A review of the sparse data
Dr David White, retired General Practitioner
Comorbidity and death from or following Covid-19 in the Pfizer/BNT162b2 trial
A review of the sparse data
Dr David White, retired General Practitioner
HART has recently covered a detailed analysis of all-cause mortality in the Pfizer BioNTech trial. This second article takes a deeper look at the comorbidities of those who died.
Obesity stood out as the greatest danger in the Pfizer/BioNTech Covid-19 vaccine trial. Of the seven people who died after catching the virus, five were obese or severely obese, with an average BMI of 41. Three were classed as severely obese. The risk of dying with covid was estimated to be about five times higher for the obese and more than ten times higher for the severely obese compared with those of healthy weight. The pattern matches warnings from the WHO and CDC that excess weight greatly increases the danger from Covid-19. Every person who died also had other health problems such as high blood pressure, diabetes, heart or lung disease, or HIV and all were aged 56 or over. The importance of risk factors has been massively downplayed in official messaging about everyone being at risk and an opportunity to address long term health improvement was wasted.
The NEJM pivotal paper reports 3 deaths due to Covid-19 in the trial. Despite enrolling more than 44,000 participants during a pandemic which locked down the world, there is a sparsity of mortality data from randomised trials. However in addition to those Covid-19 attributed deaths: 2 participants with cause of death ‘cardiac arrest’ died at 3 and 22 days after the start of a Covid-19 infection; another participant died of Covid-19 after unblinding and one participant died of Covid-19 but was not part of the safety analysis population due to being HIV positive. This review refers to these 7 participants.
The WHO, CDC, NICE and others have listed risk factors for severe Covid-19 illness, some of which are shown in Table 1 below.
| Table 1. Some of the risk factors for severe Covid-19 | |
| WHO | Obesity, quoting a French study which found the odds of requiring mechanical ventilation in Covid-19 patients admitted to intensive care were 7 times higher in those with a BMI >35 compared with those <25.Smoking, heavy alcohol consumption, physical inactivity, diabetes, cardiovascular disease, chronic obstructive pulmonary disease (COPD), cancer. |
| CDC | Age over 65, 81% of Covid-19 deaths are in people over the age of 65.The list of 21 risk factors includes: cancer, cardiovascular disease, chronic lung disease, dementia, diabetes, HIV infection, physical inactivity, smoking,overweight (BMI ≥25 kg/m2), obesity (BMI ≥30 kg/m2 or >95th percentile in children), severe obesity (BMI ≥40 kg/m2)References 19 studies including 3 meta-analyses ‘Obesity may triple the risk of hospitalisation due to Covid-19.’‘… risks of hospitalization, intensive care unit admission, invasive mechanical ventilation, and death are higher with increasing BMI.’ |
| NICE | A list of ‘Risk factors for progression to severe COVID‑19 in adults defined by the independent advisory group commissioned by the Department of Health and Social Care (June 2023)’ is included in a review of medications intended to prevent such progression, but NICE state it ‘has not been independently reviewed or verified by NICE’. The table lists more than 60 conditions including COPD, pulmonary fibrosis, HIV, dementia but does not include obesity or severe obesity.The list is based on the 2021 IAG report requested by Professor Jonathan Van Tam (the former Deputy Chief Medical Officer). The independent advisory group was constituted to identify a set of patient conditions (or cohorts) that were deemed to be at the very highest risk of an adverse COVID outcome, namely hospitalisation and death. But the report said ‘Parameters including age and obesity remained out of our remit and are not included in our recommendations though both remain significant risk factors in all data sets.’‘The QCOVID risk tool programme …originally significantly informed the work of the IAG’. A database comprising 1205 general practices and over 6 million patients in England was used to develop a Living risk prediction algorithm (QCOVID) for risk of hospital admission and mortality from coronavirus 19 in adults, which incorporates how risk increases with each additional risk factor. |
Now that the approximately half million pages of anonymised data Pfizer shared with the FDA in the licence application have been made public at the PHMPT website and in a more searchable form at the ICAN website, trial participants’ narrative reports and clinical report forms CRFs can be reviewed. This is explained in my films about the excess cardiovascular deaths in the trial and how the Covid-19 deaths were represented in the licence application. Reviewing these 7 individual cases gives an indication of which groups were at increased risk of death following infection in the trial and how risk factors interacted. The findings contradict the statements, the virus does not discriminate or that everyone was at risk. In the trial, all deaths from or following Covid-19 occurred in those with comorbidities and risk factors for severe disease.
