An Iatrogenic Imperative

Weight-Goal Pursuit and Morbidity from Adolescence to Adulthood

Victoria Sass

Thank you all for being here, and for the years of support that made any of this possible. It means more to me than I can say.

Just contextualize everything a bit before we jump into the presentation.

There’s one person I most wish were here today: my mom, who many of you know passed away at then end of March. The two of us spent much of our lives at odds with our own bodies — something we shared deeply, even when we couldn’t talk about it. This dissertation grew directly out of that experience, as a way in which to understand society and myself a little better. And it’s also for the many people, like my mom, who never had the chance to experience it differently.

At the heart of this research is something very basic and human: the desire for safety and connection — in the world and in the body that carries us through it. Health is often presented as a means of achieving that sense of security, framed as a static endpoint to be reached and defended. But a body isn’t a destination. It changes, gets sick, ages, it slips our grip; so much of what happens to it was never ours to individually control in the first place. We’re taught to believe otherwise though, asked to chase a permanence that doesn’t exist; that with enough discipline we can change the outcome. What this work argues is that the pursuit, in and of itself, can be harmful.

Roadmap

---

• Motivation
• Theoretical Framework
• Dissertation Studies
  1. Weight goal formation
  2. Weight control behaviors
  3. Health outcomes
• Limitations & Future Directions

To give you a sense of the structure of my presentation I’ll begin with a brief motivation of my research and introduce the main theoretical framework used to advance my arguments.

Then we’ll dive into to the three studies that comprise my dissertation, which build upon one another. I’ll begin with an overview of the data used throughout, before discussing each study in turn.

Then, I’ll close with a brief discussion of limitations and future directions.

Motivation

“Until we have better data about the risks of being overweight and the benefits and risks of trying to lose weight, we should remember that the cure for obesity may be worse than the condition.”

Jerome P. Kassirer, M.D., and Marcia Angell, M.D. 
Losing Weight - An Ill-Fated New Year’s Resolution
The New England Journal of Medicine (1998)

This quote, written almost 30 years ago, gets at the core of what this dissertation seeks to examine. It reads, “Until we have better data about the risks of being overweight and the benefits and risks of trying to lose weight, we should remember that the cure for obesity may be worse than the condition.”

The Weight-Centered Health Paradigm

Historical shift to anti-fat bias

Fatness often signaled wealth, health, and status in pre-industrial societies (Oliver 2006b)

Modern anti-fat bias has racialized and racist origins, forged by narratives used to justify the transatlantic slave trade and the spread of Protestant asceticism (Strings 2019)

Amplified by the fashion and beauty industries which have promoted the thin-ideal, directed especially at women (Bordo 1993; Gaesser 2002; Oliver 2006a; Wolf 2009)

Development and clinical endorsement of the Body Mass Index (BMI)

Created in the 19th century to describe the population-level distribution of body size that captures “no more than half the variation of body fat” (Keys et al. 1972; Quetelet 1842)

Turned into “ideal weight” by life insurers pricing mortality risk, then enshrined as a clinical standard by a task force funded by weight-loss drug companies (Gilman 2010; Oliver 2006a)

Its “overweight” and “obese” thresholds are arbitrary — debated for decades and unsupported by the very mortality data used to justify them (Fletcher 2014; Troiano et al. 1996)

The medicalization of body size

The medicalization of “obesity” was extremely contested, including ad hominem attacks of CDC researchers who found evidence “overweight” showed lower mortality (Flegal et al. 2005, 2009; Flegal 2021)

In 2013 the AMA voted obesity a disease, overruling its own committee’s warning that it didn’t fit the definition and would only deepen stigma (Association 2013; H-440.842 Recognition of Obesity as a Disease | AMA n.d.)

Associations between morality and body size now acquired the authority of a medical diagnosis, and weight loss became an explicit health pursuit (Campos 2004; Conrad 2007)

CLICK History of anti-fat bias - For much of history a larger body has signaled wealth and health as food was often scarce and labor physically demanding for the majority of people. Thinness as the ideal is recent. As Sabrina Strings argues, that reversal wasn’t about health: it grew from the racist narratives Europeans used to justify the slave trade, fused with Protestant ideas casting self-restraint as a sign of virtue. By the early 1800s these ideas fused into an ideology that linked Blackness to fatness and fatness to moral failure, intensifying with the growing fashion and beauty industries in the 20th century that turned thinness into a product — marketed relentlessly, and above all to women.

This historical context, Strings argues, created the widespread cultural disdain for fatness that was a necessary precursor for the pathologization of body size.CLICK

BMI - As racialized notions of body size and morality were coalescing, what’s now known as the Body Mass Index (weight/height^2) was being developed. The BMI isn’t actually a measure of health — it’s a population statistic first developed in 1842, based on white European men, that, when applied to the individual - explains no more than half the variation in body fat. It was made into a measure of individual health first by insurers pricing mortality risk and then by a task force funded by weight-loss drug companies. Crucially, the lines it draws between its categories are essentially arbitrary, contested for decades and never supported by the very data used to justify them. CLICK

Medicalization of obesity - Obesity became a ‘disease’ not because of an overwhelming preponderance of evidence - in fact, the best data & methods available at the time actually showed lower mortality rates among those classified “overweight,” results met with hostility and personal attacks of the researchers themselves. In 2013 the AMA decided to overrule its own expert committee on the issue, which had voiced concerns about the BMI’s imprecision, the lack of a clear definition as a disease, and the potential for increasing stigma. Driven by moral judgment and financial interest, body size became an official measure of health — and weight loss was recast from an aesthetic pursuit into an individually-mandated health imperative. CLICK CLICK

Theoretical Framework

Critical counter-narrative

Dieting is not effective for sustained weight-loss

Long-term, dieting fails up to ~95% of the time; most people regain everything they lost, and often more (Gaesser 2002; Mann et al. 2007)

Body weight is 70–80% heritable, and the body defends a “setpoint” — fighting back with hunger and a slowed metabolism when adequate intake decreases (Bacon and Aphramor 2014; Wardle et al. 2008)

When health does improve, it tracks with eating varied/nutrient dense foods and moving more — not the weight loss itself (Barry et al. 2014; Tomiyama, Ahlstrom, and Mann 2013)

Dieting has negative health effects

Because diets fail, most people weight cycle — and recurring bouts of loss and regain is linked to inflammation (Strohacker and McFarlin 2010), hypertension (Guagnano et al. 2000; Kajioka et al. 2002), diabetes (French et al. 1997; Montani et al. 2006; Vergnaud et al. 2008), cancer (Thompson and McTiernan 2011), and higher mortality (Diaz, Mainous, and Everett 2005; Ernsberger and Koletsky 1999; Lissner et al. 1991; Nguyen et al. 2007; Rzehak et al. 2007), compounding with each cycle

