-37. McKeown’s factors each close= ly intertwine with social class, opportunities and environmental infrastructure. In an epidemi= ological shift that did much to transform the face of health on a personal and public level, the original bio-chemical explanation was proven incorrect, and a social-environmental explanation replaced it as the largest causal factor.<= span style=3D'mso-spacerun:yes'> In the years following, a substant= ial body of work on social determinants of health developed, and has served as = the basis for and foreshadowed our current understanding of the balance between biological and social explanations for health outcomes.

Cassel ³⁸uses the insights of R. = Dubos³⁹ to demonstrate a second fundamental shift in disease epidemiology, this time in the later 20^th= century. While epidemiologica= l work early in the century had necessarily centered on destructive viruses, many = of the diseases faced later in the century were not the result of a contagious= and highly destructive agent, instead “the [most common] microbial diseases... [arose] from the activities of microorganisms that are ubiquito= us in the environment”³⁸. Thus, instead of the epidemics of = years before, it was now most often endemic factors that defined the landscape of disease. The magnitude of this change was demonstrated by the shift in mortality demographics by the 1970s. The eleven viral condi= tions which had reigned as the single largest health threat in 1900 were now attributed to less than 10% of all mortality. This represented a precipitous dro= p from the 40% of all deaths they caused at the beginning of the century. Further, the three major chronic conditions, which had previously accounted for less than 20% of all mortali= ty, now accounted for well over 50% of all mortality³³. This shift in mortality call= ed for new paradigms to understand the question, “then why this person and n= ot that one”? If all peopl= e are exposed to disease causing agents, what differentiates those who get disease and those that do not?

From this question arose the theory of gener= alized effect of the social environment on host resistance³⁸. Cassel elegantly builds the case that social support systems and other social fact= ors, such as integration, identification and expectation, all combine to create = an effect on a person that either renders them more or less susceptible to the= ir environmental disease causing agents. If the experiences are negative, such as isolation, unmet expectatio= ns, or lack of identification, then the person is more susceptible to illness a= nd will become sick more often. = If the experiences are positive, such as greater level of social support, then the person is more resistant to endemic disease causing agents, and therefore g= ets sick less often.

Cassel&#= 8217;s theory is directly evocative of a seminal work in the field of sociology, Durkheim’s Suicide^40<=
/sup>. “Durkheim was perhaps the fi= rst scholar to attempt to explain social phenomena not only in terms of individ= ual characteristics but in terms of population-based characteristics.”^41<=
/sup> His analysis of statistical data f= rom the late 1800s directly echoes the work of social epidemiologists today. By disaggregating suicide data, he= was able to demonstrate that a seemingly individual phenomenon, suicide, follow= ed different patterns in different communities. Further, he posited that these differences were explained by social environmental issues, such as common beliefs, customs, and religious ideology.&= nbsp; This work led to a long tradition of sociological studies that used = both social markers and health outcomes to explore an issue^41<=
/sup>. Increasingly, scientists from both= the field of sociology and epidemiology (as well as psychology) are working together^{42,=
43}. To the credit of these collaborati= ons, Medical Sociology is a burgeoning subfield of sociology. As well, Social Epidemiology is a promising subfield of epidemiology. While some detractors argue that social factors and biological facto= rs should remain more separate^44<=
/sup>, an increasing number of researchers and health policymakers have wholeheartedly adopted an expanded version of Durkheim’s original proposition, that social determinants provide key insight into health outcomes.&= nbsp;

Social determinants of health have been defined by the World Health Organization (= WHO) as “the economic, social and cultural factors that influence individu= al and population health both directly and indirectly, through their impact on psychosocial factors and biophysiological responses.^45<=
/sup>” Recognizing the importance and pot= ential large-scale impact of basing economic and health policy decisions on knowle= dge of social determinants, WHO’s Regional Office for Europe has launched a campaign to promote public awareness of social determinants = of health. A key component of th= is campaign is to disseminate the latest available information on the topic. To that end, WHO’s Social Determinants of Health: The Sol= id Facts⁴⁶, summar= izes the current scientific knowledge of social determinants in a straightforward framework. In this publication, WHO posits a framework for social determinants of health, organizing them into ten key categories.<= span style=3D'font-family:Arial;mso-bidi-font-family:"Times New Roman"'>

<= span style=3D'font-family:Arial;mso-bidi-font-family:"Times New Roman"'>

To provi= de a foundation for the specific topic of this thesis, a brief summary of the is= sues and evidence in each category will be presented here.

The social gradient

Perhaps = the most fundamental of social determinants of health, the social gradient of health refers to an axiom that has been proven repeatedly: health is direct= ly proportional to socio-economic status (SES). As ones socio-economic status climbs, ones health can be expected to improve as well. Further, happiness and = life expectancy are also bundled in this gradient and will improve concomitant w= ith health^47<=
/sup>. While the magnitude of the effect varies, this relationship holds true across societies and over time periods= ^48<=
/sup>. It is named a gradient because the= re is no threshold level of effect of SES on health as might be expected. The impact of SES on health starts= at the lowest measurable level, and increases gradually past the point where a person is able to reasonably afford and access all needed prevention or care interventions^47<=
/sup>. The differences are profound, “People further down the social ladder run at least twice the risk of serious illness and premature death as those near the top.” ADDIN EN.CITE <EndNote><Cite><Author>WHO Regional Office for Europe.<= ;/Author><Year>2003</Year><RecNum>2</RecNum><= record><rec-number>2</rec-number><ref-type name=3D"Electronic Source">12</ref-type><contributors><authors><a= uthor>WHO Regional Office for Europe.,</author></authors></contributors><titles>&= lt;title>Social determinants of health: the solid facts</title></titles><volume>2005</volume><numb= er>February 6th</number><edition>2nd</edition><dates><year&g= t;2003</year></dates><pub-location>Denmark</pub-locati= on><publisher>World Health Organization</publisher><isbn>92 890 1371 0</isbn>= <urls><related-urls><url>http://www.euro.who.int/document= /e81384.pdf</url></related-urls><pdf-urls><url>http= ://www.euro.who.int/document/e81384.pdf</url></pdf-urls></ur= ls><research-notes>The field of the social determinants of health is perhaps the most complex and challenging of all. It is concerned with key aspects of people’s livi= ng and working circumstances and with their lifestyles. It is concerned with t= he health implications of economic and social policies, as well as with the benefits that investing in health policies can bring. Agis D. TsourosH= ead, Centre for Urban HealthWHO Regional Office for Europe</research-notes></record></Cite></EndNote>^46<=
/sup>

The prof= ound impact of SES on health has led some to consider SES a key risk factor for disease^47<=
/sup>. But the social gradient is a macro= -level explanation of effect, encompassing many others. According to Marmot, “The so= cial gradient in health is influenced by such factors as social position; relati= ve versus absolute deprivation; and control and social participation.” Most of these factors, along with = other components of SES, interact with other social determinants listed here.