Results
Table 2 below lists causes of death, BMI, comorbidities, risk factors and some other conditions. The first 5 participants died in the blinded, placebo-controlled part of the trial in the safety population and are therefore included in NEJM table S4. The 6th participant (12291083) was one of the 200 HIV positive participants and so not part of the safety population described in the NEJM paper. The 7th participant (10271191) died after unblinding (and declining vaccination) 6 days before acute respiratory failure was diagnosed, a sequence of events that clearly illustrates how the healthy vaccinee effect biases observational studies. These last two participants do not appear in the NEJM paper, but are shown in the licence application in table 32.
| Table 2. BMI, risk factors for severe Covid-19 and comorbidities in 7 participants who died of or following Covid-19 in the Pfizer/BNT162b2 trial. Cause of death is as shown in NEJM paper table S4 for the first 5 participants. Participant age is between 56 and 68 unless stated, 4 were male. | |||||
| Participant number | Cause of death | mRNA Covid-19 vaccine received | Days from start of Covid-19 to death | BM kg/m2 | Risk factors for severe Covid-19 infection Comorbidities Other information |
| 12521010 | Covid-19 pneumonia # | Pfizer/BNT162b2 | 1 | 29.2 | Age 81 Congestive heart failure Atrial fibrillation Cerebrovascular accident Hypertension Hyperlipidaemia |
| 11401117 | Cardiac arrest | Pfizer/BNT162b2 | 22 | 38.1 | Coronary artery disease Hypertension Hypercholesterolaemia Hyperglycaemia Taking vitamin D Obesity |
| 10881126 | Cardiac arrest | Placebo | 3 | 40.7 | Coronary artery disease Hypertension Hypercholesterolaemia Severe obesity |
| 10941112 | Acute respiratory failure Pneumonia Covid-19 # | Placebo | 23 | 34.2 | Type II diabetes Asthma Hypercholesterolaemia Vitamin D deficiency Obesity |
| 3 participants were not eligible to be included in the NEJM paper. The next participant was included but should have been removed from the study because of receipt of an out of study Covid-19 vaccine | |||||
| 10841470 | Multiple organ dysfunction syndrome Covid-19 # | Placebo and Moderna | 12 | 41.2 | Pulmonary fibrosis Hypertension Hyperlipidaemia Severe obesity |
| In addition 1 participant died after unblinding, 1 participant was HIV positive so not in the safety population. As neither are listed in NEJM table S4, cause of death is shown as reported in licence application table 32. | |||||
| 12291083 | Covid-19 pneumonia | placebo | 3 | 51 | Asthma Hypertension HIV infection Type II diabetes diagnosed 8 days before death Severe obesity |
| 10271191 | Respiratory failure Covid-19 | placebo | 31 | 21.8 | Lung neoplasm Emphysema Hypertension Hyperlipidaemia |
# cause of death reported as Covid-19 in NEJM table S4
Cause of death was attributed to Covid-19 in 5 cases and cardiac arrest in 2. Of these 2, Covid-19 was a contributory cause on the death certificate of participant 10881126 (but was not mentioned in the NEJM table as shown above). For participant 11401117, the only COD given is cardiac arrest and the death certificate entry is not specified. In this case, Covid-19 was diagnosed on the basis of ‘sniffles’ and a positive blood test for Covid-19 antibodies.
One placebo group participant (10841470) received Moderna vaccine during the trial but was not removed from the analysis, despite 388 others being disqualified for receiving out-of-trial vaccines. Had this participant been appropriately excluded, there would have been 2 Covid-19 deaths in each group during the blinded period. Steve Kirsch talks about it on X here. Excluding this participant would reduce the total number of deaths in the placebo group to 13 and leave the number in the vaccine group at 15.