There are psychological costs as well: body dissatisfaction, low self-esteem, disordered eating, and internalized stigma (Catling and Malson 2012; Cogan and Ernsberger 1999; Ikeda, Crawford, and Woodward-Lopez 2006; Kassirer and Angell 1998; Lawrence, Hazlett, and Abel 2012; Pieterman 2007; Puhl and Heuer 2010; Rees et al. 2011)

Crucially, these are the same biological pathways — chronic inflammation, stress-hormone dysregulation, allostatic load — that the field blames on fatness itself (McEWEN 1998; Tomiyama 2014)

CLICK CLICK Dieting’s failure - However, the most common treatment prescribed — dieting — almost never works long-term. It fails up to ninety-five percent of the time, and most people regain everything, often more. The reason is biological: body weight is seventy to eighty percent heritable and the body defends a setpoint weight similarly to the way it regulates internal temperature, responding to less food with more hunger and a slower metabolism. When dieting does improve health markers, the gains come from eating better and moving more — not from the weight loss itself. CLICK

Dieting is harmful - Because diets so reliably fail, most people engaged in them end up yo-yo dieting, or weight cycling which is linked to inflammation, high blood pressure, diabetes, and higher mortality, with the damage compounding each time. The additional psychological costs of perceived failure include body dissatisfaction, low self-esteem, disordered eating, and internalized stigma. And the key takeaway here is that the biological wear shown to be associated with dieting is the same biology the obesity field points to when it claims that fatness damages health. In other words, the signature blamed on the fat body is equally consistent with the strain of trying to shrink it. CLICK CLICK

Framework for analysis

Three theoretical lenses allow for the paradigm’s own logic to be reversed — locating the harm not in the body, but in the pursuit of changing it.

Healthism (Crawford 1980)

Health is recast as an individual moral duty and therefore illness is attributable to a lack of discipline

Expands medical jurisdiction into daily habits, making routine personal choices matters of clinical concern rather than personal preference

Structural causes - poverty, environment, and inequality - get deprioritized and recast as personal failings, simultaneously offering an illusion of control while the conditions go unchallenged

Life-course perspective

Individual health trajectories are the accumulation of exposures, behaviors, and orientations that are shaped by their timing, duration and ordering across developmental stages (Dannefer 2003; Elder 1998; Geronimus et al. 2006; Kuh et al. 2003; Lynch and Smith 2005; Shanahan 2000).

Trajectories and their consequences are additionally conditioned by the historical period in which they unfold (Elder 1998; Ryder 1965).

Sensitive periods, such as adolescence and the transition to young adulthood, are developmental windows in which exposure to particular conditions or experiences have an outsized effect on subsequent trajectories (Ben-Shlomo and Kuh 2002; Shanahan 2000).

Embodiment (Krieger 2001; Krieger 2005)

The social and material conditions of our lives are literally incorporated into the body.

CLICK To ask whether pursuing thinness itself could be harmful requires stepping outside the paradigm’s own logic — and three theories provide the framework for my analyses. CLICK

The first is healthism — Robert Crawford’s term for what happens when health stops being a state of the body and becomes a measure of one’s character. Individuals are held responsible for it, and to fall short is read as a moral failing. Healthism also extends medicine into everyday life such that individual choices - what you eat, how you move, how you handle stress - become matters of clinical concern rather than personal preference. And by recasting structural problems as personal ones, it simultaneously shifts focus from the systems that overwhelming influence health while offering individuals an illusion of control. That’s why a failing pursuit doesn’t get abandoned — it gets redoubled. CLICK

The second is the life-course perspective which argues that early life experiences develop into trajectories and their effects accumulate over decades. These effects are additionally conditioned by the historical period in which individuals experience them, such that one’s orientation towards their own body carries different significance depending on the social and economic moment that frames it. Crucially, a weight goal formed during the developmentally sensitive period of adolescence - where the body naturally changes due to puberty and becomes a more salient object of scrutiny and comparison - is likely to be more durable and develop into a lifelong orientation. CLICK

Lastly is the ecosocial theory of embodiment which argues that human bodies literally incorporate the material and social conditions of their existence, refusing the partition of the social and the biological into separate explanatory domains. It insists that the chronic activation of stress physiology, the metabolic consequences of dietary practice, and the psychological residue of repeated self-evaluation are not metaphors for social experience but its physical instantiation. CLICK CLICK

Dissertation studies

Turning now to the dissertation, I’ll go through the research questions of each study and the data used for all three before describing the main results of each.

Research questions

1. Weight goal formation
   What individual, household, and contextual characteristics predict membership in distinct weight-goal trajectories across adolescence and early adulthood?

2. Weight-control behaviors
   How does weight-goal trajectory membership translate into engagement in weight-control behaviors (both frequency and extremity)? Does a history of dieting independently predict escalation into clinically extreme practices?

3. Health outcomes
   Do weight-goal trajectories and the behavioral engagement they motivate predict cardiometabolic and mental-health morbidity (and their comorbidity) in adulthood? To what extent is body size implicated in this association?

This diagram gives a crude but succinct overview of the dissertation to help orient each study’s research question within the broader theoretical relationships. CLICK

The first study seeks to better understand the process and stability of weight-goal formation and asks what individual, household, and contextual characteristics predict membership in distinct weight-goal trajectories across adolescence and early adulthood. CLICK

The second study takes those weight goal trajectories and asks to what extent they translate into engagement in weight-control behaviors, looking at both behavior frequency and extremity. Additionally, it asks whether dieting itself, the most ubiquitous and commonly prescribed weight-loss behavior, independently predicts engagement in more and more clinically extreme weight-control behaviors. CLICK

Lastly, the third study uses those same weight-goal trajectories and weight-control behavior measures to understand whether they predict higher rates of cardiometabolic and/or mental-health diagnoses and their co-occurence. It also tests to what extent these associations can be independently attributed to body-size itself. CLICK

Data

• The National Longitudinal Study of Adolescent to Adult Health¹ (Add Health)
  • Nationally-representative, school-based cohort study of U.S. adolescents in grades 7-12, drawn via stratified, clustered probability sampling
  • Baseline (Wave I) survey in 1994-1995, with follow-ups at:

• Wave II (1996)
• Wave III (2001-2002)

• Wave IV (2008-2009)
• Wave V (2016-2018)

CLICK Data come from the National Longitudinal Study of Adolescent to Adult Health, which is an ongoing, nationally representative survey of adolescents who were in grades 7-12 during the 1994-5 academic year. A subsample of almost 21,000 students were selected for a more-extensive in-home survey at Wave I and these respondents were followed-up at Waves II, III, IV, & V, by which time the cohort was in their late 30s to early 40s.