Stress

In contr= ast to SES, the bio-physical effects of stress can be easily measured. Under stress the human body has an immediate neuroendocrine response, releasing both adrenalin and cortisol, w= hich in turn trigger a host of physical responses. An evolutionary adaptation, this reaction assists in dealing with short-term, possibly lethal threats. While short-term, possibly-l= ethal threats are less common as humans diverge from their hunter-gatherer ancest= ry, short term stress such as might be experienced under a work deadline, follo= wed by reward, has been shown to be generally health promoting^49<=
/sup>. But evidence shows that frequent or long-term stress may be mal-adaptive, essentially creating “wear and = tear on the organism”^50<=
/sup>, and leading to a broad rang= e of health problems. Research on = stress related health problems demonstrates adverse outcome in such disparate area= s a mental health to immune functioning to cardiovascular disease.^{46,=
51,
52}

Mal-adap= tive, long-term stress can be a result of many factors, among them: inferior soci= al position, low sense of control, social isolation, low self-esteem, insecuri= ty, and anxiety. Further, the eff= ects of stress have been shown to accumulate over the life-span. With repeated stress, a person’s physiological response changes, creating a new lev= el of stress-reactivity. In 1998, McEwan coined the term “allostatic load” to describe the physic= al damage caused by accumulated stress.^53<=
/sup>

Early life

Regardless of culture or local health systems, = health in later life is strongly influenced by early life experiences⁵⁰. Health stressors in pregnancy can = have a direct and long-term effect on the child.&= nbsp; Likewise, early life experiences impacting growth and social develop= ment affect the biological development of a child, ultimately impacting later he= alth outcomes. Lastly, many health habits are set early in life, affecting a person’s predisposition to risky healthy behaviors such as smoking or poor diet.^{46, 54}<= !--[if supportFields]> Education affects this rubric as w= ell, as youth who achieve greater levels of education are prone to better health behaviors, higher SES, and concomitant access to health care. As summed up by Wadsworth, “A principal process in the association of childhood social factors with adult illness is the accumulation of vulnerability during childhood and adult life.”⁵⁰ (p. 50).

Social exclusion

The category of social exclusion includes pover= ty, social exclusion, and discrimination. While this demonstrates the intertwined nature of the subcategories,= it also makes the macro level category of social exclusion broad and unwieldy.= Social gradient is a phenomenon that affects all classes. In compa= rison, poverty as a specific health risk is listed separately here, because of the stark deleterious impact it has on health.= Epidemiological studies have shown the close relationship between poverty, addictions, and mental health issues⁴⁶. Evidence is accumulating to sugges= t that material conditions are the foundation for this progressive slide into ill = health. Lack of access to adequate housing, nutrition, respite, and medical care, employment, and opportunities to participate in society compromises ones physical health, mental health, and drives choices about healthy behaviors.&nb= sp; As with the allostatic load, the effects of poverty are also being s= hown to be cumulative, decreasing resistance over the lifespan. Increasingly researchers are broad= ening this perspective, focusing not only at poverty, but at the newer category of “social exclusion”. Social exclusion includes poverty, but adds a second key phenomenon,= the “process of marginalization” (p. 222) ⁵⁰. Exclusion can result from legal me= ans (i.e. lack of civil rights protections), from lack of access to normal soci= al goods or services, or from denied ability to participate in social production. Discrimination in= all forms creates some level of social exclusion. Embedded in this category is the b= ody of research on race and health. = This literature convincingly proposes that discrimination, while very hard to measure directly, is the reason behind a portion of the adverse health outc= omes in minority populations. Additionally, population-wide differences in health status which may= be currently attributed to identifiable barriers such as financial restriction= s, often have roots in historically discriminatory behaviors.^55-58

Work

While being unemployed creates obvious health challenges, the issues surrounding health and employment are receiving increased attention. Every job environment mixes issues of stress, control, and satisfaction routinely. These issues have been proven to h= ave a direct bearing on long-term health. In Brit= ain, the Whitehall studies followed workers for year in an effort to explore the extra morbidi= ty displayed by some workers over others.&nbs= p; As hypothesized, level of control over the job, variety, and colleag= ue support all correlated with better health.^59-63 In practice, these factors were mo= st often found in higher-paying jobs, providing a partial explanation for the phenomena of social gradient and health.

Unemployment

Unemploy= ment and conversely, job security, can have a variety of affects on health. Lowered income is an immediate out= come, and this can often bring stress. But less obvious outcomes are becoming better understood. Research increasingly shows that j= ob security is protective to mental health.&n= bsp; Likewise having a job provides “latent consequences” suc= h as time structure, self-esteem and respect of others. The loss of these is theorized as = the true threat of unemployment and its associated decrease in health. ADDIN EN.CITE <EndNote><Cite><Author>Jahoda</Author><Year>1= 979</Year><RecNum>65</RecNum><record><rec-number= >65</rec-number><ref-type name=3D'Journal Article'>17</ref-type><contributors><authors><autho= r>Jahoda, M.</author></authors></contributors><titles><tit= le>The impact of unemployment in the 1930s and the 1970s.</title><secondary-title>Bulletin British Psychosocial Sociology</secondary-title></titles><periodical><full-= title>Bulletin British Psychosocial Sociology</full-title></periodical><pages>309-14</page= s><volume>32</volume><dates><year>1979</year&= gt;</dates><urls></urls></record></Cite><C= ite><Author>Fryer</Author><Year>1987</Year><R= ecNum>64</RecNum><record><rec-number>64</rec-number= ><ref-type name=3D'Book Section'>5</ref-type><contributors><authors&= gt;<author>Fryer, D.</author></authors><secondary-authors><author>Fry= er D.,</author><author>Ullah P.,</author></secondary-authors></contributors><titles= ><title>Monmouthshire and Marienthal: sociographies of two unemployed communities. </title><secondary-title>Unemployed People</secondary-title></titles><dates><year>1987&= lt;/year></dates><pub-location>Milton Keynes</pub-location><publisher>Open University Press</publisher><urls></urls></record></Cite>= ;</EndNote>^{64,=
65} Further, the impact of unemployment begins with the threat of it, as workers with low job security showed incre= ased mental health and physical problems over those with greater job security.^66<=
/sup>