A recent HART bulletin identified a third potential Covid-19 death in the vaccine group as participant 10891073 who died from an exacerbation of COPD chronic obstructive pulmonary disease, but no Covid-19 test result was available despite the test being required by the protocol. This participant is not included in the table above as there was no positive Covid-19 test result, but was also severely obese with a BMI of 40.1.
Risk factors
All deaths from or following Covid-19 were in participants aged 56 or older.
The high disease burden in those that died of or following Covid-19 is striking. All had comorbidities. Two had coronary artery disease, 1 had cerebrovascular disease, 2 had chronic lung disease – 1 of which also had a history of lung neoplasm, 1 was HIV positive and the final participant had type II diabetes.
Apart from the participant who was 81 years of age and the participant with a history of lung neoplasm, all 5 other participants were obese or severely obese and had a BMI above 34 kg/m2. The average BMI in these 5 participants was 41 with 3 having a BMI over 40. The highest BMI was 51. In the trial 34.4% of the almost 44,000 participants had a BMI of 30 or above.
Six of the seven participants who died were hypertensive. Of the 5 with medication history listed, 1 was taking an angiotensin-converting enzyme (ACE) inhibitor and 2 were taking an angiotensin-II receptor antagonist.
All of those who died were stated to have hyperlipidaemia or hypercholesterolaemia apart from the participant with a BMI of 51. Medication history is only given for 4 of these 6 in the narrative report and all 4 were taking a statin.
Three participant narratives reference vitamin D. Participant 10941112 had been diagnosed with vit D deficiency a year before death but pre admission medication is not listed and no mention is made of giving vit D in hospital, despite remdesivir, steroids, zinc, vitamin C, convalescent plasma and ventilation being listed. Participant 11401117 had been taking vit D at an unknown dose for the previous 5 years for general health. His narrative report describes his COVID-19 infection as: ‘”sniffles” which was not considered to be serious’. Participant 10881126 was advised to consider taking vit D and C by his primary care physician. HART has previously reviewed the importance of vitamin D and found 17 studies confirmed the risk of death from Covid-19 was higher in those with vitamin D deficiency and a study of Vitamin D prophylaxis, in Mexican healthcare workers, demonstrated a highly statistically significant 78% reduction in becoming infected.
Regarding ethnicity: 3 participants were ‘White Non-Hispanic/non-Latino’; 2 ‘Hispanic/Latino’; and 2 ‘Black or African American’. The 3 largest ethnic groups or ethnicities reported in the demographic characteristics of participants at baseline, table 1, were: 82% ‘White’; 25.9% as ‘Hispanic or Latino’; and 9.6% as ‘Black or African American’.
The under representation in the trial of the group at highest risk by age, the elderly, is of concern. A meta analysis by Levin et al found:
‘The estimated age-specific IFR is very low for children and younger adults (e.g., 0.002% at age 10 and 0.01% at age 25) but increases progressively to 0.4% at age 55, 1.4% at age 65, 4.6% at age 75, and 15% at age 85.’
By these estimates the risk of death was 7,500 times higher in 85 year olds than in 10 year olds.
Only 4% of participants were over the age of 75 according to table 3 of the NEJM paper. In September 2020 a protocol amendment stated: ‘removed the upper age limit for inclusion in Phase 2/3 in order to evaluate … those over 85 years of age’. I have been unable to establish the number of participants over the age of 80 in the trial but 1 participant with cardiovascular disease died of Covid-19 at age 81.
The importance of obesity and severe obesity as a risk factor for death following infection stands out. The proportion of obese and severely obese participants in the trial was 34.4% whereas 71% of participants that died of or following Covid-19 were obese or severely obese. Obese participants (BMI ≥30 kg/m2) had about 4.8 × higher risk of death from or following Covid-19 (CIs 0.9–24.7). The excess of Covid-related deaths among obese participants (5 of 7) was borderline but not conventionally statistically significant (exact Fisher p-value ≈ 0.052).
