Sex Bias in Lyme Diagnosis

Home (Main Menu)

To: lyme-aid@egroups.com
Source: LTShepler@aol.com
Date: Tue, 13 Jun 2000
Subject: [Lyme-aid] PLEASE READ! Feder, Borrelia & SEX

Lynn T. Shepler (LTShepler@aol.com) wrote:

Friends/colleagues:

Here is a most provocative review. Please read! Fascinating disjunct findings in males versus females with persistent symptoms of Lyme disease unearthed from an otherwise dull Feder publication!!

A copy of an old article by Feder at UConn surfaced in my brief case today:

"Persistence of Serum Antibodies to Borrelia burgdorferi in Patients Treated for Lyme disease," Clin Inf Dis 1992;15:788-93.

The article contains an entire page of data for 32 patients (see p. 790 of the article) all of whom are said to meet the CDC surveillance criteria for Lyme disease -- 18 females, 14 males. All received "appropriate antibiotics." [sic] I started reworking the data while sitting in a restaurant and continued for about 5 hours at home.

These bozos missed the most interesting thing about their data!! It quickly becomes apparent that the findings are radically disjunct when the data is separated by the SEX of the subjects. Take a look.

Feder tells us there are 9 patients out of the 32 with persistent symptoms of Lyme disease (28% of the total). The data shows these are 5 females and 4 males.

For the five females with persistent symptoms all had Erythema Migrans (EM) and were followed an average of 19 months from the time of presentation. At presentation, only one had a positive IgM ELISA, and only one had a positive IgG titer (defined as 1:320 or greater). The other four all had IgG titers of 1:80 or less (undetectable) at the time of presentation. Average of the IgG titer for all five women was 314 at the time of presentation.

So, yea, you got it --- only TWO of the women would have been picked up by the CDC two-tiered scheme at the time of presentation. If not for the Erythema Migrans, 3/5 (60%) would have been missed! (Yea, teensy numbers and Yes it would be nice to repeat in a larger population)

Of the four males with persistent symptoms, only one had an Erythema Migrans (EM). None had a positive IgM at presentation, although the IgG titers at the time of presentation for the four males was respectively 1:5120, greater than 1:5120, greater than 1:5120, and 1:1280. [In order to create an average number, I artifically gave the value 5400 to anything reported as greater than 5120.] The average IgG titer for the four men was thus 4,300.

Contrast that with what they found in the women !!! (I did not do any tests of statistical significance, but there is a dramatic difference -- 4,300 versus 314!) Note that ALL of these 4 males would have been picked up on a Lyme ELISA according to these titers.

Then contrast this with the IgG titers found in the five women with persistent symptoms at T2 (less than 80, less than 80, less than 80, less than 80, 1:1280, respectively). [To do some calculations, I artificially gave the value of 50 to anything reported as less than 80. (T1 = time of presentation; T2 = followup.) The calculated average was thus 296.

For the men, the IgG titers reported at T2 -- 1:5120, greater than 5120, 1:650, 1:1280, respectively. Using my system (greater than 5120 = 5400), the average IgG titer for the guys with persistent symptoms was 3,110 (compared to 296 for the gals).

Looking at the IgG western blots done at T1 (presentation) and T2 (follow-up), the number of bands on tests from the 18 females female, was 4.56, contrasted with the average of 14 males at 6.14.

THINK ABOUT THE RAMIFICATIONS OF THIS!!!! ON THE AVERAGE, WOMEN WHO HAD DOCUMENTED Erythema Migrans AND WHO MET THE CDC CRITERIA FOR DIAGNOSIS (THE MOST STRINGENT CRITERIA USED!!!) DID NOT MEET THE DRESSLER CRITERIA !!! 4.56 BANDS DOES NOT MEET THE CUT OFF OF 5 POSITIVE BANDS!! AND, IF THEY DON'T MEET IT, WHO CAN???? THIS IS THE GROUP OF WOMEN YOU WOULD EXPECT TO DEMONSTRATE THE MOST VIGOROUS ANTIBODY TITERS !!!