Of note for the purposes of these studies is that the Add Health cohort moved through adolescence and into early adulthood precisely as the weight-centered health paradigm was consolidating its institutional authority, encountering a cultural and medical environment in which the imperative to manage body weight was both more medicalized and more commercially saturated than it had been for earlier cohorts.

Included in all three studies (unless otherwise noted) are variables derived from Wave I that capture theoretically important individual-, household-, and neighborhood-level measures of interest.

1. A Project of the Carolina Population Center (Harris et al. 2019)

Individual-level variables

• Sex (binary: male/female)
• Age (in years)
• Race
  • Asian/Pacific Islander
  • Black
  • Multiracial/Other Race
  • Native American
  • White
• Latino/a (binary: no/yes)
• U.S. Born (binary: no/yes)
• Hours of Sleep (average)
• Met MVPA¹ Recommendations (binary: no/yes)

• BMI Classification
  • ‘Underweight’
  • ‘Healthy Weight’
  • ‘Overweight’
  • ‘Obese’
• Weight Self-Image
  • Very Underweight
  • Slightly Underweight
  • About the Right Weight
  • Slightly Overweight
  • Very Overweight

CLICKIndividual-level variables include key demographic indicators, CLICK 2 baseline behavioral controls for sleep and physical activity, and 2 key body-size measures: BMI classification and weight self-image.

Categories in purple denote the reference for that variable with the exception of race where each category was estimated versus all-others-combined rather than imposing an arbitrary reference on a categorical variable with no inherent ordering.

1. Moderate-to-Vigorous Physical Activity

Household-level variables

• Food Choice Autonomy (binary: no/yes)
• Proportion Dinner w/ Parent(s) (per week)
• Social Origins Score¹
• Parent Graduated High School (binary: no/yes)

CLICK Household-level measures include whether respondents reported getting to make their own decisions about what they eat, frequency of family meals, a composite index of socioeconomic advantage, and whether either parent graduated high school.

1. Composite index constructed from measures of parental education, parental occupation, household income, and household receipt of public assistance

Neighborhood-level social variables

• Neighborhood SES Disadvantage¹
• Violent Crimes per/100,000
• Neighborhood Feels Safe (binary: no/yes)
• Metropolitan Area (binary: no/yes)

CLICK Neighborhood-level social measures include a composite index of socioeconomic disadvantage, the rate of violent crime, a subjective assessment of neighborhood safety, and a metropolitan area indicator.

1. Composite index constructed from tract-level proportions of female-headed households, poverty, public assistance receipt, low educational attainment, and unemployment.

Neighborhood-level environmental variables

• Climate (nearest climate station)
  • Principal component 1 & Principal component 2
• Physical Activity Resources (3km)
  • Principal component 1 & Principal component 2
• Neighborhood Navigability / Walkability (3km)
  • Ratio of streets to intersections: Beta Index
  • Prevalence of dead-end street configurations: Cul-de-sac Density
  • Proportion of roads classified Local, Neighborhood, and Rural Road: CFCC A40-A48
• Land Use / Cover (3km)
  • Degree of land use fragmentation/uniformity: Landscape Patch Density
  • Developed, recreational land: Class 4 Area
  • Undeveloped/natural land: Class 5 Area

CLICK Also included are a battery of neighborhood environmental measures — climate, physical-activity resources, walkability, and land use — to capture the built and natural context respondents are embedded in.

Weight goal formation

Outcome variable

• Weight goal:
  • Lose weight
  • Maintain weight
  • Gain weight
  • Not trying to do anything about weight

Wave I

Wave II

Wave III

The first study seeks to understand what predicts membership in 6 theoretically-derived weight-goal trajectories. CLICK These orientations were constructed from a survey item available in Waves I-III which asked, “Are you trying to lose weight, gain weight, or stay the same weight?” with options for Lose, Maintain, Gain, and Nothing.

Outcome variable

• Weight goal:
  • Lose weight
  • Maintain weight
  • Gain weight
  • Not trying to do anything about weight

• Weight goal trajectory:
  • Consistent Lose
  • Consistent Maintain
  • Consistent Gain
  • Consistent Nothing
  • Changing Lose/Maintain
  • Changing Goals

Trajectories were created from combined Wave I-III responses to this question for each respondent. Consistent trajectories represent individuals who reported the same weight goal across survey waves and Changing trajectories reflects those who changed goals at least once across waves. Changing Lose/Maintain was separated from the broader Changing Goals trajectory due to its theoretical similarity with Consistent Lose and Maintain.

The distribution of weight-goal trajectories shows that the majority of respondents were changing goals between waves, followed by Consistent Lose as the next largest trajectory. Notable from these descriptives is the smallest trajectory category - Consistent Nothing, signifying the broad engagement in some active weight-goal trajectory among respondents during their adolescent and early adult years.

As the largest category, Changing Goals warrants some additional descriptives…

Broken down by distinct weight goals we can see that the majority of this group was engaging in weight maintenance and weight loss, along with no weight goals. Additionally, it was less common among respondents to cycle between three different weight goals than it was to cycle between two.

Results

The method for this study is a series of survey-weighted multinomial logistic regression models, with Consistent Nothing as the reference, adding one domain of predictors at a time. Results are from the final, fully-estimated model.

One of the clearest findings from the first study is the asymmetry of recruitment into different weight-goal trajectories. Each row is a trajectory with points to the right of the solid vertical line more likely among girls and women and to the left among boys and men. Relative to males, females have about five times the odds of membership in Consistent Lose, three times the odds of Changing Lose/Maintain, but only a fifth the odds of Consistent Gain. In other words, the same cultural system motivates female bodies toward reduction and male bodies toward gain.

This figure layers in the two body-size measures to make visible how all three significant predictors combine to produce these trajectories. It shows the survey-weighted average predicted probabilities across their joint distribution. Columns are weight self-image, from very underweight to very overweight; within each panel the x-axis is BMI classification; the top row is men, the bottom women. The headline is the relative power of the two. The clinical label does predict loss goals — but self-image predicts them far more strongly, at every level of the distribution. Focusing on the teal line, Consistent Lose, in the bottom row: for women, simply perceiving oneself as slightly overweight or very overweight all but eliminates the other trajectories in favor of loss — across every BMI category, including clinically healthy and even underweight bodies. The same logic runs in reverse for men and Consistent Gain. The clinical category the paradigm treats as its rational basis is a weaker driver of the goal it’s meant to produce than the internalization of body size norms.

