Study Finds Direct Association Between Cardiovascular Disease & Periodontal Bacteria

Fullosseousflap

Wet-Gloved Dentist
10+ Year Member
7+ Year Member
Researchers report this week that older adults who have higher proportions of four periodontal-disease-causingbacteria inhabiting their mouths also tend to have thicker carotid arteries, a strong predictor of stroke and heart attack.

Read about it here.
 

ajmacgregor

Senior Member
10+ Year Member
15+ Year Member
May 24, 2003
122
4
40
new york
Visit site
Status
Okay, I'll be the first one to take the bait...

I think we need to see the actual abstract and obtain more information from the study authors regarding their analytic methods and results before we can jump to any conclusions.

1. Nowhere in the abstract does it state whether the authors controlled for the effects of potential confounders such as smoking, LDL levels, physical activity levels, previous CVD, family history of CVD, diet, body mass index, or socioeconomic status. All of these factors could be associated with intimal thickening and C-reactive protein levels, independent of periodontal disease status.

2. They report that there was an association between the number of periodontitis-inducing bacteria and the C-reactive protein level. However, they fail to report what measure of association and a p-value (who knows if it's statistically significant). For example, what if the measure of association was an Odds Ratio, and individuals with high levels of bacteria had 1.5 times the odds of developing CVD than those with low levels of bacteria. This is all well and good, but what if the 95% CI was 0.7 - 2.3. Since the confidence interval contains the null value of 1, it's not statistically significant.

3. There is a definite selection bias in that they have chosen a population (mean age = 69 years) that already has an elevated risk for CVD. Since CVD takes years to develop, how can one be sure that the present CVD was caused by long-standing periodontitis? Also - they state that they will follow the patients for three-years to evaluate the progression of disease. How will they account for all of the confounders present in #1?

4. C-reactive protein is used as a surrogate outcome for CVD in this study. CRP is elevated in acute inflammatory processes, including periodontitis. Therefore, the CRP levels here may just be the baseline measurements, and we will need to see the 3-year follow up data. Also - given the selection to the bias, there's potential for regression to the mean, in that the population that was selected for the study is very likely to have higher CRP levels than the general population, on average.

I don't mean to be so nitpicky about these things. I just think we should all take a step back and not jump to any conclusions without critically analyzing the study.
 

12YearOldKid

10+ Year Member
5+ Year Member
Oct 31, 2004
775
3
Status
Dentist
I definitely see your point, aj. That was exactly my thought when I first heard of this connection a few years ago. The trouble is that there is no real way to eliminate the selection bias. CVD is one of the most common diseases in the US - as is periodontal disease. And both increase with age. You would find similar correlation between wrinkles and menopause, or baldness and prostate cancer, or a million other obviously non-causal relationships.

I think it is an interesting idea that should be studied, but it is going to take a MASSIVE study to get some really solid evidence of this link. And even when we do find a correlation in a case like this it will be almost impossible to prove causation. After all, we don't even have a definitive study proving that bacteria are responsible for human caries. It's going to be quite a few years yet before the cardiologist starts making referrals to his friendly, neighborhood periodontist.
 
OP
F

Fullosseousflap

Wet-Gloved Dentist
10+ Year Member
7+ Year Member
12YearOldKid said:
I definitely see your point, aj. That was exactly my thought when I first heard of this connection a few years ago. The trouble is that there is no real way to eliminate the selection bias. CVD is one of the most common diseases in the US - as is periodontal disease. And both increase with age. You would find similar correlation between wrinkles and menopause, or baldness and prostate cancer, or a million other obviously non-causal relationships.

I think it is an interesting idea that should be studied, but it is going to take a MASSIVE study to get some really solid evidence of this link. And even when we do find a correlation in a case like this it will be almost impossible to prove causation. After all, we don't even have a definitive study proving that bacteria are responsible for human caries. It's going to be quite a few years yet before the cardiologist starts making referrals to his friendly, neighborhood periodontist.
There are a number of studies which discuss the role of bacteria and caries.

Would you like some citations? Or you can go to PubMed and look up John Featherstone at UCSF Dental School. There were some great papers about a year ago or so in the California Dental Association Journal.

With regards to CVD, this paper was just presented and I will link to the full paper when it is available. :)
 

tx oms

Welcome to Thunderdome
10+ Year Member
5+ Year Member
Oct 24, 2004
694
2
40
USA
Status
Fullosseousflap said:
There are a number of studies which discuss the role of bacteria and caries.

Would you like some citations? Or you can go to PubMed and look up John Featherstone at UCSF Dental School. There were some great papers about a year ago or so in the California Dental Association Journal.

With regards to CVD, this paper was just presented and I will link to the full paper when it is available. :)
Flap, the guys that wrote the Circulation article published the same study in about '03, if memory serves me. Search for the authors' names on PubMed and you'll see what I mean. The latest article just rehashes what they already published.

I read the '03 article and wasn't that impressed, as I recall. The authors claimed to control for the variables mentioned in the above posts but I don't think there control was really believable. There are only about three groups of authors I can find that are heavily interested in this topic; therefore, the articles on this topic seem pretty incestuous to me. Plus, each group defines perio disease differently, a fact they all mention in their discussions.

My other gripe is that they claim there is a link to CVD because they can find a relationship between perio disease and increased carotid artery intima thickness. I'm not sure that carotid intimal thickness is an independent risk factor for MI/CVA. Maybe it is. Finally, the way that this study defines perio disease is to correlate the number of missing teeth to severity and chronicity of disease. I guess this has been proven in previous population studies, but I don't know.
 