WHEREAS, THE GROUP AVERAGE FOR THE NUMBER OF BANDS DEMONSTRATED ON THE IgG WESTERN BLOT FOR MALES WAS 6.14 --- SO YOU COULD ARGUE THE AVERAGE GROUP OF MALES WITH DOCUMENTED ACTIVE LYME DISEASE MADE THE CUT OFF FOR A POSITIVE IgG WESTERN BLOT, WHEREAS THE AVERAGE VALUE FOR THE NUMBER OF BANDS FOR THE FEMALE GROUP DID NOT!!!!

If these findings are replicated in other papers on antibody findings in Lyme disease, this is an excellent example of how sexist critiera, or criteria thought to be "neutral" can exclude women from the very definition of a disease. (note here all the implications for "chronic fatigue syndrome")

Now, let's look at what happens with the number of bands in the nine patients with persistent symptoms of Lyme disease when you separate out the data by sex. The average number of bands seen on the western blots of the GUYS jumps to 8.00 !!! The average number of bands seen on the western blots for the WOMEN increases slightly but only up to 4.80 (still, doesn't meet the 5 positive band cut off!!! tut tut tut )

Note also that the IgG titers for these 5 women with persistent symptoms does not even make the cut off for a positive IgG titer at presentation at an average value of 314 (the average titer for the guys was 17,200) !!!!! Do you see!!! a value of 314 versus 17,200!!!! (The cut off for a POSITIVE IgG TITER is 1:320. At an average value of 314, these women would be told their Lyme test is "negative" or "equivocal." Brushed off.

Lucky for them they had an EM (Erythema Migrans), that's all I can say -- otherwise they would be told they had "chronic fatigue syndrome" or some such bullshit). At time T2, the average of the women's titer fell to 296, and the average titer for the four guys fell to 12,440. (that's well within a value to be detected with a Lyme ELISA --- BUT NOT SO FOR THE WOMEN!)

Another interesting observation: it appears that women overall (the 18 female patients) produce, on the average, more bands in the IgM than males on the average, particularly during the EARLY phase of the disease. At T1, the average number of bands seen on IgM western blots was 1.78 (females), 1.07 (males). Hmmmmmm. Why?

For all female patients (n=18), the screening tests used for Lyme disease (either IgG OR IgM) at the time of presentation were positive in only 13 patients (72%); for males (n=14), these screening tests were positive in 13 patients (93%). When you look at the IgG western blots for both genders, for the 5 women Lyme patients missed by the screening tests, the number of positive bands reported back on their IgG western blots was, respectively, 4 bands, 4 bands, 6 bands, 4 bands, 4 bands; for the 1 man missed by the screening tests, his IgG western blot demonstrated 5 positive bands.

(Note: with the exception of the woman who had 6 positive bands, although these 4 women had documented Erythema Migrans, they would have been missed by TWO Lyme screening tests AND the IgG western blot - where interpretation is based on the Dressler criteria of 5 positive bands --- even though they had very significant reactivity to be found on their IgG Lyme western blots ... geez, these gals can't win .... think about it, though.... 4/18 female patients with a documented EM can't get to a positive test the way it's rigged up!!! -- that's 22% or over 1 out of 5...women with "real" Lyme disease! )

If you would apply these statistics to, say, a population of 2000 patients with CDC-surveillance criteria Lyme disease of mixed gender, the application of these screening tests would miss 278 female patients, 71 male patients. Thus, of the patients who are missed, 80% (278/349) are FEMALE.

Using my artificial technique described above (I'd like to publish this, so if anyone knows a better way to estimate this, let me know), the average IgG titers of the 18 women at the time of presentation was 1,008; the average IgG titer of the 14 men was 2,386 (THE AVERAGE LYME TITER FOR A GUY WAS GREATER THAN DOUBLE THE AVERAGE LYME TITER FOR A FEMALE). As mentioned above, the average number of bands on western blot was also increased for the males: 4.56 (females) and 6.14 (males).