Weight-control behaviors

If study one asks what motivates aspiration, study two asks what behaviors that aspiration motivates.

Outcomes & Methods

• Lose / Maintain
  • Dieted
  • Exercised
  • Purged
  • Took diet pills
  • Took laxatives
  • Fasted
  • Used diuretics
  • Took food supplements
  • Weight-loss program
  • Other

• Gain
  • Ate different foods
  • Ate more foods
  • Exercised
  • Lifted weights
  • Took food supplements
  • Took steroids

• Universally-surveyed
  • Skipped 2+ meals / day
  • Used anabolic steroids
  • Used performance-enhancing drugs
  • Binge eating
  • Excessively exercise
  • Specialized diet

Total
Behaviors

Ever
Extreme

Total
Extreme

If a respondent reported an active weight goal in any given wave they were asked a follow-up set of questions about what behaviors they were engaged in to achieve that goal. CLICK

Lose and Maintain were asked the same set of questions, CLICK Gain a slightly different set and those with no weight goal weren’t asked any. CLICK Additionally included were behaviors asked of everyone that were related to weight-control. CLICK

Behaviors were categorized as non-extreme and extreme based on existing literature and counts of each type were summed across Waves I-III for each respondent.

Three related models were estimated, each with a different outcome. CLICK The first used the total count of weight-control behaviors, CLICK the second a binary indicator for ever engaging in an extreme behavior, CLICK and the third a count of total extreme behaviors, estimated only on those who reported any extreme engagement. I carry the trajectories forward as the exposure, now with Consistent Maintain as the reference — the group that stays engaged with weight management but isn’t trying to change their body size.

I’ll show results from two samples: everyone with an active goal, and a subsample of those exposed to the dieting for weight loss / maintenance question where dieting is excluded from the outcome and used as a binary predictor of engagement in other behaviors and/or escalation into more extreme behaviors. All covariates from study one were included in these models with the exception of the two body-size related measures.

Extreme behaviors highlighted in yellow.

Results

Reading left to right, the outcomes get more severe: frequency of any behavior, then crossing into any extreme behavior, then frequency among those already engaged in extreme behaviors. The top block is the all-behaviors sample, the bottom the dieting subsample. Notable across the progressively escalating outcomes are the increasing magnitudes of each of these statistically significant effects.

This makes the gradient explicit for the all behaviors sample. Each line is the same predictor and the different points are their effects from each of the three models; the y-axis is on the log scale and shows the effect relative to Consistent Maintain for the four trajectories and relative to Male for Female. For the loss-oriented and unstable trajectories — Consistent Lose especially — the effect climbs steadily from the first model on the left to the third model on the right. It’s weakest where behavior is most common and strongest where it’s most harmful: Consistent Lose engages in extreme behaviors at about 3.5 times the rate relative to Consistent Maintain.

For girls the first model actually shows lower overall engagement in weight-control behaviors compared to boys but in models 2 and 3 we can see that they are about 25% more likely to cross into extreme behavioral engagement and, among those engaged, have a 40% higher rate of extreme behaviors, per question asked.

The one break is Consistent Gain — elevated early and on-par with Consistent Lose in crossing into extreme behaviors, but is not significantly different from Consistent Maintain in their rate of engagement in extreme behaviors.

The dieting subsample isolates the paradigm’s single most sanctioned prescription, and asks whether dieting carries risk on its own, net of the trajectory. It does — and it follows a similar escalating effect size across the three models.

On overall frequency of any behavior dieting doesn’t substantively broaden the repertoire. But its effects become more pronounced in models 2 and 3 where dieters are 33% more likely to cross into extreme behavior, and engage in those behaviors at roughly 66% higher rates. And the trajectory lines stay elevated right alongside it. So a sustained orientation toward loss and a history of acting on it through dieting are distinct, additive routes into extreme practice.

Health outcomes

The final study takes up the paradigm’s most consequential claim — that the pursuit of a normative body size will result in better health outcomes.

Outcomes & Methods

The primary analytic sample used for this analysis was the dieting sub-sample from study 2…

Outcomes & Methods

• Cardiometabolic Diagnoses
  • Diabetes
  • Hypertension
  • High cholesterol
  • Heart disease
  • Heart failure
  • Arterial disease
  • Atrial fibrillation
  • Blood clots
  • Kidney disease
  • Stroke

…and the outcomes are cumulative counts of self-reported, provider-diagnosed conditions in adulthood, ten cardio-metabolic…

Outcomes & Methods

• Cardiometabolic Diagnoses
  • Diabetes
  • Hypertension
  • High cholesterol
  • Heart disease
  • Heart failure
  • Arterial disease
  • Atrial fibrillation
  • Blood clots
  • Kidney disease
  • Stroke

• Mental Health Diagnoses
  • Depression
  • Anxiety
  • Post-traumatic stress disorder
  • Attention-deficit/hyperactivity disorder

…and four mental health.

Using these outcomes three different types of models were estimated…

Outcomes & Methods

• Cardiometabolic Diagnoses
  • Diabetes
  • Hypertension
  • High cholesterol
  • Heart disease
  • Heart failure
  • Arterial disease
  • Atrial fibrillation
  • Blood clots
  • Kidney disease
  • Stroke

• Mental Health Diagnoses
  • Depression
  • Anxiety
  • Post-traumatic stress disorder
  • Attention-deficit/hyperactivity disorder

Cardiometabolic

Mental health

Comorbidity

…a series of progressive specifications for both count outcomes, and an unconstrained bivariate logistic regression of the fully specified model in which each count outcome is turned into a binary indicator of any diagnosis to enable estimation of the marginal effects of each diagnosis type…

Outcomes & Methods

• Cardiometabolic Diagnoses
  • Diabetes
  • Hypertension
  • High cholesterol
  • Heart disease
  • Heart failure
  • Arterial disease
  • Atrial fibrillation
  • Blood clots
  • Kidney disease
  • Stroke

• Mental Health Diagnoses
  • Depression
  • Anxiety
  • Post-traumatic stress disorder
  • Attention-deficit/hyperactivity disorder

Cardiometabolic

Mental health

Comorbidity

…as well as their co-occurence, or comorbidity.