Because the NEJM paper does not state how many participants were severely obese (BMI of 40 or more) the relative risk for this group can’t be directly calculated. Those with a BMI over 30 are slightly underrepresented in the trial at 34.4% compared to the US rate of 41.9%. If we assume those with severe obesity are underrepresented by a similar proportion (0.82), then using US population BMI figures of 9.2% severely obese, gives an estimated 3312 severely obese participants in the trial. While the exact number is uncertain, it provides a reasonable basis for calculating relative risk. This would equate to an estimated risk ratio (a measure of how much more likely an event is in one group compared to another) for the severely obese (BMI≥40kg/m²) as compared to the not obese (BMI<30) of 13 (95%CI 2.17, 78.5) albeit with wide CIs due to small event numbers, with a p-value of 0.0094, using a two-tailed Fisher’s exact test, statistically significant even with these small numbers. As the number of deaths is small the odds ratio (a similar measure that compares the odds of an event rather than the direct risk) is similar at 13.04 with similar CIs.
By way of comparison a review paper by Martono et al gives odds ratios for coronary artery disease/cardiovascular disease of 1.85-5.19, COPD of 2.01, diabetes of 1.82-3.68 and age over 65 of 6.06, meaning severe obesity could be one of the most important risk factors for death from Covid-19. It is surprising that this paper lists and gives odds ratios for many risk factors but not for obesity or severe obesity and does not even mention either condition.
In addition, 4 of the 31 non Covid-19 deaths were in severely obese participants with an average BMI of 42.6. Those with severe obesity had 1.83 times the odds of a non Covid death compared to non-obese participants, OR 1.83 (95% CI: 0.66-5.09, p>0.29). Of these 4, one death was attributed to COPD as mentioned above, 2 were attributed to cardiac arrest and one to cardiopulmonary arrest. Only one had a post-mortem and the result is not known. Cardiac arrest was the sole cause of death on the death certificate of 1 and what was on the death certificates of the other 3 is not specifically stated.
Death was not attributed to severe obesity, even as a contributory cause, in any of the 7 severely obese participants who died in the trial, despite other conditions such as hypertension and diabetes being mentioned. In the trial there were 7 deaths from or following Covid-19, and 7 deaths in the severely obese. The former condition locked down the world, the latter condition is not even recorded on death certificates and therefore ignored in mortality statistics.
Mortality statistics for England and Wales are available at the nomis website. Underlying cause of death is shown by ICD-10 code, with E66 coding for obesity but no separate code for severe obesity. In 2013 the number of deaths with underlying cause of death obesity was 0.086 of total deaths and this rose to 0.123% by 2024. ICD-11 is not used for mortality data collection in the UK but it has introduced a code 5B81.2 for severe obesity (BMI≥40kg/m²). In the UK, obesity is rarely recorded as the underlying cause of death, even in severely obese patients.
Do these findings on obesity agree with published findings?
Broadly yes, despite the small numbers, but with an important exception shown below.
Public Health England states: ‘The evidence consistently suggests that people with Covid-19 who are living with overweight or obesity compared with those of a healthy weight, are at an increased risk of serious Covid-19 complications and death.’
383 hospitalised patients, admitted in January and February 2020 were included in a 2020 paper from Shenzhen, China, which found after adjusting for characteristics, comorbidities and treatment, men who were obese (BMI ≥28 kg/m2) versus those who were normal weight were at increased odds of developing severe Covid-19 (OR 5.66, 95% CI 1.80–17.75, P = 0.003).
In April 2020, a letter to the NEJM described how 35.8% of patients with Covid-19 admitted to intensive care in New York were obese (BMI ≥30 kg/m2) compared to 43.4% who required invasive mechanical ventilation.
A French study published in June 2020 found the odds of requiring mechanical ventilation in Covid-19 patients admitted to intensive care were 7 times higher in those with a BMI ≥30 kg/m2 compared with those <25.
The CDC states: ‘Obesity may triple the risk of hospitalisation due to Covid-19’ and
‘… risks of hospitalization, intensive care unit admission, invasive mechanical ventilation, and death are higher with increasing BMI.’
The UK Joint Committee on Vaccination and Immunisation: advice on priority groups for COVID-19 vaccination included morbid obesity in its main risk groups.
A systematic review showed: ‘Our results suggest that obese individuals are 1.5 times more likely to experience severe outcomes and 1.09 times more likely to die when compared to non-obese individuals with COVID-19 disease’.
A surprising discrepancy that requires explanation.