Regarding the ages of the patients who reported persistent symptoms: the average age of the 5 women was 49 years old; males 61 years old. The overall average age of the 18 women was 55 yo, for the 14 men it was 52 years old. So the males who did poorly were from an older group; whereas the females who did poorly were slightly younger than the group average for females. One wonders whether this could relate to the hormonal status of the female patient and effects of hormonal status on immune function.

Of note, when you look at the test results for the 5 women with persistent symptoms (remember, all had documented Erythema Migrans), screening tests would have missed 3/5 of the women or 60% !!!! The findings for women are thus very different compared to what you see for males with persistent symptoms.

Here are the findings published for the 5 FEMALE SUBJECTS WITH PERSISTENT SYMPTOMS:

(T1 = time of presentation; T2 = followup.)

FEMALE SUBJECT #3 -- IgM ELISA neg at T1& T2; IgG titer of less than 80 at T1&T2; IgG western blot (T1) with bands 24, 41, 66, 75; IgG western blot (T2) with bands 25, 41, 66; IgM western blot (T1&T2) with bands 41, 60.

FEMALE SUBJECT #14 - IgM ELISA pos at T1, neg at T2; IgG titer of 1:80 at T1, <80 at T2; IgG western blot (T1) with bands 25, 41; IgG western blot (T2) with band 41; IgM western blot (T1) with bands 25, 66; IgM western blot (T2) with band 66.

FEMALE SUBJECT #16 - IgM ELISA neg at T1&T2; IgG titer of 1:80 at T1, <80 at T2; IgG western blot (T1&T2) with bands 18, 25, 31, 34, 41, 66; IgM western blot (T1&T2) with band 66.

FEMALE SUBJECT #18 - IgM ELISA neg at T1&T2; IgG titer of 1:80 at T1, <80 at T2; IgG western blot (T1&T2) with bands 25, 41, 60, 66; IgM western blot (T1) with bands 25, 41, 66: IgM western blot (T2) with bands 41, 66.

FEMALE SUBJECT #20 - IgM ELISA neg at T1&T2; IgG titer (T1&T2) of 1:1280; IgG western blot (T1&T2) with bands 18, 25, 31, 41, 60, 66, 75, 88; IgM western blot (T1&T2) with bands 25, 41, 66.

Here are the findings published for the 4 MALE SUBJECTS WITH PERSISTENT SYMPTOMS:

MALE SUBJECT #4 -- IgM ELISA neg at T1&T2; IgG titer of 1:5120 at T1&T2; IgG western blot (T1&T2) with bands 18, 25, 31, 34, 41, 60, 66, 75, 88; IgM western blot (T1&T2) with bands 60, 66.

MALE SUBJECT #8 - IgM ELISA neg at T1&T2; IgG titer of > 5120 at T1&T2; IgG western blot (T1&T2) with bands 18, 25, 31, 41, 60, 66, 75, 88; IgM western blot (T1&T2) with no bands.

MALE SUBJECT #17 - IgM ELISA neg at T1&T2; IgG titer of > 5120 at T1, and 1:640 at T2; IgG western blot (T1&T2) with bands 18, 25, 31, 41, 60, 66; IgM western blot (T1&T2) with band 66.

MALE SUBJECT #19 - IgM ELISA neg at T1&T2; IgG titer of 1:1280 at T1&T2; IgG western blot (T1&T2) with bands 18, 25, 31, 34, 41, 60, 66, 75, 88; IgM western blot (T1&T2) with band 66.

Note the average number of bands on IgG western blot for the 9 patients with persistent symptoms: for females, 4.80 (T1), 4.40 (T2); for males, 8.00 (T1), 8.00 (T2). The women never really make it over the 5 positive band criteria. The guys get the prize of being told they have a diagnosis!! Congratulations --- you have LYME DISEASE !!!! Well, girls, we don't know WHAT you got, but just beat it ! OK???!

COMMENTARY: NIH has had guidelines for analysis of data by gender in place since at least the early 1980's. This paper was published in 1992! Although this paper does not mention any PHS funds involved in generating this data, it is still shameful.