Social support

Social isolation and social integration form two ends of a continuum that strongly affects the health of individuals. The more socially isolated a person is, the more their physical and mental health suffers, and the higher their risk for early mortality. Marriage, a proxy for long-term partnership, has been consistently shown to have a protective health effect. In long-term analysis= of Alameda County study data, the least socially integrated people were twice = as likely to die as those on the high end of the scale.^67<=
/sup> Importantly, it is now understood = that the variety as well as the quantity of social support received is important. People who receive social support from different areas of their life are more protected against deleterious health than those who do not.^{68,=
69}

Addiction

Addictio= n and low SES share a symbiotic relationship.&nb= sp; While it is agreed that addictions often drive individuals SES downwards, it is also understood that people succumb to addictions as a res= ult of low SES. “People tur= n to alcohol, drugs, and tobacco to numb the pain of harsh economic and social conditions.” (p.24)^46<=
/sup>. Interaction with each of these fac= tors is the constant social pressure exerted by highly effective marketing campa= igns for alcohol and tobacco. In addition to lowered SES, addictions carry their own health burden, including increased social exclusion, morbidity and mortality. Notable among all addictions, toba= cco use accounts for over 400,000 deaths in the U.S. per year. Tobacco is distinguished for being= the single leading cause of mortality both in this country and worldwide .^70<=
/sup>

Food

From a g= lobal perspective, the relationship between food and health is most severe in man= y of the developing countries where access to subsidence nutrition is a challenge. But the interactio= n of food and health outcomes is not limited by availability. The factors of poverty, dietary ha= bits, and promotion of nutritionally unsound food products combine to ensure food remains a health issue whether a family is waiting to get bags of rice, or lives within easy walking distance of a fully-stocked grocery store. Increasingly, obesity is the primary dietary concern. Over= the last ten years, the prevalence of obesity has doubled in the U.S. ADDIN EN.CITE <EndNote><Cite><Author>Caterson</Author><Year>= ;2002</Year><RecNum>48</RecNum><record><rec-numb= er>48</rec-number><ref-type name=3D"Journal Article">17</ref-type><contributors><authors><= author>Caterson, Ian D.</author><author>Gill, Timothy P.</author></authors></contributors><titles><tit= le>Obesity: epidemiology and possible prevention</title><secondary-title>Be= st Practice & Research Clinical Endocrinology & Metabolism</secondary-title></titles><periodical><full= -title>Best Practice & Research Clinical Endocrinology & Metabolism<= /full-title></periodical><pages>595</pages><volume&= gt;16</volume><number>4</number><dates><year>= 2002</year></dates><urls><related-urls><url>h= ttp://www.sciencedirect.com/science/article/B6WBD-47C4446-2/2/2a5eac0b199bb= f633c48e1ddbf9027b6 </url></related-urls></urls></record></Cite>&= lt;/EndNote>^71<=
/sup> Other countries follow similar pat= terns leading to recent labeling of obesity as a worldwide epidemic.^72<=
/sup> This epidemic is generally conside= red to be a byproduct of easy availability of high-energy foods combined with a low demand for physical exercise.^73<=
/sup> ^74<=
/sup> The medical consequences are profound. It is estimated that obesity is the second leading cause of preventable death in the U.S. and al= most one third of cancers worldwide are diet-related.^{70,=
75}

Transport

For year= s, our cities and towns have gradually become car-centered. Options for bicycling or walking as transportation are discouraged through the structural development of our li= ving areas. This shift has adverse= ly impacted health through environmental effects, by increasing social isolati= on, and through the population-wide reduction in physical activity.^76<=
/sup> “The solid facts are that wa= lking and cycling benefit health, while motor vehicles damage health.”^50<=
/sup> Lack of physical activity has been= added to tobacco and obesity to form a triumvirate of the top preventable health concerns facing the = U.S. Meanwhile environmental health bur= dens cluster in areas of high poverty, aggravating the health of populations alr= eady lower on the social gradient.^77-=
79 While it is difficult to conceptua= lize the current car-based society changing significantly, the World Health Organization stressed the need for policy-makers to move towards more pedestrian and bicycle friendly policies to combat these health problems.^46<=
/sup>

Transgender Health

There have been documented accounts of transgen= der people existing across the broadest range of geographical, racial, ethnic, = and historical contexts⁸. While it is generally accepted that the phenomena of being transgender crosses all social boundaries, less is known about the prevalence of transgender people within the general population. The best approximates are analyses done on transsexuals presenting for care or = sex change interventions. A 1996 = study from the Netherlands used this method to estimate that 1/11,900 born males = and 1/30,400 born females were transsexual¹¹. This measures solely the extreme end of the gender identity continuum, and of th= ose, the even smaller number yet who pursue care for their gender identification issues. Actual prevalence of transgender people is hypothesized to be much higher for several reasons, including the number of people whose gender identity issues fluctuate above= and below the clinical threshold and the relative invisibility of many female-bodied transgender people who can pass as male without medical interventions.⁸⁰ = This prevalence is probably much higher in urban areas, as transgender people relocate to them in an effort to gain a greater level of anonymity or acces= s to like-minded individuals.