Body-size exposure measures

Additionally, to adjudicate between the argument being advanced in this dissertation and the argument of a weight-centered health paradigm, the two body size measures were transformed to capture accumulated exposure to non-normative clinical classification and subjective appraisal across Waves I-III.

Body-size exposure measures

BMI classification · cumulative departure from ‘Healthy Weight’

+1
underweight

+1
overweight

+2
obese

Reference · 0
Healthy Weight

For BMI, for example, relative to the normative reference BMI category of ‘Healthy Weight’, two directional measures were summed across waves…

Body-size exposure measures

BMI classification · cumulative departure from ‘Healthy Weight’

← under-classification

over-classification →

no +2 tier

+1
underweight

+1
overweight

+2
obese

Reference · 0
Healthy Weight

…such that each respondent had a value from 0-3 for how often they’d been classified as under the normative weight and a similar value from 0-6 for how often they’d been classified as over the normative weight.

Body-size exposure measures

BMI classification · cumulative departure from ‘Healthy Weight’

← under-classification

over-classification →

no +2 tier

+1
underweight

+1
overweight

+2
obese

Reference · 0
Healthy Weight

Weight self-image · cumulative departure from ‘About the Right Weight’

+2
very underweight

+1
slightly underweight

+1
slightly overweight

+2
very overweight

Reference · 0
About the Right Weight

The same rationale applied to the weight self image appraisals…

Body-size exposure measures

BMI classification · cumulative departure from ‘Healthy Weight’

← under-classification

over-classification →

no +2 tier

+1
underweight

+1
overweight

+2
obese

Reference · 0
Healthy Weight

Weight self-image · cumulative departure from ‘About the Right Weight’

← under-image

over-image →

+2
very underweight

+1
slightly underweight

+1
slightly overweight

+2
very overweight

Reference · 0
About the Right Weight

Each wave scored +1 or +2 based on proximity to the normative category, summed across Waves I–III and separately by direction
↓
bmi_over · bmi_under · wsi_over · wsi_under

…with the only difference being the symmetrical nature of the over-/under- scales that doesn’t exist for the BMI classification system.

This resulted in 4 body-size cumulative exposure measures, 1 over and 1 under for both body size variables.

This construction enabled an informal mediation analysis of their independent effects on health outcomes relative to trajectories and behaviors, with the central test being whether those focal predictors’ effects survive adjustment for body size — and through which channel they travel.

Results

This figure traces the trajectory effects across the full sequence of models, cardiometabolic on top, mental health below, with Consistent Maintain again the reference. The left portion adds successive blocks of controls; the right, after the dotted line, adds the body-size mechanisms. Several things stand out.

First, the active-goal trajectories carry more of both kinds of morbidity, and the effects are stable through every social and contextual adjustment, with the behaviors block actually revealing a stronger effect for the Changing Goals while attenuating the other two trajectories. In the fully adjusted models, risk of disease burden is on the order of 40–85% higher for cardiometabolic outcomes and 23–37% higher for mental health.

Second — the unstable Changing trajectories survive even after accounting for clinical classification and self-image, on the far right. Consistent Lose largely doesn’t — its cardiometabolic burden runs almost entirely through accumulated classification and self-appraisal in the overweight direction (as we’ll see on the next slide). But a body-size-as-cause account has no explanation for why an unstable pattern of goals would damage cardiometabolic health independent of how heavy someone was ever classified as being. It is consistent, however, with the theoretical arguments advanced in this dissertation.

For the next two slides I’ll present results from the bivariate logistic regression model of the fully-specified model with both body size exposures included, which enables an estimation of the co-occurence, or comorbidity of any cardiometabolic and mental health diagnosis.

Clinical classification, is shown in pink, and self-image, in red. Over-classification is on the right and under-classification on the left.

One thing to note about all of the effect sizes presented on this slide is that they are relatively small. In other words, the independent effects of body size are not as strongly predictive of either a cardiometabolic diagnosis, a mental health diagnosis, or their co-occurrence that one would believe from a weight-centered health perspective.

But looking at what the effect sizes do suggest, a weight-centered account predicts that self-image should be redundant once the clinical measure is in the model — a noisier reading of the “objectively” classified body. Instead, each carries an independent, comparable association with cardiometabolic burden. And for mental health, in the middle row, it’s the perceived body, not the measured one, that increases the odds of a diagnosis. Jointly, each additional deviation above feeling ‘About the Right Weight’ increases the odds of comorbidity by 24%.

Looking at under-classification on the left, however: being clinically underweight is protective — but perceiving oneself as underweight is not statistically significant. This asymmetry is interesting because it provides some evidence of the arbitrary nature of the BMI’s own classification system. If the normative BMI category ‘Healthy Weight’ were a genuine marker of health, one would expect there to be negative health effects in both directions, not just one. The evidence is limited but this suggests that what matters isn’t deviation from a normative weight in any direction but the stigmatized direction specifically — being and feeling over the line that our culture and medical-industrial complex marks for correction. This lends support to the argument that stigma itself is implicated in the very health conditions the paradigm attributes to body size alone.

From the same model, the effects of weight control behaviors align with what the theoretical framework predicts.

Extreme behavior stands apart as a distinct mental-health risk with 28% higher odds of any diagnosis for each additional extreme behavior reported.

Crucially though, dieting itself is elevated across all three outcomes — and most strongly for comorbidity. Dieting doesn’t just raise each outcome separately; it concentrates their co-occurrence, raising the odds of having a diagnosis in both families 58% higher than for those with no history of dieting. In other words, the paradigm’s most prescribed instrument is the one most tied to carrying both kinds of disease at once.

Limitations & Future Directions

Limitations

• Inability to definitively prove causation
• Unmeasured confounders
• Measurement bias
  • Self reports
  • Extreme behavior a proxy for clinical risk
  • Weight self-image inherently subjective
  • Weight goal doesn’t distinguish underlying motivations

As with any study, several limitations warrant mentioning. CLICK

The data used in these analyses are observational and thus it is not possible to establish definitive causal relationships between weight goal trajectories, weight control behaviors, body size, and health outcomes. CLICK

The study design can’t exclude every unmeasured common cause of these main predictors and diagnoses, so it adjudicates between two accounts rather than settling causation outright. However, calculated E-Values suggest that a confounder would have to be implausibly strong and specifically structured, generally stronger than anything I measured, to explain the main findings away. CLICK

The measures carry their own limits as well. Most measures are self-reported which introduces the risk of measurement error, recall bias, or social desirability bias, which could affect the precision of the estimates. CLICK The behavioral items weren’t built to diagnose disordered eating, so the extreme-behavior measure is a proxy for clinical risk, not a clinical assessment. CLICK Self-image is a single scale that can’t separate the thin ideal from the muscularity ideal — a real limitation for the men in the sample — CLICK and the goal items don’t distinguish an aesthetic motive from a medical one, or a goal a clinician prescribed from one a person set for themselves.