Although this is a paper by Feder, it is even more surprising to pick up publications by Steere and Sigal (i.e., rheumatologists) that MAKE NO ATTEMPT TO ANALYZE DATA BY GENDER because 'we know' that autoimmune diseases ... occur in much higher numbers in females. So, if you thought GENDER was a big factor in your area of specialty, it doesn't seem like it should be a radical leap to analyze the data by gender, but these guys don't get it.

Nowhere in Feder's publication does he note ANY of these findings that are SO striking. Not a peep. Feder concludes in the paper that "nine of the 32 patients had persistent or recurrent symptoms, and ELISA and immunoblot were not helpful in identifying these nine patients." Well, you and I dear Reader can share this delicious secret. The man is dead wrong!

While I did not belabor it here, the findings for the symptomatic patients separated by the sex of the subjects can be distinguished

(1) from the findings for the group of females without persistent symptoms and

(2) the group of males without persistent symptoms.

Note the average IgG titer at T1 for symptomatic female patients --- 314; females without persistent symptoms ---1275; for symptomatic males --- 17,200; males with no persistent symptoms --- 1620.

           Graph of IgG data by Gender at
         Presentation (T1) and Follow-up (T2)

M      | 
a  T1  |     NO                              SX
l      | 
e  T2  |    NO                       SX
       |
       |             "SX" means "has symptoms"
F      |             "NO" means "no symptoms"
e  T1  | SX  NO 
m      | 
a  T2  | SX NO
l      |
e      |
       ------------------------------------------
       0          5000         10,000      15,000 

                             IgG titer

At T2, symptomatic females --- 296; females without persistent symptoms --- 674; symptomatic males --- 12,440; males without persistent symptoms ---1005. The numbers don't speak for themselves, they SHOUT.

Women, of course, are accorded their due in the publication. Among the 4 male authors and 3 additional males mentioned in the acknowledgement section, there is one woman. Dr. Feder thanks her, "Alice Iacobucci who patiently prepared the manuscript."

WE NEED TO GET RID OF THESE BOZOS!!! Obviously, the data here has very signficant implications for chronic fatigue syndrome. As I have stated here before, chronic fatigue syndrome, for the most part, is simply seronegative Lyme disease coupled with the use of gender "neutral" criteria (or, alternatively, diagnostic criteria used to apply to men with a specific disease state) to define a disease state that PRESENTS DIFFERENTLY IN WOMEN !!!!

If anyone should like their own copy of this publication but can't get to a medical library to unearth it, please backchannel me with your mailing address and I will send you a personal copy.

Yours, Lynn T. Shepler, MD, JD

Postscript: EM = erythema migrans

Also, it does occur to me that it may not be possible to "average" titers and that there there may be a different way to undertake what I am calling "averages." If anyone has thoughts about this, don't hesitate to share them. In any case, no matter how this data ultimately gets refined, the significant differences based on gender are irrefutable.

Date: Wed, 14 Jun 2000 01:03:57 EDT

Friends/colleagues:

This post contains two parts:

(1) summary of ability, sorted by sex of the subject, of different screening tests to successfully pick up CDC-defined Lyme cases (data from Feder's paper);

(2) implications of clear findings of SEX BIAS to call HALT to the use of the CDC surveillance criteria for Lyme disease (logical goal).

SUMMARY OF SCREENING TEST RESULTS BY SEX OF SUBJECT (FASCINATING!!)

To summarize in a more succinct manner from my earlier post (i.e., see, PLEASE READ! Feder, Borrelia & SEX) here are some fascinating statistics. Keep in mind that the authors (Feder, Gerber, Luger, and Ryan --- an all guy line up, I should point out...yes, equal opportunity in medicine is bullshit...) tell us that ALL of the 32 patients "fulfilled the Lyme disease national surveillance case definition of the Centers for Disease Control." There are 18 females, 14 males. 72% of the females reported EMs, 57% of the males.

FOCUS IS THE SEX BIAS OF THE SCREENING TESTS FOR LYME DISEASE

IgG titer (positive = 1:320 or greater) missed 45% (8/18) of the females, 14% (2/14) of the males.