In the review that follows, data on risks and outcomes varies significantly by the gender vector of the people being studied. Most of the needs as= sessments cited present data on respondents by two categorical groupings: male-to-fem= ale gender vector (MTF) and female-to-male gender vector (FTM). A few needs assessments separate t= his further and also identify another group, MTF male identified people<= !--[if supportFields]>= ADDIN EN.CITE <EndNote><Cite><Author>McGowan</Author><Year>= 1999</Year><RecNum>46</RecNum><record><rec-numbe= r>46</rec-number><ref-type name=3D'Book'>6</ref-type><contributors><authors><a= uthor>McGowan, C.K.</author></authors></contributors><titles><t= itle>Transgender Needs Assessment</title></titles><dates><year>1999</ye= ar></dates><pub-location>New York City</pub-location><publisher>The HIV Prevention Planning = Unit of the New York City Department of Health</publisher><urls><= /urls><research-notes>needs assessment</research-notes></record></Cite><Cite>&l= t;Author>Transgender Health Action Coalition</Author><Year>1997</Year><RecNum>45</R= ecNum><record><rec-number>45</rec-number><ref-type name=3D'Book'>6</ref-type><contributors><authors><a= uthor>Transgender Health Action Coalition,</author></authors></contributors>= ;<titles><title>Final Report by the Transgender Health Action Coalition (THAC) to the Philadelphia Foundation Legacy Fund</title><alt-title>Needs Assessment Survey Project</alt-title></titles><dates><year>1997</y= ear></dates><pub-location>Philadelphia, PA</pub-location><publisher>The Transgender Health Action Coalition</publisher><urls></urls><research-notes>n= eeds assessment</research-notes></record></Cite></EndNote&g= t;^{81, 82}. This group would include those who only display their gender variance occasional= ly, such as crossdressers or drag queens. This a priori separation of people is upheld by the analyses, which demonstrate that each of these groups show different levels of health risks. In general, people who= were natal males and are now female-identified, MTFs, display the most health ri= sks and problems. To help clarify= the data being presented below, it is specified if the finding refers to people= on the “male to female” gender vector (MTF), or those on the “female to male” gender vector (FTM). This is not meant to be indicative= of self-identity of any people in these groups, or to unnecessarily conflate t= he possible differences in experiences by full-time versus part-time gender non-conformists. Some b= elieve it is a myth that transgender people can be separated meaningfully by these basic categories, since the range of self-identity is so varied in comparis= on.⁸³ = Indeed, like current arguments about the validity of race as a meaningful question = on surveys⁸⁴, it is likely that these basic categories are actually a rough proxy for much more complex phenomena. In particular, a more effective me= asure of transgender identity in terms of social discrimination might be a combination measure of “how much of your life do you appear to others= as gender normative” and “when you are seen as gender variant, do people perceive your birth gender as male or female”. These two measures wou= ld more successfully capture the twin issues of passing and the possible great= er societal bias against transgenders who were natal males (MTF). In the absence of more discriminat= ory measures, FTM and MTF are used here as an expediency to identify difference= s in major population subgroups. If neither MTF nor FTM is specified then all groups were included. =

Complicated relationship with the medical com= munity

Debates over pathology and the need for medical interventions combine with a history of discrimination to create a complica= ted relationship between the transgender and medical communities.

Medical diagnoses, such as Gender Identity Diso= rder, are currently necessary accompaniments to desired hormonal or medical interventions⁸⁰. Occasionally, the process of diagnosis also provides access to insur= ance reimbursement. Nonetheless, advocates have debated for years over whether being transgender should be de-pathologized in an effort to reduce current stigma^{85, 86}. Burke explores the dark side of pathologizing transgender identities in her book = Gender Shock. In it she examines the disqu= ieting case histories of youth who have been subject to behavioral and medical the= rapy to try and ‘correct’ their GID⁸⁷. In 1997, GenderPAC, a national gender advocacy group, passed a resolution embo= dying the widespread desire for change,

THEREFORE BE IT RESOLVED THAT, GenderPAC affi= rms that no one -- gay, lesbian, bisexual, transgender, intersexed, or straight= -- should have to accept being pathologized as mentally ill in order to attain wholeness, completeness and civil equality; BE IT FURTHER RESOLVED THAT, GenderPAC supports GID reform (rather than eradication), where "reform" means another diagnosis -- possibly physical -- that does not pathologize transgender people or gender-variant youth and children.= 221;88

As community members struggle over the issues of self-determination and acceptance, this debate reflects the one preceding t= he removal of ‘homosexual’ from the Diagnostic Statistical Manual = in the early 1970s⁸⁹. The difference that has likely drawn this debate out longer is the existence of medical treatments that alleviate the stress associated with being transgender. Hormones and sur= gical interventions often allow gender variant people to “pass” as th= eir desired gender in public. Not passing can bring great stress, and is highly correlated with mental health problems⁹⁰. Unlike homosexuals, transgender people have a specific need for medical interventi= ons, and in the case of hormones, ongoing medical care to supervise those interventions. This need can = have a positive outcome, as transgender health clinics have provided a platform for researchers to cull important information about these rare populations^{11, 91, 92}.

Despite this high need for medical care, resear= ch shows transgender people have lower than average ability to access care and often experience blatant and sometimes flagrant discrimination from medical providers. A transgender needs assessment from Massachusetts relays the story of one transgender per= son who was summarily discharged from the emergency room despite having a fract= ured vertebrae and a concussion⁹³. Another needs assessment tells the story of a transgender person who was told to “see a veterinarian” as a medical doctor was a “doctor for people” (p.22)⁹⁴. In 2001, the documentary Southern Comfort won critical acclaim for portraying = the last phases of the life of Robert Eads, who was refused treatment for his ovarian cancer by 20 doctors. In a well-publicized incident in Wa= shington, DC, Emergency Medical Techn= icians from the Fire Department stopped treating a female car accident victim when they realized she had male genitalia. This delay was later proven pivotal in her death. In a further discriminatory twist,= the Fire Department named the later court-mandated sensitivity training programs after the victim’s male instead of female name⁹⁵. These stories are not isolated examples, but indicative of a widespread phenomenon. In a community ne= eds assessment in Philadelphia, fully 81% of respondents reported being denied medical services because they were transgender⁸². As demonstrated, the fears of discrimination by health care providers are very real. Even if a person has not directly experienced discrimination, anecdotal information from community members and friends is sufficient to act as a deterrent to health care.

In addition to a fear of discrimination, access= to healthcare by transgender people is limited by lack of provider education.<= span style=3D'mso-spacerun:yes'> In the same Philadelphia survey, 67% of the transge= nder (non-crossdresser) population reports having to regularly educate their hea= lth care providers about transgender-related health issues. Disturbingly, this figure stays st= atic even among people who have health insurance, and therefore greater latitude= to choose their providers⁸². Other community needs assessments consistently uphold the problem of lack of education among providers^{81, 94}. This lack of education creates a personal barrier in choosing to access health care, = but it also creates a hard-to-quantify lower quality of care for transgender pe= ople who are accessing health care.

Because of prevalent discrimination and inadequ= ate provider education, avoidance of healthcare by transgender people has become widespread. Often, people res= trict interaction with the health care community to efforts to obtain the hormone= s or surgery to bring their body into better accordance with their gender identi= ty. In the transgender community, stor= ies of people who avoid medical care at all costs are legion, and the possibility = of being incapacitated and needing emergency care is a major fear for many. In studying transgender health it = is important to understand and consider the complex issues related to medical care, because they often interact negatively with many of the other social factors determining health outcomes.

Health outcomes

There are two bodies of research that inform cu= rrent knowledge about transgender health outcomes. One, the most easily accessible, i= s the literature published in the academic arena. Augmenting this information is a s= eries of community needs assessments. Some of these data are less accessible, partly because reports are s= elf-published, and no list of all needs assessments exists. To demonstrate the scope of this research, a profile of the needs assessments identified and used in this re= view is presented in the table below.