Many of these limitations point the same way: toward data designed to model weight goals directly over time — where they come from and how they shift — and toward qualitative work on the small group who never develop a weight goal orientation.

Future Directions

• Perceived, not measured, size drove findings in every study
• Medical mistrust and the wellness-industrial complex
• Influence of internet culture and social media
• Intensifying body image ideals and behaviors among boys & men
• Othering the human body
• Dismantling the weight-centered health paradigm

CLICK One thread runs through all three studies and sets up numerous potential avenues for future work. At every stage — goal formation, behavior, and the morbidity that followed — the internalized sense of being too large did more work than the measured fact of body size: the mark of a culturally installed standard rather than adiposity itself. And what gives that standard its grip is medical endorsement — the medicalization of obesity recast the smaller body as the healthier one, so the demand to attain an arbitrarily-defined normative body size comes coupled with the obligation of a moral imperative. CLICK

A weight-centered health apparatus that prescribes a regimen that is hard to sustain, ineffective on its own terms, and implicated in the harm it claims to avert gives people good reason to doubt it — and when trust in expertise fails, the pursuit of health does not stop, it migrates toward a largely unregulated and non-evidenced-based health and wellness market. A paradigm that breeds that distrust is poorly placed to object to what rushes in to fill the void and there is ample opportunity to better understand how people individually define health and their avenues for improving it within the constraints of our healthism-oriented culture. CLICK

The Add Health cohort predates the current social media era, which has entirely restructured the basic character of how the weight-centered health paradigm’s messaging is transmitted and internalized. Social media platforms create environments of persistent body comparison and self-surveillance that internalize the outsider’s gaze as a near-constant psychological presence. For present-day adolescents navigating the developmental period these studies examine, social media has substantially raised both the intensity of body-image pressure and the behavioral infrastructure available for responding to it. The same platforms do carry real countercurrents — body-positivity, anti-diet, and Health at Every Size to name a few. Extensions of this work would be well-suited to examining the ways in which this particular cohort-effect plays a role in the relationships current adolescents have with their bodies and perceptions of health. CLICK

The muscularity ideal among boys and men, evidenced by the asymmetric sex recruitment into weight goal trajectories, is another avenue of research that warrants further investigation. More recently, the emergence of explicitly appearance-focused online communities organized around so-called “looksmaxxing” — the practice of systematically modifying one’s appearance to maximize perceived physical attractiveness — represents a new and largely undocumented front in the behavioral ecology of body management among young men. Many of the behaviors promoted by these online communities are objectively extreme - including bone-smashing, limb-lengthening surgery, drug abuse, and extreme exercise and restrictive dietary protocols, yet a eating-disorder infrastructure already ill-equipped for men is even less prepared to address behavioral risk organized around a hegemonic-masculinity body-optimization framework intensified by social media feedback loops. CLICK

Set within a longer cultural lineage, none of this is fundamentally new. As Sabrina Strings argues, the equation of fatness with disorder and thinness with virtue took shape within a racialized Western project of distinguishing the rational, self-governing person from the body and its appetites. Its newest form is the idiom of optimization, in which the body is discussed as something to be hacked, programmed, and upgraded, tuned like a machine toward greater output and longevity — an idiom now made literal in the life-extension and transhumanist ambitions of a segment of the technological elite. Set beside Strings’ account, these projects look less like novelty than continuity: another attempt to escape the condition of being an embodied animal with needs and limits rather than a system to be engineered, and one that, as Émile Torres has argued, carries forward the eugenic impulse to rank and perfect human bodies. The weight-centered health paradigm is one ordinary, daily expression of that same drive to discipline the unruly body toward an idealized and governable form. Its ubiquity is also what makes it so damaging: lodged in routine medical advice and daily self-scrutiny, the demand reaches nearly everyone and has harmful consequences for individuals’ health and quality of life. CLICK

Which points to the most defensible direction of all: continuing the rich sociological tradition of moving the work of health away from the individual body and toward the conditions that produce it, but applied directly to the case of the pathologization of body size. Findings not presented today bear this out — across all three studies, what most protected adolescents was social connection and security: regular family meals above all, alongside neighborhood safety and a more advantaged social origin. What protects health lies largely outside the narrow domain the paradigm polices. And as long as research keeps body size at its analytic center — even research meant to expose the paradigm’s harms — it will continue to mistake the forest for the trees. A substantial portion of the ill health charged to fat bodies appears, on this evidence, to be the ill health of the campaign waged against them. Setting the body’s size aside as the object of concern — and taking up in its place the pursuit imposed upon it, the unequal conditions that determine who is enlisted and who is harmed, and the structural supports on which health genuinely depends — is the reorientation toward which these findings point. It is, in the end, an empirical correction and a moral one at once: a way of ceasing to produce the very harm we have so long misread as the fault of the bodies that bear it.

Thank you!

References

Association, American Medical. 2013. “Report of the Council on Science and Public Health - Report 3-A-13.”

Bacon, Linda, and Lucy Aphramor. 2014. Body Respect: What Conventional Health Books Get Wrong, Leave Out, and Just Plain Fail to Understand about Weight. Body Respect: What Conventional Health Books Get Wrong, Leave Out, and Just Plain Fail to Understand about Weight. Dallas, TX, US: BenBella Books.

Barry, Vaughn W., Meghan Baruth, Michael W. Beets, J. Larry Durstine, Jihong Liu, and Steven N. Blair. 2014. “Fitness Vs. Fatness on All-Cause Mortality: A Meta-Analysis.” Progress in Cardiovascular Diseases 56(4):382–90. doi:10.1016/j.pcad.2013.09.002.

Ben-Shlomo, Yoav, and Diana Kuh. 2002. “A Life Course Approach to Chronic Disease Epidemiology: Conceptual Models, Empirical Challenges and Interdisciplinary Perspectives.” International Journal of Epidemiology 31(2):285–93. https://www.ncbi.nlm.nih.gov/pubmed/11980781.

Bordo, Susan. 1993. Unbearable Weight: Feminism, Western Culture, and the Body. Berkeley: University of California Press.

Campos, Paul F. 2004. The Obesity Myth: Why America’s Obsession with Weight Is Hazardous to Your Health. New York: Gotham Books.