OP
F

Fullosseousflap

Wet-Gloved Dentist
10+ Year Member
7+ Year Member
tx oms said:
Flap, the guys that wrote the Circulation article published the same study in about '03, if memory serves me. Search for the authors' names on PubMed and you'll see what I mean. The latest article just rehashes what they already published.

I read the '03 article and wasn't that impressed, as I recall. The authors claimed to control for the variables mentioned in the above posts but I don't think there control was really believable. There are only about three groups of authors I can find that are heavily interested in this topic; therefore, the articles on this topic seem pretty incestuous to me. Plus, each group defines perio disease differently, a fact they all mention in their discussions.

My other gripe is that they claim there is a link to CVD because they can find a relationship between perio disease and increased carotid artery intima thickness. I'm not sure that carotid intimal thickness is an independent risk factor for MI/CVA. Maybe it is. Finally, the way that this study defines perio disease is to correlate the number of missing teeth to severity and chronicity of disease. I guess this has been proven in previous population studies, but I don't know.
Yes, you are correct.

The authors, just presented, I believe yesterday at the IADR and am awaiting the reports back form the meeting.

Did anyone go? Have the full paper?

I would be interested to see if there was new material/research incorporated in their presentation or just a reiteration.
 

tx oms

Welcome to Thunderdome
10+ Year Member
5+ Year Member
Oct 24, 2004
694
2
40
USA
Status
My library is suppose to send me the full article electronically. I'll post it when I get it. Bottom line: it will be difficult to impossible to prove that perio disease is an independent risk factor for CV disease b/c the overwhelming majority of people with perio disease have about 12,249 other risk factors for heart disease.
 
OP
F

Fullosseousflap

Wet-Gloved Dentist
10+ Year Member
7+ Year Member
tx oms said:
My library is suppose to send me the full article electronically. I'll post it when I get it. Bottom line: it will be difficult to impossible to prove that perio disease is an independent risk factor for CV disease b/c the overwhelming majority of people with perio disease have about 12,249 other risk factors for heart disease.
Great and Thanks... will look forward to it.

Yes, agreed....indeed at least 12,249!
 

12YearOldKid

10+ Year Member
5+ Year Member
Oct 31, 2004
775
3
Status
Dentist
Fullosseousflap said:
There are a number of studies which discuss the role of bacteria and caries.

Would you like some citations? Or you can go to PubMed and look up John Featherstone at UCSF Dental School. There were some great papers about a year ago or so in the California Dental Association Journal.

With regards to CVD, this paper was just presented and I will link to the full paper when it is available. :)
Of course. I wasn't disputing the link between caries and bacteria. I was just pointing out how even something that is apparently simple can really frustrate researchers. Everybody has caries and everybody has the bacteria that cause it. Any fool with an agar plate can demonstrate a correlation there, but to prove causation is a lot more difficult. We still haven't done the definitive study which would be to raise a control group of biologically "sterile" humans and see if they develop decay. But obviously, that experiment can't be performed (even though it has been done in animals). All the other human studies out there just are just clever ways of skirting around the issue.

And the link between CVD and perio is going to be even harder to tease out because CVD is multifactorial (I have it from a good source that there are at least 12,249 contributing factors. ;) ) Still very interesting though.
 

rahmed

SDN Angel
10+ Year Member
7+ Year Member
Sep 10, 2004
732
0
Rochester, NY
Status
Dentist
12YearOldKid said:
Of course. I wasn't disputing the link between caries and bacteria. I was just pointing out how even something that is apparently simple can really frustrate researchers. Everybody has caries and everybody has the bacteria that cause it. Any fool with an agar plate can demonstrate a correlation there, but to prove causation is a lot more difficult. We still haven't done the definitive study which would be to raise a control group of biologically "sterile" humans and see if they develop decay. But obviously, that experiment can't be performed (even though it has been done in animals). All the other human studies out there just are just clever ways of skirting around the issue.

And the link between CVD and perio is going to be even harder to tease out because CVD is multifactorial (I have it from a good source that there are at least 12,249 contributing factors. ;) ) Still very interesting though.
To come to a definitive conclusion between Bacteria and Caries in human population one has to repeat something similar to Vipeholm Study which would be impossible and unethical to say the least in these day and age. But I am digracing here. I think we all are hearing about this alleged correlation study for last few years and I still don't understand how these groups of researchers are finding sustained funding! Aren't there more important and genuine issues crying for fund?

rahmed
 

tx oms

Welcome to Thunderdome
10+ Year Member
5+ Year Member
Oct 24, 2004
694
2
40
USA
Status
Is anyone aware of exactly how independent risk factors of CVD are confirmed? I know many of them are based on studies like Framingham, but how exactly are the risk factors confirmed? How do you jump from a correlation to a definant risk factor?

Like I said above, people with perio have lots of other risk factors so how can you be sure the perio is truly an independent risk factor, but for that matter how is high cholesterol confirmed as a risk factor? I mean, can you find anyone with high cholesterol who has no other risk factors? Probably not. So how do you absolutely control for the other risk factors? Does the data show that people with "x" risk factors plus high cholesterol have a higher risk of CVD than those with "x" factors alone; therefore, high cholesterol is an independent risk factor?
 

rocknightmare

Senior Member
7+ Year Member
15+ Year Member
Apr 13, 2004
1,124
1
Status
Dental Student
Hell I can associate that everytime there is a fire there is going to be a fire truck nearby within minutes. The question that needs to be raised is not if there is an association but why is there an association.