2 - Elise Paradis, Ph.D.

Th e O besit y R esearch Paradigm:
An Analysis of the 100 Most-cited Articles
Bilal Al-Khashalee MD and Elise Paradis1 PhD
1) Department of Anesthesia, Postgraduate Medical Education, and The Wilson Centre, University of Toronto Faculty of Medicine.
INTRODUCTON
The literature on obesity has been growing at an
exponential pace (Figure 1). Previous research criticized
obesity scholarship for being sensationalist and
1,2
misrepresenting epidemiological data. Comprehensive
reviews of the research on obesity have had mixed results.
While some suggest that obesity exerts a massive morbidity
3
and mortality toll , recent epidemiological studies
4
question these results. Our aim was to trace a portrait of
decades of obesity research and document its paradigmatic
findings. Given the size of this literature (~1 million articles
since 1880), in this exploratory study we focused on the
100 most-cited articles on the topic. What can we learn
about obesity from these articles?
R E S U LT S
Figure 1: Obesity Articles in PubMed Database, as Count and
Percentage,1960-2014
Figure 3: Type of Study in the Sample (n = 89)
Table 1: Co-morbidities and physiological impact of obesity
discussed in sample articles
METHODS
Co-morbidities
We downloaded metadata on the 100 most-cited articles on obesity as a topic found in the Web of Science Database,
and downloaded all articles in PDF form. We checked that all were still being cited today. Two of the 100 could
not be located, for a final sample size of n = 98. Articles were coded and analyzed using 59 categories that were
Figure 2: Extent to which obesity is covered in 100 most-cited
articles, Web of Science
Using our 5-pt scale, we found that only 22 articles are primarily about obesity. The 9 articles (9%) that are not about obesity were excluded from analysis;
2 more could not be located, for a final sample of 89 (Figure 2).
developed iteratively. We used pre-existing literature
as a guide, and grew the coding scheme as we read
through the articles. The coding scheme included:
type of study, population studied, etiology, comorbidities, physiological impact, known mitigators
of obesity, and whether the article discussed obesityrelated mortality. A 5-point scale was developed to
rate the extent to which obesity was discussed in the
article (Figure 2). Authors coded all 98 articles using
that 5-pt scale and resolved disagreements. Data
were tabulated and mapped over time.
Nervous
Cardiovascular
Renal
Respiratory
Cancers
Reproductive
Metabolic (total)
Diabetes
Metabolic syndrome
Mental health
GastroIntestinal
Total
n
8
48
6
8
14
0
58
54
13
2
4
68
%
9
54
7
9
16
0
65
61
15
2
4
76
Physiological Impact
n
11
14
2
3
1
3
43
30
9
2
3
54
%
12
16
2
3
1
3
48
34
10
2
3
61
Most studies in the sample are epidemiological (Figure 3). There were twice as many
studies on animal models as humans (36 vs. 18). One study combined lab techniques
and epidemiological methods. About 40% of articles documented the etiology of
obesity, which is mostly attributed to genetic and behavioral causes. The most frequent
co-morbidities and physiological impacts listed were metabolic (diabetes and
metabolic syndrome) and cardiovascular. One quarter of articles mentioned
mitigators such as stress, weight cycling, cardiorespiratory fitness, or poverty. The
impact of obesity on the healthcare system is also mentioned. Mortality was explored
in depth only in 7 articles, but mentioned in 39 others. Proposed solutions included
diet, exercise and medications, with rare mentions of surgery. Other solutions such
as education and psychotherapy were not mentioned (Table 1).
Reviews in our sample were not centrally about obesity, but rather about metabolic
syndrome, inflammation, adipose tissue as endocrine system, and HDL. The only
exception positioned obesity as a global epidemic, and discussed the etiology of
obesity as a disease, several comorbidities, its physiological impact, but also noted
3
the difficulty of ascertaining its impact on mortality.
CONCLUSION
A large proportion of analyzed articles are not primarily about obesity, but rather investigate the association between obesity and other diseases. From
epidemiological studies we learn that the association between body mass index and mortality / morbidity is more complex than the public health discourse
(fat = bad) would imply. Twice as many studies were about animal models as were about humans, suggesting that we still know very little about obesity.
REFERENCES
1) Bacon L and L Aphramor. 2011. Weight Science: Evaluating the Evidence for a Paradigm Shift. Nutrition Journal, 10:9. 2) Campos P, Saguy A, Ernsberger P, Oliver E, Gaesser G. 2006. The epidemiology of overweight and obesity: public health crisis or moral panic? International Journal of Epidemiology. 35(1):55–60.
3) Kopelman PG. 2000. Obesity as a medical problem. Nature, 404(6778):635-43. 4) Flegal KM, Kit BK, Orpana H, Graubard BI. 2013. Association of all-cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta-analysis. JAMA. 309(1):71-82