Catling, L., and Helen Malson. 2012. “Feeding a Fear of Fatness? A Preliminary Investigation of How Women with a History of Eating Disorders View ‘Anti-Obesity’ Health Promotion Campaigns.”

Cogan, Jeanine C., and Paul Ernsberger. 1999. “Dieting, Weight, and Health: Reconceptualizing Research and Policy.” Journal of Social Issues 55(2):187–205. doi:10.1111/0022-4537.00112.

Conrad, Peter. 2007. The Medicalization of Society: On the Transformation of Human Conditions into Treatable Disorders. Medical Humanities Collection PAH. Baltimore: Johns Hopkins University Press.

Crawford, Robert. 1980. “Healthism and the Medicalization of Everyday Life.” International Journal of Health Services 10(3):365–88. doi:10.2190/3H2H-3XJN-3KAY-G9NY.

Dannefer, Dale. 2003. “Cumulative Advantage/Disadvantage and the Life Course: Cross-Fertilizing Age and Social Science Theory.” The Journals of Gerontology: Series B 58(6):S327–37. doi:10.1093/geronb/58.6.S327.

Diaz, Vanessa A., Arch G. Mainous, and Charles J. Everett. 2005. “The Association Between Weight Fluctuation and Mortality: Results from a Population-Based Cohort Study.” Journal of Community Health 30(3):153–65. doi:10.1007/s10900-004-1955-1.

Elder, Glen H. 1998. “The Life Course as Developmental Theory.” Child Development 69(1):1–12. doi:10.2307/1132065.

Ernsberger, Paul, and Richard J. Koletsky. 1999. “Biomedical Rationale for a Wellness Approach to Obesity: An Alternative to a Focus on Weight Loss.” Journal of Social Issues 55(2):221–60. doi:10.1111/0022-4537.00114.

Flegal, Katherine M. 2021. “The Obesity Wars and the Education of a Researcher: A Personal Account.” Progress in Cardiovascular Diseases 67:75–79. doi:10.1016/j.pcad.2021.06.009.

Flegal, Katherine M., Barry I. Graubard, David F. Williamson, and Mitchell H. Gail. 2005. “Excess Deaths Associated with Underweight, Overweight, and Obesity.” JAMA 293(15):1861–67. doi:10.1001/jama.293.15.1861.

Flegal, Katherine M., Barry I. Graubard, David F. Williamson, and Mitchell H. Gail. 2009. “Response to "Biased Corrections or Biased About Corrections.” Obesity (Silver Spring, Md.) 17(5):940–40. doi:10.1038/oby.2009.13.

Fletcher, Isabel. 2014. “Defining an Epidemic: The Body Mass Index in British and US Obesity Research 1960–2000.” Sociology of Health & Illness 36(3):338–53. doi:10.1111/1467-9566.12050.

French, S. A., A. R. Folsom, R. W. Jeffery, W. Zheng, P. J. Mink, and J. E. Baxter. 1997. “Weight Variability and Incident Disease in Older Women: The Iowa Women’s Health Study.” International Journal of Obesity 21(3, 3):217–23. doi:10.1038/sj.ijo.0800390.

Gaesser, Glenn A. 2002. “Big Fat Lies: The Truth about Your Weight and Your Health.” Big Fat Lies : The Truth about Your Weight and Your Health.

Geronimus, Arline T., Margaret Hicken, Danya Keene, and John Bound. 2006. “‘Weathering’ and Age Patterns of Allostatic Load Scores Among Blacks and Whites in the United States.” American Journal of Public Health 96(5):826–33. doi:10.2105/AJPH.2004.060749.

Gilman, Sander L. 2010. Obesity: The Biography: The Biography. Oxford, UNITED KINGDOM: Oxford University Press, Incorporated.

Guagnano, M. T., E. Ballone, V. Pace-Palitti, R. Della Vecchia, N. D’Orazio, M. R. Manigrasso, D. Merlitti, and S. Sensi. 2000. “Risk Factors for Hypertension in Obese Women. The Role of Weight Cycling.” European Journal of Clinical Nutrition 54(4, 4):356–60. doi:10.1038/sj.ejcn.1600963.

H-440.842 Recognition of Obesity as a Disease | AMA. n.d. Retrieved April 5, 2023. https://policysearch.ama-assn.org/policyfinder/detail/*?uri=%2FAMADoc%2FHOD.xml-0-3858.xml.

Harris, Kathleen Mullan, Carolyn Tucker Halpern, Eric A. Whitsel, Jon M. Hussey, Ley A. Killeya-Jones, Joyce Tabor, and Sarah C. Dean. 2019. “Cohort Profile: The National Longitudinal Study of Adolescent to Adult Health (Add Health).” International Journal of Epidemiology 48(5):1415–1415k. doi:10.1093/ije/dyz115.

Ikeda, Joanne P., Patricia B. Crawford, and Gail Woodward-Lopez. 2006. “BMI Screening in Schools: Helpful or Harmful.” Health Education Research 21(6):761–69. doi:10.1093/her/cyl144.

Kajioka, Taeko, Shigeki Tsuzuku, Hiroshi Shimokata, and Yuzo Sato. 2002. “Effects of Intentional Weight Cycling on Non-Obese Young Women.” Metabolism - Clinical and Experimental 51(2):149–54. doi:10.1053/meta.2002.29976.

Kassirer, Jerome P., and Marcia Angell. 1998. “Losing Weight — An Ill-Fated New Year’s Resolution.” New England Journal of Medicine 338(1):52–54. doi:10.1056/NEJM199801013380109.

Keys, A., F. Fidanza, M. J. Karvonen, N. Kimura, and H. L. Taylor. 1972. “Indices of Relative Weight and Obesity.” Journal of Chronic Diseases 25(6):329–43. doi:10.1016/0021-9681(72)90027-6.

Krieger, N. 2005. “Embodiment: A Conceptual Glossary for Epidemiology.” Journal of Epidemiology and Community Health 59(5):350–55. doi:10.1136/jech.2004.024562.

Krieger, Nancy. 2001. “Theories for Social Epidemiology in the 21st Century: An Ecosocial Perspective.” International Journal of Epidemiology 30(4):668–77. doi:10.1093/ije/30.4.668.

Kuh, D., Y. Ben-Shlomo, J. Lynch, J. Hallqvist, and C. Power. 2003. “Life Course Epidemiology.” Journal of Epidemiology and Community Health 57(10):778–83. doi:10.1136/jech.57.10.778.