Profile of Current North Amer= ican Transgender Health Needs Assessments
City	Year	Lead Author	Mode	n	Target Population(s)	Cite
At= lanta, GA^{96, 97}	1993	El= ifson, K.W.	in= terviews	53	MTF prostitutes	2
Bo= ston, MA⁹⁴	2000	Fa= llas, G.	qu= alitative	40	Tr= ansgender	1
Bo= ston, MA^{93, 98}	1999	Kammerer<= span style=3D'font-size:9.0pt;font-family:Arial'>, C.A.	et= hnographic		Vu= lnerable TG	2
Ch= icago, IL⁹⁹	2002	Ke= nagy, G.P.	su= rvey	81	Tr= ansgender	2
Connecticut		fo= rthcoming	su= rvey		Tr= ansgender	--
Ho= uston, TX¹⁰⁰	2003	Ri= sser, J.M.	su= rvey	67	MT= F only	3
Los Angeles, CA¹⁰¹	2001	Re= back, C.	in= terviews	244	MT= F only	1
Mi= nneapolis, MN¹⁰²	1998	Bo= ckting, W.O.	fo= cus groups	19	Tr= ansgender	2
Ne= w York City, NY⁸¹	1999	Mc= Gowan, K.	mi= xed	94	Tr= ansgender	1
Ph= iladelphia, PA 1⁸²	1997	TH= AC	su= rvey	109	Tr= ansgender	1
Ph= iladelphia, PA 2¹⁰³	1997	Ac= tionAIDS	su= rvey	70	Vu= lnerable TG	1
Ph= iladelphia, PA 3⁸³	2003	Sa= usa, L.	in= terviews	24	Tr= ansgender youth	5
Pu= erto Rico^104-106	2000	Ro= driquez-Madera	su= rvey	50	MT= F only	2
Qu= ebec, ON¹⁰⁷	1999	Na= maste, V.K.	qu= alitative	5	FT= M only	2
San Francisco, CA 1^{108, 109}	1999	Cl= ements-Nolle, K.	fo= cus groups	100	Tr= ansgender	2
San Francisco, CA 2¹¹⁰	2001	Cl= ements-Nolle, K.	su= rvey	515	Tr= ansgender	2
San Francisco, CA 3¹¹¹	2001	Ro= se, V.	mi= xed	71	MT= F Afr. Amer.	4
San Francisco, CA 4^{112, 113}	2004	Ne= moto, T.	in= terviews	332	MT= F People of Color	2
Va= ncouver, BC¹¹⁴	2003	Go= ldberg, J.	su= rvey	187	Tr= ansgender	1
Vi= rginia¹¹⁵	2004	He= ndricks, M.L.	fo= cus groups	46	Tr= ansgender	3
Wa= shington, D.C.^{2, 116}	2000	Xa= vier, J.M.	su= rvey	252	Tr= ansgender	2
&n= bsp;	&n= bsp;	&n= bsp;	Total n	2359	&n= bsp;	&n= bsp;
&n= bsp;	&n= bsp;	&n= bsp;	&n= bsp;	&n= bsp;	&n= bsp;	&n= bsp;
Related Research
HI= V+ health experience¹¹⁷	1998	Sc= hilder, A.	fo= cus group	10	HI= V+ MTF TG	2
Na= tional Violence Study¹¹⁸	2001	Lo= mbardi, E.L.	su= rvey	402	Tr= ansgender	2
Pa= tient satisfaction¹¹⁹	2004	Bo= ckting, W.O.	su= rveys	1017	TG= & non-TG	2
Pr= ovider info. needs¹²⁰	2004	Fi= kar, C.R.	su= rvey	152	TG= health providers	2
Pr= ovider training needs¹²¹	2005	Lu= rie, S.	in= terviews	13	TG= health providers	3
&n= bsp;	&n= bsp;	&n= bsp;	&n= bsp;	&n= bsp;	&n= bsp;	&n= bsp;

Guid= e to type of cite &= nbsp; : 1 – report only. 2 – published article.= 3 – forthcoming. 4 – conference paper.<= span style=3D'mso-spacerun:yes'> 5 – dissertation.=

As can be seen, twenty community health needs assessments exist, spanning fourteen regions or metro areas and incorporati= ng over 2,300 respondents. Many = of the studies were funded by HIV prevention dollars and all are non-probability samples. Both of those factors could arguably lessen their import. Sampling strategies might bias towards finding people who are at hig= her risk for HIV, and therefore at higher risk for concomitant health issues. Similarly, non-probability samplin= g is subject to many biases in design and outreach, any of which can impact the findings. Still, the findings should not be noted for a variety of reasons. First, many of the needs assessmen= ts used the best available sampling science considering the population characterist= ics.¹²² = That is, the sampling techniques were designed to capture a wide diversity of people= , by gaining access through existing community leaders or organizations, then us= ing a snowball or social networks approach to getting sufficient inreach to the communities. In total, the le= vel of inreach was excellent, as findings from over 2,000 transgender people are presented in these reports. S= everal of the needs assessments collected data in different modes to enhance valid= ity, most often focus groups supplemented with interviews. Importantly, the body of research = shows internal validity, as similar patterns are echoed across each individual study. A recent publication i= n the area of smoking prevalence among homosexual populations upheld the findings from a similar group of community-based surveys, showing how their internal validity correlated to later population-based findings¹²³. All of these factors increase the credibility of these needs assessments. The time has been taken to examine= this credibility first because the findings in this body of research can be troubling. The review that fo= llows will draw from both these needs assessments and the academic literature.

Sexually transmitted diseases

HIV funding has been provided for many of the a= bove needs assessments because of the stark reports of high HIV prevalence in ea= rly needs assessments. In many ci= ties, one of the most visible elements of the transgender communities is a subpopulation of MTF sex workers. There are often separate “strolls” or streets for the transgender versus natal-female sex workers. Likewise, there are sometimes bars= where transgender sex workers congregate and meet potential customers.^{98, 124} = The existence of this sub-community brings up intriguing questions, both about = the “johns” that support this work, and the factors that have influenced the development of this sub-community. Some available work in this area i= s deliberately not referenced here due to debates over its scientific merit. ¹²⁵^126-128 = In the early 1990s, literature from other countries began to report levels of HIV sero-prevalence that ranged from 11% to 86% among these transgender sex workers.^{83, 129-133} = Elifson and Boles’ 1993 study first explored this population in the U.S. In that Atlanta-based study, 68% o= f the transgender MTF sex workers were HIV+⁹⁷. Further, a social organization analysis of the same data showed HIV prevalence of 81% in one transgender community, the one which was “mo= st committed to transvestitism, and consequently… socially isolated̶= 1;.⁹⁶ A previous study by the same authors lent perspective to these findings, as t= hey found only 27% of non-transgender male prostitutes were HIV+.¹³⁴ = Later full-community studies routinely reported HIV prevalence rates over 20%, wi= th 25% in Washington D.C.², 21% in New York City⁸¹, and 35% in San Francisco¹¹⁰. In repeated studies, transgender people tell of low-SES and low-self-esteem driving them into this sub-population of sex workers, and t= he high risk sex that ensues. The following excerpts from previous studies demonstrate this issue.=