The BMJ QCOVID prediction algorithm paper, Copyright © 2021, Oxford University Innovation Limited, shows hazard ratios for many risk factors but BMI is dealt with as a continuous variable. The researchers used a more flexible mathematical approach to model how BMI affects the outcome. Instead of assuming that risk increases in a straight line as BMI goes up, they allowed the relationship to curve or change shape.This was done using a ‘fractional polynomial’ method which applied special calculations (in this case involving the square root and natural logarithm of BMI) to better capture how changes in BMI relate to risk.
To test how the calculator worked in practice, I conducted a simple experiment – I entered my details at my actual weight (normal BMI 20.4), then repeated the calculation after adding 30kg and 60kg. The results were unexpected. My risk of death within 90 days of testing positive for Covid-19 was: with a normal BMI (20.4) – risk 1 in 72; 30kg heavier and obese with a BMI of 31.2 – risk 1 in 188; 60 kg heavier and severely obese with a BMI of 42 – risk 1 in 118. Apparently, if I put on a lot of weight I could halve my risk of death. I consider myself healthy now but according to their algorithm, my risk of death would be considerably less if I was severely obese.
It is of note that obesity remained out of the remit of the UK government’s independent advisory group on the highest risk clinical subgroups and the IAG was ‘originally significantly informed’ by the QCOVID risk tool programme. In whose remit was obesity? How did the U.K. government inform itself about, plan, manage, and allocate resources for, what may have been one of the most important risk factors for Covid-19 severe illness and death, both in terms of elevated individual risk and high prevalence, rivalling extreme old age?
The findings of this review of deaths from or following Covid-19 support the findings of early observational studies and the conclusions of the WHO and CDC, that risk of death from Covid-19 increases with obesity and severe obesity and found obese participants (BMI ≥30 kg/m2) had about 4.8 × higher risk of death from or following Covid-19 (CIs 0.9–24.7) than the non obese in the trial.
A missed opportunity
In addition to obesity, the WHO and CDC listed physical inactivity as a risk factor for severe disease. Instead of locking gyms and closing parks and beaches and confining people to their homes for 23 hours a day, if some sort of intervention was thought essential, then a programme of national fitness, healthy eating, weight and vitamin D optimisation would at least have been less harmful than lockdowns and could have improved national physical and mental health and resilience for many years to come. This would have reduced pressure on the NHS, without the harm caused by emptying hospitals and withdrawing services, as the UK Covid-19 Inquiry recently heard.
A lack of COVID-19 and total death benefit
The trial was not designed to be powered enough, or perhaps designed to be not powered enough (too brief, too small) to use death as a primary endpoint, but the lack of Covid-19 and total death benefit shown, by what sparse randomised trial data we have, does not support a lockdown and wait for a vaccine policy, vaccine passports, vaccine mandates or vaccination. Instead the worrying safety signals from the excess cardiovascular and total deaths in the vaccine group mean there was not enough data to proceed, or proceed with extreme caution and full transparency with scrupulous monitoring of recipients. How did a safety signal for excess deaths in the vaccine group become, in an Orwellian transfiguration, the marketing motto ‘safe and effective’?
Conclusion
In summary, each of the 7 participants that died of or following Covid-19 had:
a BMI over 34, or age over 80, or a diagnosis of cancer;
and one of the following:
coronary artery disease, cerebrovascular disease, severe chronic lung disease, HIV infection or type II diabetes;
and
hyperlipidaemia and/or hypertension;
and
were over the age of 55.
Although previous reviews of risk factors for severe Covid-19 illness and death have isolated risk factors and calculated risk ratios or similar, this review illustrates how risk factors can powerfully work together in individual patients albeit using only a small number of examples.
Obese participants (BMI ≥30 kg/m2) had about 4.8 × higher risk of death from or following Covid-19 (CIs 0.9–24.7) and the risk appears to be significantly higher in the severely obese. The findings not only confirm the importance of obesity and severe obesity as risk factors for death from or following Covid-19 but suggest the risk may have been underestimated. The question arises, why was obesity excluded from consideration by the IAG, the group set up to assess high risk conditions? The comorbidities of the trial participants who died of or following Covid-19 reflect the metabolic health crisis affecting much of the world. A unique and powerful opportunity to start tackling the metabolic health crisis and make long term health improvements in the UK was wasted.