Lawrence, Shawn A., Rebekah Hazlett, and Eileen Mazur Abel. 2012. “Obesity Related Stigma as a Form of Oppression: Implications for Social Work Education.” Social Work Education 31(1):63–74. doi:10.1080/02615479.2010.541236.

Lissner, Lauren, Patricia M. Odell, Ralph B. D’Agostino, Joseph Stokes, Bernard E. Kreger, Albert J. Belanger, and Kelly D. Brownell. 1991. “Variability of Body Weight and Health Outcomes in the Framingham Population.” New England Journal of Medicine 324(26):1839–44. doi:10.1056/NEJM199106273242602.

Lynch, John, and George Davey Smith. 2005. “A LIFE COURSE APPROACH TO CHRONIC DISEASE EPIDEMIOLOGY.” Annual Review of Public Health 26:1–35. doi:10.1146/annurev.publhealth.26.021304.144505.

Mann, Traci, A. Janet Tomiyama, Erika Westling, Ann-Marie Lew, Barbra Samuels, and Jason Chatman. 2007. “Medicare’s Search for Effective Obesity Treatments: Diets Are Not the Answer.” American Psychologist 62:220–33. doi:10.1037/0003-066X.62.3.220.

McEWEN, BRUCE S. 1998. “Stress, Adaptation, and Disease: Allostasis and Allostatic Load.” Annals of the New York Academy of Sciences 840(1):33–44. doi:10.1111/j.1749-6632.1998.tb09546.x.

Montani, J.-P., A. K. Viecelli, A. Pr’evot, and A. G. Dulloo. 2006. “Weight Cycling During Growth and Beyond as a Risk Factor for Later Cardiovascular Diseases: The ‘Repeated Overshoot’ Theory.” International Journal of Obesity 30(4):S58–66. doi:10.1038/sj.ijo.0803520.

Nguyen, Nguyen D., Jacqueline R. Center, John A. Eisman, and Tuan V. Nguyen. 2007. “Bone Loss, Weight Loss, and Weight Fluctuation Predict Mortality Risk in Elderly Men and Women.” Journal of Bone and Mineral Research 22(8):1147–54. doi:10.1359/jbmr.070412.

Oliver, J. Eric. 2006a. “Fat Politics: The Real Story Behind America’s Obesity Epidemic.” Fat Politics : The Real Story Behind America’s Obesity Epidemic.

Oliver, J. Eric. 2006b. “The Politics of Pathology: How Obesity Became an Epidemic Disease.” Perspectives in Biology and Medicine 49(4):611–27. doi:10.1353/pbm.2006.0062.

Pieterman, Roel. 2007. “The Social Construction of Fat: Care and Control in the Public Concern for Healthy Behaviour.” Sociology Compass 1(1):309–21. doi:10.1111/j.1751-9020.2007.00009.x.

Puhl, Rebecca M., and Chelsea A. Heuer. 2010. “Obesity Stigma: Important Considerations for Public Health.” American Journal of Public Health 100(6):1019–28. doi:10.2105/AJPH.2009.159491.

Quetelet, Adolphe. 1842. A Treatise on Man and the Development of His Faculties. Edinburgh : W. and R. Chambers.

Rees, Rebecca, Kathryn Oliver, Jenny Woodman, and James Thomas. 2011. “The Views of Young Children in the UK about Obesity, Body Size, Shape and Weight: A Systematic Review.” BMC Public Health 11(1):188. doi:10.1186/1471-2458-11-188.

Ryder, Norman B. 1965. “The Cohort as a Concept in the Study of Social Change.” American Sociological Review 30(6):843–61. doi:10.2307/2090964.

Rzehak, Peter, Christa Meisinger, Gabriele Woelke, Sabine Brasche, Gert Strube, and Joachim Heinrich. 2007. “Weight Change, Weight Cycling and Mortality in the ERFORT Male Cohort Study.” European Journal of Epidemiology 22(10):665–73. doi:10.1007/s10654-007-9167-5.

Shanahan, Michael J. 2000. “Pathways to Adulthood in Changing Societies: Variability and Mechanisms in Life Course Perspective.” Annual Review of Sociology 26:667–92. https://www.jstor.org/stable/223461.

Strings, Sabrina. 2019. Fearing the Black Body: The Racial Origins of Fat Phobia. NYU Press.

Strohacker, K., and B. K. McFarlin. 2010. “Influence of Obesity, Physical Inactivity, and Weight Cycling on Chronic Inflammation.” Frontiers in Bioscience (Elite Edition) 2:98–104. doi:10.2741/e70.

Thompson, Henry J., and Anne McTiernan. 2011. “Weight Cycling and Cancer: Weighing the Evidence of Intermittent Caloric Restriction and Cancer Risk.” Cancer Prevention Research (Philadelphia, Pa.) 4(11):1736–42. doi:10.1158/1940-6207.CAPR-11-0133.

Tomiyama, A. Janet. 2014. “Weight Stigma Is Stressful. A Review of Evidence for the Cyclic Obesity/Weight-Based Stigma Model.” Appetite 82:8–15. doi:10.1016/j.appet.2014.06.108.

Tomiyama, A. Janet, Britt Ahlstrom, and Traci Mann. 2013. “Long-Term Effects of Dieting: Is Weight Loss Related to Health?: Weight-loss Diets and Health.” Social and Personality Psychology Compass 7(12):861–77. doi:10.1111/spc3.12076.

Troiano, R. P., E. A. Frongillo, J. Sobal, and D. A. Levitsky. 1996. “The Relationship Between Body Weight and Mortality: A Quantitative Analysis of Combined Information from Existing Studies.” International Journal of Obesity and Related Metabolic Disorders: Journal of the International Association for the Study of Obesity 20(1):63–75. https://www.ncbi.nlm.nih.gov/pubmed/8788324.

Vergnaud, A.-C., S. Bertrais, J.-M. Oppert, L. Maillard-Teyssier, P. Galan, S. Hercberg, and S. Czernichow. 2008. “Weight Fluctuations and Risk for Metabolic Syndrome in an Adult Cohort.” International Journal of Obesity 32(2, 2):315–21. doi:10.1038/sj.ijo.0803739.

Wardle, Jane, Susan Carnell, Claire Ma Haworth, and Robert Plomin. 2008. “Evidence for a Strong Genetic Influence on Childhood Adiposity Despite the Force of the Obesogenic Environment.” The American Journal of Clinical Nutrition 87(2):398–404. doi:10.1093/ajcn/87.2.398.

Wolf, Naomi. 2009. The Beauty Myth: How Images of Beauty Are Used Against Women. Harper Collins.