“A lot of people in the trans community= have to make ends meet.... Just like with prostitution, there is a high rate of = HIV with trans prostitutes who are trying to make it that way. And they’ll get men who tell= them, ‘I’ll give you and extra $10 or $20 if you don’t use a condom.’ And they’= ;ll go for it” (p.19)^{81. “Shake your tits, shake your=
tush,
and have some guy slip you a twenty, fifty, hundred dollars. It builds up your self-esteem!R=
21;
(p.7)¹⁰⁹}

Overall, the issue of sex work among MTFs is th= ematic in the needs assessments. It = is difficult to gauge what percent of the MTF community engages in sex work, but problems and issues of the sex workers clearly shape the experiences of the larger community. <= /span>

In the instances where FTM transgenders are ana= lyzed as a separate group, they do not show an unusually high prevalence rate for= HIV^{81, 108}= , but a San Francisco qualitative study documents increased risk-taking in this group. According to these respondents, th= is higher risk taking is linked to a desire for acceptance, hormonal changes in temperament, or increased disassociation with one’s body.¹⁰⁹

Transgender studies show high levels of sexually transmitted diseases in addition to HIV^{81, 98}. In the Atlanta study of MTF sex workers, 79% of participants tested positive for syphilis and 76% for Hepatitis B.⁹⁷ = Sharing hormone needles is a contributing factor to the spread of sexually transmit= ted diseases. <= /p>

Silicone use

For many transgender people, achieving a body image congruent with their internal im= age can have two beneficial effects: it can ease personal distress or gender dysphoria, and it can alleviate violence or discrimination stemming from publicly perceived gender incongruence. Hormones and gender confirma= tion surgeries are often used to achieve this end. Another option is injecting silicone. Injecting silicone = can quickly produce body contours indicative of the desired gender presentation, both for MTFs and FTMs (who can use it to achieve “muscle bulk”= in desired areas). Silicone can = be a faster and cheaper option than either hormones or surgery and its use in the transgender communities is widely documented.^{81, 129, 135-138} As it is illegal, silicone injection is never medically supervised, instead it is a ‘back alley’ procedure. The difficulty in obtaining medical grade silicone and lack of medical supervision has led to wide variations in the quality of material being injected.&nb= sp; According to Bay, the most common form of injected silicone is a non-medical grade purchased from hardware stores or automotive supply house= s.¹³⁵ Additional reports show that materials as diverse as mineral oil, linseed oil, flax oil, and automotive transmission fluid have all been inje= cted to simulate the effects of silicone.^{139, 140}

The effects of long-term silicone exposure have= fired a debate in the arena of breast implants, with studies pointing to increase= s in cancer, joint-related, and neurological disorders in affected women.= ¹⁴¹ = This debate led to the 1992 banning of silicone breast implants in the U.S. by the Food and Drug Administration. A later Institute<= /st1:PlaceType> of Medicine report = found much of this research did not stand. While the scientific debate over the health effects silicone implants continues, the IOM has provided a warning about the complications that aris= e, and the resultant need for further medical care to remedy complications.¹⁴¹ = Silicone use in the transgender communities carries a profile of different risks abo= ve those experienced by women with silicone breast implants. First, the silicone is free-inject= ed into the body, not encased in a pouch as with a breast implant. This increases the direct exposure= to the compound. Second, as demonstrated above, medical-grade silicone is often not available, and noth= ing is known about the risks of the available substitutes. Third, the quantities at issue are vastly different, in a 2001 study of transgender MTFs, the average injection quantity was 4.5 liters.¹³⁹ = Fourth, the illegal practitioners are not required to adhere to the standards of sterility used in an operating procedure.&= nbsp; Fourth, free injected silicone (or its substitutes) is not bound to = any one location, shifting is commonly reported, and no remedy is fully effecti= ve in treating this problem. Fifth, silicone injection is painful and concomitant use of drugs to alleviate that pain is often reported, thus introducing another risk.^{83, 135} = In Hage et al.’s 2001 review of the issues faced by 15 transsexuals presenting for treatment, they listed dangers ranging from change in skin color to sev= ere respiratory failure. They also warned of the long latency period before complications, up to 24 years. In sum, they were unequivocal, cal= ling the effects “devastating” and ending with “In view of the potential dangers, feminization by injection therapy should be soundly condemned.”¹³⁹

Mental health issues

A recent article by Nuttbrock, Rosenblum and Blumenstein reviews the stages of maturation of research on the mental heal= th of transgender populations.⁹⁰ Initial reports portrayed transgender people as “miserable souls”.¹⁴² = Later research started to show transgender populations as emotionally stable and = not displaying higher rates of mental illness than the general population. Contrasting with this are consistent needs assessment findings showing alarmingly high rates of depression, anxiety, and suicide ideation.= In the largest San Francisco needs assessment, 21% of the sample had been hospitalized for a mental health condition, and 32% reported a past suicide attempt.¹⁰⁸ = These numbers were stable for both gender vectors, FTM and MTF. In the Washington D.C. needs asses= sment, 35% of respondents reported suicide ideation and 16% reported past suicide attempts.² A Philadelphia needs assessment found 39% of respondents had attempted suicid= e.⁸² = In a later San Francisco study of MTF of color 40% of the respondents reported b= eing depressed, and 29% had attempted suicide.¹¹² = In comparison, even the most generous estimate of the rate of suicide attempts= in the general population is 00.3%.^{143, 144} =

These blatantly high rates of past suicide attempts prompts the re-examination of= the sample population. Presuming = some transgender suicide attempts are successful, it follows that the available population of transgender people who have responded to these surveys are a truncated representation of the whole.&nbs= p; It is necessarily missing a group of people, those that have succeed= ed in committing suicide. To fol= low this theory further, the populations surveyed here are likely to be mentally healthier or more resilient than the missing group, in effect biasing the a= lready stark findings on mental health problems downwards. Further it should be noted that the suicide attempt rates are equally high for the FTM population, the group th= at otherwise experiences less adverse health outcomes.