IgM ELISA missed 78% (14/18) of the females; (13/14) 93% of the males.

IgG western blot (positve = 5 positive bands) missed 56% (10/18) of the females, 21% (3/14) of the males.

IgM western blot (positive = 41 band, plus one among bands 18, 25, 31, 34) missed 78% (14/18) of the females, 93% (13/14) of the males.

IgG titer + IgM ELISA missed 28% (5/18) of the females, 7% (1/14) of the males.

IgG titer + IgG western blot missed 39% (7/18) of the females, 7% (1/14) of the males.

IgG titer + IgM ELISA + IgG western blot missed 22% (4/18) females, 0% males (0/14).

IgG titer + IgM ELISA + IgG western blot + IgM western blot missed 17% (3/18) females, 0% (0/14) of the males.

SO, 17% (3/18) OF ALL FEMALE PATIENTS WITH CDC SURVEILLANCE CRITERIA LYME DISEASE DID NOT HAVE A POSITIVE TEST ON ANY OF THE FOUR TESTS !!!! COMPARED TO 0% OF THE MALES !!!

IF ONLY AN IgG TITER + IgG WESTERN BLOT IS USED (NOT AN UNUSUAL REGIMEN IN DOCTORS' OFFICES), THIS MISSED 39% OF WOMEN (COMPARED TO 7% OF MEN) !!!!

IF A WOMAN WANDERS INTO A DOCTOR'S OFFICE WITH RECENT SX OF LYME (NO EM) AND HER DOCTOR ORDERS JUST A LYME IgG TITER, FEDER'S DATA ON WOMEN WITH DOCUMENTED CDC-DEFINED LYME DISEASE SUGGESTS THAT 45% WILL HAVE A NEGATIVE TEST!!! (COMPARED TO 14% OF THE GUYS). AND, TYPICALLY, THESE WOMEN ARE LUCKY THEY GET EVEN THAT. MANY DOCTORS WILL NOT EVEN TEST FOR IT.

FEDER'S DATA SUGGEST THAT EVEN IF A FEMALE PATIENT IS LUCKY ENOUGH TO WALK INTO A DOCTOR'S OFFICE WHO IS WILLING TO DRAW OUT ALL THE BIG GUNS AND ORDER EVERYTHING (AND RISK HAVING THE MEDICAL BOARD BREATH DOWN HIS/HER NECK FOR ORDERING EXCESSIVE AND INAPPROPRIATE TESTS), 17% OF FEMALE PATIENTS WITH BONA FIDE DISEASE MAY STILL NOT OBTAIN A "POSITIVE" TEST AND TREATMENT FOR THE DISEASE BECAUSE *ALL* OF THESE TESTS ARE SEX BIASED, PARTICULARLY BIASED AGAINST ACHIEVING POSITIVE TEST RESULTS IN THE FEMALE POPULATION WITH ACTIVE LYME DISEASE. (NOTE THAT *ALL* OF THESE TESTS SHOWED SIGNIFICANT DIFFERENCES BASED UPON SEX AS DEMONSTRATED BY FEDER'S DATA.)

SECOND PART: SEX BIAS IN LABORATORY TESTS AND THE CDC SURVEILLANCE CRITERIA

The destructiveness of the CDC surveillance criteria has been noted for years. For at least the past 10 years, the Senate Committee on Labor & Health has been calling on the CDC to change its criteria. Nothing happens.

Sex is incendiary. This might be enough to burn down the house. It's not as incendiary as charges of race bias, but it might just be enough. If other papers support interpretations of the data as can be ferretted out from Feder's old publication, this could be enough. Clean. Simple.

Reminds me of strategies for bringing down the death penalty. Proving the death penalty is arbitrarily enforced and used to kill innocent people is enough to halt state killings. DNA evidence shows innocent people are killed. That's enough. You don't have to get to the deeper, divisive questions. What? Even George Bush has called a moratorium?

