This article is based on internet visitor statistics and this is a fairly accurate representation on who seeks information on the disease or gluten free products and who is mostly likely to be diagnosed with the disease, even though celiac disease does affect other races to some degree.
A previous Gluten Free Pages online research article looked at the average celiac age and sex who visit online celiac and gluten free sites in America. This was done because real world statistics on the typical celiac are scant, and there has been no analysis performed on the online visitors. It showed that there was an overwhelming bias towards online search by middle aged women. It also noted that the disease affects both sexes equally, so this bias was probably caused by shopping habits and primary care given duties in average households. The statistics for this research also showed that women make up about 60-65% of gluten free site visitors to the sites used in this analysis.
This article looks specifically at the profiles for age, race and income level of visitors to major celiac and gluten free sites. The raw data was sourced from quantcast and three celiac and three gluten free sites were chosen based on their high monthly visitor numbers. Noting that unless a site is of sufficient size and coded for quantcast, no statistics will be available to quantify their visitor demographics. The site demographics are from: celiac (dot com), celiac (dot org), celiaccentral (dot org), gluten (dot net), Glutenfree (dot com), Gluten (dot net).
The proportion (prop %) below indicates in absolute terms, the segments visitor proportion to the site, while a more significant value is often the index % column. This is because the Index compares how a segment (such as 3-12 year olds), visit these specific (gluten free or celiac site) compared to the average internet site. For example an index of 120 means the segment visits a GF site 20% more than they do the average internet site. Thus the index gives an indication of who in the population is more affected by the disease (searches more) and who is under-represented.
The three celiac and three gluten free sites show the weighted averages below.
CELIAC SITES
GF SITES
age
Prop %
Index %
Prop %
Index %
3-12
1
22
3
42
13-17
2
15
3
27
18-34
29
99
26
89
35-49
40
144
41
148
50+
28
117
27
115
Ethnicity
Prop %
Index %
Prop %
Index %
Caucasian
91
117
91
117
African American
3
37
2
27
Asian
1
28
1
23
Hispanic
4
44
5
60
Other
1
54
1
63
Affluence
Prop %
Index %
Prop %
Index %
$0-$30K
14
76
14
79
$30-60K
23
88
24
88
$60-100K
27
96
31
110
$100K+
36
129
31
113
TABLE SUMMARY
AGE
The proportion statistics show that for celiac sites that 97% of visitors are over 18 with the major segment being 35 to 49 (40%). The proportion for GF sites is similar with the 35-49 age segment representing 41%. Even more significant is how the statistics compares to the internet average. Only the older two age segments are over-represented: 35-49 age segment (Celiac 144%, GF 148%) and 50+ age (celiac 117%, GF 115%).
As celiac diagnosis methods are becoming better, it appears that more of the 35-49 age segment are being diagnosed and seeking information or just choosing gluten free as a healthy alternative. Long diagnosis periods of 10+ years, and the extremely low coeliac rate of diagnosis in the USA (5%) could be a major cause of the younger ages not be represented. Perhaps the older population also has some inbuilt immunity to acquiring the disease (similar to the relatively low levels of peanut allergy in the older age groups, compared to infants). Or maybe the older age segment just doesn’t use computers that much to search for their GF information.
Ethnicity
The preamble to this is that celiac disease gene is much more prominent in Europeans (Caucasians), While some African Americans have found to develop CD, Asian have a very low rate. Also note that as the US is predominantly ‘white’ and whites probably have a higher access level to the internet than other races, that they make up the major proportion of visitors to these kinds of sites. Wikipedia gives the following ethnicity proportions of American citizens: ‘White alone’ (75%), White Hispanic and Latino Americans (65%), Hispanic or Latino (15%), Black or African American alone (12%), Asian alone (5%).
Taking these population proportions into account, the ethnicity online profile shows that both proportions and index are VERY over represented by Caucasians. Caucasians proportion for both types of site are: (celiac sites 91%, GF sites 91%). The online statistics show that African Americans make up a proportion of total site visitors of about 2.5%, Asians 1% and Hispanics 4.5%. So allowing for mixed races, non white races (African American, Hispanic, Asian) make up about 20% of their expected proportion of visiting the sites.
So we know that the white race in absolute numbers is over represented for searching on these kinds of sites. Given that blacks and Asians tend to have less susceptibility to celiac disease because of their genes, then this might be understandable. However, the Caucasian INDEX for both celiac and GF sites is the same at 117%. This shows that of the sites that Caucasians visit on the net, that 17% more visit these kinds of sites. This is a small but significant over representation by whites.
In this current research we look at the wealth or annual income of individuals (predominantly) within the US. The weighted average values for the affluence segments show that the proportion of visitors to both celiac and gluten free sites is roughly averaged between the three highest income segments ($30-60K, $60-100K, $100K+) around 20 to 30%, with a much smaller proportion of visitors from the 0-$30K segment. In absolute terms, the lowest income level ‘$0 to $30K have a proportion of 14% of visitors to these kinds of sites. Proportion of visitors in the $30-60K income segment (celiac 23%, GF 24%), $60-100K proportion (celiac 27% GF sites 31%) and over $100K (celiac 36% GF sites 31%). Even more significant is the Internet indexes, The $0-$30K segment = 77.5% index, $30-60K = 88%, $60-100K = 103%, $100K+ = 121%.
This shows a definite bias towards celiac coming from the more affluent intent users as the index for the two poorest segments are 22.5% and 12% UNDER the average internet user income level, while the two HIGHER income segments are 3% and 21% over the average. This result supports the theory that more affluent people/ states/ countries have more resources to diagnose health issues and gain treatment.
CONCLUSIONS
While previous studies have suggested a middle aged female bias in diagnosis, within a race celiac disease is believed to affect all members of a race equally. In the past, age profiles have suggested that older people are more often diagnosed, this has usually been because stresses that may bring out gluten intolerance or celiac disease may not happen until later in life, and diagnosis is often delayed up to ten years on average.
Regarding race, the proportion of Caucasians is very much higher than the proportion of Caucasians in America (75% proportion of population, and 91% of gluten free and celiac site visitors). While this can be somewhat explained by both celiac disease affecting the white race much more than the other segments, as well as Caucasians on average being more affluent, and so having access to better healthcare diagnostics. Mixed races may also be reporting their ancestory as white (eg 'white hispanic' races may record as white on the internet demoagraphics).
A significant bias of 25% LESS visitors to these sites for the lowest income group and 20% more for the highest income group supports the theory that affluence leads to high rates of diagnosis. As wheat and other gluten grains are the ingredients of snack food, fast food and high quality restaurant foods it is unlikely that the diets of the different affluence segments are a major contributor to creating celiac disease, its just the wealth factor allowing access to better health treatment.
How the statistics were created
Each demographic (age, race, affluence) is segmented into 4 or five segments by quantcast (e.g. wealth groupings are ( $0-$30K, $30-60K, $60-100K, $100K+ ). Each one of these groups records the proportion that the group comprises of the total visitors to the site AND an Index value. The proportion is obviously relative to the demographic representation in the country (eg the US is mainly white, so the proportion of white visitors would be expected to be the largest). However a much more useful statistic is the INDEX value. The index compares the visitor proportion to that expected by the average internet visitor. Thus and index of 100 would mean that just as many people from a given grouping visit other sites, as they do this type of site (gluten free or celiac), however an index of 150 means that the segment is 50% over-represented compared to the average internet user, for that segment.
The raw results for the three ‘celiac’ sites and three ‘gluten free’ sites show that regardless of size, that they tend to have very similar visitor segment demographics. This means that there is no bias on what type of gluten free associated site is chosen.