Practiti= oners debate the cause of these mental health problems. Some argue that that childhood tra= uma can be the cause of both transgenderism and the later mental health problems. Others posit the me= ntal health problems are a result of body image dissonance. Another body of practitioners posi= t that mental health problems in the transgender population emerge from “liv= ing in a hostile environment”.^{90, 145} In the Nuttbrock study, the level of mental health problems was dire= ctly associated with broad acceptance of the desired gender identity, a construct the authors named transgender ident= ity affirmation.⁹⁰ While needs assessments sometimes report the suicide attempts as bei= ng related to being transgender, this vague category can cover internal gender dysphoria distress and the stress caused by lack of societal acceptance.^{2, 82} A Boston needs assessment offers this explanation “the social stigma that TG face from early in life is translated via internalization and fear into psychological problems, notably low self-esteem and even loathing, often to the point of suicidal tendencies” (p.60)⁹³.

The ment= al health issues of transgender people are likely related to a host of social characteristics, including stress, discrimination, and isolation. In any case, the extremely high re= ports of mental health problems and suicidality by community needs assessments ma= y be the clearest beacon to the role social determinants are playing in health outcomes. <= /p>

Additional health outcomes

Prolonged hormone use may put transgender people at increased risk for a number of adverse health outcomes, including: cardiovascular disease, diabetes, thromboembolic events, and liver abnormalities. Bockting’= ;s 2003 article reviewing transgender health literature summarizes the current information on risks. He note= s that the information in these areas is “not well characterized” (p. 7). This statement that refle= cts the limited research on transgender populations, reliance on related non-transgender hormone studies, and the sometimes conflicting outcome data from available studies.⁹¹

Both femininizing and masculinizing hormones can lead to increase risk for heart disease. Masculinizing hormon= es increase a persons’ HDL cholesterol and decrease their LDL cholestero= l, putting the population of FTMs at increased risk. Feminizing hormones can do the opp= osite, but ancillary effects show no health benefit from this change, and increased heart disease has been noted for women on hormonally similar oral contraceptives. The small population-specific information is contradictory. While there are case reports suppo= rting this increased risk from each population, a retrospective study shows no overall increased morbidity.^{91, 146}

Feldman&= #8217;s 2003 article reviews information showing that a potential increase in estro= gen use can lead to an increased risk of diabetes, with case reports from the M= TF population supporting this hypothesis.¹⁴⁷ In a broader study, Asscheman reports a 45-fold increase in thromboembolic events among MTF transexuals.¹⁴⁸ Information on osteoporosis is controversial, since both estrogen and testosterone maintain bone density, but there is a likelihood of higher lev= els of osteoporosis in some transgenders who are not fully adherent to hormone therapy, or have discontinued it and have lower levels of internal hormone production due to gender confirmation surgery. In light of this limited informati= on, Feldman recommends weight bearing exercise for all transgender people on hormone therapy.⁹¹

In Feldman’s review, he notes that cancer risks for transgender people a= re similarly “not well understood” (p. 9).⁹¹ Hormone therapy and surgery both change a persons risk profile for cancer, but the effects of this are largely unknown, and there is a dearth = of screening guidelines. Chest s= urgery for FTMs often leaves some breast tissue that still constitutes a risk for breast cancer. Lack of educat= ion may be a problem as people can falsely perceive themselves to be free from = this risk. While MTF transgenders = seem to be at low risk for breast cancer, Feldman recommends routine female screening guidelines be followed as a preventative measure. Masculinizing hormones can atrophy= the cervix, causing abnormal Pap smears, or cause uterine bleeding. Pre-operative FTM transgender peop= le may also be at increased risk for ovarian cancer. And while large studies are not available, cancer has been reporting in the both the prostrate and neo-vagi= na of MTF transgender people.⁹¹

Social determinant information

The information on medical outcomes above is se= verely limited by the available research in this area, in all it shows only the beginning of a full picture of the health status of transgender people. The following information about so= cial determinants of health starts to broaden the full scope of this picture, because it provides the foundation for many other health problems to ensue. These outcomes would p= erhaps have already been documented and presented above if more research had been available. In the old adage a= bout the difference between public health and medicine it is said that the doctor fishes people out of the river and treats them, while the public health practitioner goes upstream to see what is causing the people to jump in the river. In looking at social determinants we head upstream in our pursuit of health causation, and here = we find ample problems indeed. <= o:p>

Social gradient

The needs assessments as a body show transgender people clustering on the lower ends of the SES spectrum, with an unusually = high number living in outright poverty. &n= bsp; The 1997 Philadelphia TG needs assessment found the average income of all respondents to be between $15,000 and $19,999.⁸² = The Washington D.C. needs assessment found 31% of respondents reported incomes = of less than $10,000 per year (29% had no income). On that survey, only 61% of the respondents held a full time job, which were overwhelmingly clustered in the lower paying service industry. In accordance with the level of underemployment, 47% of this group had no heal= th insurance at all.² = In New York City the= average income was $2,700 per month, but that varied greatly by group, with part-ti= me MTFs reporting over double the income of FTMs.⁸¹ There is no theory that would substantiate more transgender people being born into low SES versus other classes, so while t= he surveys only present a static snapshot of SES, they provide clear informati= on that factors related to being transgender conspire to move people lower on = the SES scale. =

Stress

In Philadelphia, 30% of MTF respondents reported they expected to have a shorter than average lifespan. Their reasons were = the “stress and pressure of being transgender,” diseases, or fear of being killed (p.18).¹⁰³ Poverty, discrimination, and = other categories of social determinants of health cause a high level of stress for many people in this group, and many of these effects are addressed under ot= her subheadings. My discuss= ion of stress here will focus on the stress engendered by the exceptionally high l= evel of violence transgender populations face.&= nbsp;

In 1997, one of the leading gender advocacy organizations, GenderPAC, conducted a survey about transgender community members’ experiences with violence.&= nbsp; They attempted to get respondents from every subsection of the communities, in all collecting data from 402 people in a span of 12 months. Respondents were bala= nced in the FTM and MTF gender vectors but the authors caution that Latinos and African Americans were underrepresented.&n= bsp; In all, the findings are sobering, 47% of the sample had been assaul= ted and 60% had experienced violence or harassment (including verbal abuse) wit= hin their lifetimes. Approximatel= y 21% of both MTF and FTM transexuals reported experiencing rape or attempted rape. The study makes the distinction that “visibly transgender” people are more likely to experience violence, and found that transgender youth have a greater likeli= hood of violence than older people. “Transgender youth who disclose their status are scorned, attacked, and locked into or thrown out of their homes.” (p.98).¹¹⁸ =