Note that Feder's data suggests that the CDC criteria (as far as lab tests go) is not fixable. Even if you routinely ordered ALL 4 tests simultaneously, this regimen still missed 3/18 women with CDC-defined LD. This suggests that there can be NO lab criteria for the diagnosis of Lyme disease. Heck, we all knew that ...

If anyone has off-hand knowledge of other papers to re-analyze in this fashion, let me know. Thanks.

Lynn Shepler, MD, JD

Addendum:

Note that if the tests were constructed with more of an emphasis on findings in the IgM, males would be disadvantaged. In Feder's data, males were much less likely to test positive on the IgM western blot or to test positive on the IgM ELISA. If we lived in a world where you had to have a positive IgM western blot to get treated, few guys would make the grade.

Lynn T. Shepler (LTShepler@aol.com) wrote:

Lynn wrote:

... As pointed out to me by a scientist who works in this area, it is incorrect to think of titers as linear -- it is a logarithmic scale ...

Jack Smith writes:


           Graph of IgG data by Gender at
         Presentation (T1) and Follow-up (T2)
                   (Logarithmic)
       |
M      |    
a  T1  |                            NO       SX
l      |
e  T2  |                          NO SX
       |
       |              "SX" means "has symptoms"
       |              "NO" means "no symptoms"
F      |
e  T1  |                      SX        NO
m      | 
a  T2  |                       SX NO
l      |
e      |
      ------------------------------------------
       0        10        100       1000    10^4  

   IgG titer  (plotted logarithmically
               to the base 10)

I would strongly advise against plotting the data with log (IgG titer) as one of the coordinates. This suppresses the effect clearly evident in the untransformed data. It brings to mind D's remark about "Liars, Damned Liars, and Statiticians."

Jack Smith

NINE REASONS FOR FALSE NEGATIVE LYME DISEASE BLOOD TESTS

The Lyme Disease Foundation (LDF), in their brochure entitled "LDF Frequently Asked Questions About Lyme Disease" lists the following nine reasons for false negative Lyme disease test results.

1. Antibodies against Bb are present, but the laboratory is unable to detect them. [Borrelia burgdorferi (Bb) is the Lyme disease bacteria.]

2. Antibodies against Bb may NOT be present in detectable levels in a patient with Lyme disease because the patient is currently on, or has recently taken, antibiotics. The antibacterial effect of antibiotics can reduce the body's production of antibodies.

3. Antibodies against Bb may NOT be present in detectable levels in a patient with Lyme disease because the patient is currently on or has previously taken anti-inflammatory steroidal drugs These can suppress a person's immune system, thus reducing or preventing an antibody response.

4. Antibodies against Bb may NOT be present in detectable levels in a patient with Lyme disease because the patient's antibodies may be bound with the bacteria with not enough free antibodies available for testing.

[For this reason, some of the worst cases of Lyme disease test negative -- too much bacteria for the immune system to handle.]

5. Antibodies against Bb may NOT be present in detectable levels in a patient with Lyme disease because the patient could be immunosuppressed for a number of other reasons, and the immune system is not reacting to the bacteria.

6. Antibodies against Bb may NOT be present in detectable levels in a patient with Lyme disease because the bacteria has changed its makeup (antigenic shift) limiting recognition by the patient's immune system.

7. Antibodies against Bb may NOT be present in detectable levels in a patient with Lyme disease because the patient's immune response has not been stimulated to produce antibodies, i.e., the blood test is taken too soon after the tick-bite (8-6 weeks).

Please do not interpret this statement as implying that you should wait for a positive test to begin treatment.

8. Antibodies against Bb may NOT be present in detectable levels in a patient with Lyme disease because the laboratory has raised its cutoff too high.

9. Antibodies against Bb may NOT be present in detectable levels in a patient with Lyme disease because the patient is reacting to the Lyme bacteria, but is not producing the "right" bands to be considered positive.

Lyme Disease Foundation
1 Financial Plaza
Hartford, CT 06103
(860)525-2000
fax (860)525-TICK
Lyme Disease National Hotline (800)886-LYME
mailto:lymefnd@aol.com

Bug War -- Top -- Home -- What's New