Indicative of the level of violence endemic in = the community, the only “celebration” specific to the transgender communities is National Transgender Day of Remembrance, a time to gather and honor those killed by gender violence in the year past. Between the years of 2002-2004, Da= y of Remembrance participants have collected information on an average of 13 transgender murders per year.¹⁴⁹ = Many of these murders were marked by a phenomenon common to hate crimes, that of a level of aggression exceeding what is needed to achieve death, commonly referred to as “overkill”.&nbs= p; In the Winter 2003 edition of the Southern Poverty Law Centers Intelligence Report, they note that the number of transgender people murder= ed in hate crimes exceeds the number of non-trans people killed in all other h= ate crimes combined.¹⁵⁰ = The possibility that anti-trans murders outweigh all other hate crimes is particularly grim considering the relative size of the transgender populati= on in comparison to the combined population of all minority groups. This level of violence is echoed b= y the needs assessments, in one of the Philadelphia needs assessments, 83% of MTF and 48% of FTM respondents experienced direct violence, including physical abuse and forced sex. As well, 77% of MTF respondents in= that survey were either “unsafe” or “uncomfortable” in public due to being transgender.¹⁰³ In the Houston needs assessment, 25% of the MTF sample had been forced into sex by their primary partner, and 48% had been = hit by a primary partner.¹⁰⁰ = In San Francisco, 59% of both FTM and MTF reported a history of physical abuse or forced sex.¹⁰⁹

Violence in the transgender community is exacer= bated by insensitive or abusive responses by the public servants entrusted to res= pond to such emergencies. Stories = of police or emergency personnel aggravating the situation are rife. One FTM Philadelphia relays his experience, “I got shot. When the EMT’s got there they just loo= ked at me and said, “I’m not going to touch it.” (p.12)¹⁰³ = In some cases, violence may not be related to anti-transgender sentiments but erupt= s as a natural outgrowth of the socially unstable communities into which many transgender people are pushed. In the words of Lisa Mottet from the National Gay and Lesbian Task Force’= ;s Transgender Civil Rights Project, "Look at the victims. Because they are transgendered, th= ey have to be in places that are extremely dangerous to begin with. Even if they're assaulted or killed for reasons other than hate, they still wouldn'= t have been targeted if they weren't transgendered, because they'd be able to stay= in school, have family support, and hold down jobs.”¹⁵⁰ In all, each of these factors combines to crea= te a culture of lived violence within the transgender communities, “When a transgendered person is victimized, it doesn't just affect her friends and family — it terrifies a whole community of people who can't help feel= ing they might be next.”¹⁵⁰ This sentiment is stated most succinctly by on= e New York City needs assessment respondent, “We are lucky to be alive day-to-day without people killing us.” (p.28)⁸¹

Early life

Many of the social determinants of early life a= re characterized as factors influencing birth and early developmental years, a time in which many transgender people would not yet have been identified. But while the earliest of influenc= es may not be affected by one’s transgender status, once a gender variant identity emerges early life can become adversarial for impacted youth. Transgender youth report a high le= vel of school- and home-based harassment, and are at increased risk for depression, addiction, and suicide.^{83, 151-154} High levels of harassment have the additional outcome of pushing many transgender youth into homelessness, either because they are forced out by their families or as they seek to escape an untenable environment. The Gay Lesbian Straight Education Network’s (GLSEN) School Climate Report demonstrates the high levels = of transgender harassment in middle and high school. Eighty one percent of transgender = youth surveyed reported verbal harassment, 46% sexual harassment, 46% had their property damaged or stolen, and 24% reported physical harassment.= ADDIN EN.CITE <EndNote><Cite><Author>Kosciw</Author><Year>2= 004</Year><RecNum>105</RecNum><record><rec-numbe= r>105</rec-number><ref-type name=3D"Report">27</ref-type><contributors><aut= hors><author>Kosciw, J.G.</author></authors></contributors><titles><t= itle>The 2003 National School Climate Survey: The school-related experiences of our = nation’s lesbian, gay, bisexual and transgender youth.</title></titles><dates><year>2004</year&g= t;</dates><pub-location>New York</pub-location><publisher>GLSEN</publisher><urls&g= t;</urls></record></Cite></EndNote>¹⁵⁵ = They further noted that higher levels of harassment are linked with lower levels= of college aspirations. As with = the figures on suicide, this sample is potentially truncated as it is based on a survey of youth in school, and some transgender youth will have left school= to avoid the harassment. In Sausa’s 2003 dissertation on transgender youth, 96% of the (small) sa= mple experienced verbal harassment, and 75% dropped out of school.⁸³ =

Likewise, home-based harassment and violence is commonly reported in the transgender populations.⁸³ = In the words of Mara Keisling, the Executive Director of the National Center for Transgender Equity, "The classic profile, is a 13-year-old who's thrown out of the house when she decides to transition. She's kicked out of school for wearing girls' clothes. She can't get a job because her ID says 'Andre' but she looks like a girl. What's going to happen? Most likely, she= 'll end up in a situation that makes her especially vulnerable — living in shelters and low-income neighborhoods, doing sex work as a matter of survival."¹⁵⁰ Sausa notes that for transgen= der youth, “Traditional support systems such as family or friends were absent…. This lack of support and understanding led youth to feel isolated, alone, and pressured. Those who had even one supportive person or family member in their lives, excelled and could better advocate for themselves.” (p.111)⁸³ = This isolation and stress conspires to create immediate health risks, 83% of the sample reported suicide ideation, 79% used illicit drugs, and 96% reported having recent sex under the influence of drugs or alcohol.⁸³ =

Social exclusion

Social exclusion for a transgender person occur= s on a myriad of different levels. In personal interactions or institutionalized discrimination, a transgender pe= rson is often made to feel the extent to which they are not desirable members of= the community. “So, when yo= u go out into the street, you get these stares, like you’re a freak from o= uter space somewhere. And you go i= nto a restaurant or something and everyone is over there giggling. But to me this is normal.” (p. 510)¹⁰² = From the daily interactions, to the persistent threat of violence, difficulties in getting employment, housing, health providers, and a supportive community environment, transgenders report living in a world that is seldom welcoming, and too often openly hostile. In this environment, basic life needs such as housing are often difficult to meet. In the New York City needs assessment, housing= was identified as the second most community concern. Overall 18% of the MTF and 9% of t= he FTM group reported currently being homeless.⁸¹ = In a Philadelphia needs assessment, 51% of MTF and 55% of FTM describe their liv= ing situation as temporary.¹⁰³ = In the Washington, D.C. needs assessment, 48% of respondents reported housing as o= ne of their top three current needs.