Header
Updated: 13-Feb-2001 NATO Information

 

HEALTH CONCERNS RELATED TO BALKAN DEPLOYMENTS

At a previous Ad Hoc Group the UK presented an update on the retrospective cohort study, that was published in the Lancet last year ¹ , that compared a UK Gulf War veterans cohort with a comparable number of personnel who did not serve in the Gulf (Era group). As a result of discussions at the COMEDS meetings held on 16 January 2001 we have subsequently examined this study data in relation to cancer, and specifically lymphatic and haematopoietic cancers. We have compared these (as requested at COMEDS) to the UK population using Standardised Mortality Rates.

Analysis of the data over the periods 1991-2000 and for 1995-2000 reveals that for 'all cancer' deaths both the Gulf and Era cohorts have statistically significantly reduced mortality at the 5% level when compared to the standardised population of the UK. A statistically significant reduced mortality at the 5% level is also shown for Lymphatic and Haematopoietic cancer during the period 1991-2000 for both the Gulf and Era cohorts. Restricting the data to 1995-2000, to allow for latency effects from service in the Gulf, Lymphatic and Haematopoietic cancers continue to show a reduced mortality for the Gulf and Era cohorts. However, only in the Era cohort does this reach significance at the 5% level when compared to the UK population. The data is presented in Annex A. I would emphasise that this study is a prospective one, following up the cohort into civilian life after they have left the Services.

You will also note that the total number of cancer deaths in the Gulf cohort was less than those in the control group (69 versus 77) and that the total number of deaths from all lymphatic and haematopoietic neoplasms was higher in the Gulf cohort (13 versus 9). However, neither of these differences is statistically significant, either for the whole period (1991-2000) or for 1995-2000.

We have separately looked at all deaths of serving personnel and at medical discharges from the Armed Forces due to Lymphatic and Haematopoietic cancers from 1990 - 2000. Overall there was no evidence to support the fact that discharges and deaths from Lymphatic and Haematopoietic cancers over the last few years represented any real increase above that observed during the earlier part of the decade. This data is at Annex B.

As the kidney is the target organ for uranium toxicity the professor of Military Medicine, a nephrologist, has also looked at the number of renal biopsies undertaken on soldiers prior to and post to the Gulf War. He found no significant difference between the number of biopsies undertaken before or after the war. A precis of his findings is at Annex C.

In conclusion, we have sought evidence of an increase in lymphatic and haematopoietic neoplasms in UK Service personnel in the period from the Gulf war to the present. Neither the prospective epidemiological study of mortality, nor our records of deaths and discharges from the Armed Forces, provide support for an increase."

  1. Mortality among UK Gulf War veterans. LANCET 1 Jul 2000 No.9223:17-21

Annexes:

  1. Cancer deaths -Gulf and Era cohorts
  2. Deaths and discharges due to Lymphatic and Haematopoietic cancer
  3. UK Army renal biopsies pre and post Gulf War.

Annex A To D/SG(Plans)950/2 (Dated 05 Feb 2001)

All cancer deaths
1 Apr 91 - 31 Dec 00
  Estimates Actual SMR Lower 95% Upper 95%
Gulf 119 69 58 46 74
Era 119 77 65 52 81
1 Jan 95 - 31 Dec 00
  Estimates Actual SMR Lower 95% Upper 95%
Gulf 87 54 62 47 81
Era 87 58 66 51 86

 
All lymphatic and haematopoietic neoplasms
1 Apr 91 - 31 Dec 00
  Estimates Actual SMR Lower 95% Upper 95%
Gulf 23 13 57 33 98
Era 23 9 39 21 76
1 Jan 95 - 31 Dec 00
  Estimates Actual SMR Lower 95% Upper 95%
Gulf 15 9 59 30 112
Era 15 3 20 6 60

The analyses for the periods 1991 - 2000 and 1995 - 2000 show that for "All cancer" both the Gulf and Era cohorts have statistically significantly reduced mortality at the 5% level (Confidence interval does not include 100).

A statistically significantly reduced mortality at the 5% level is noted for Lymphatic and Haematopoietic cancer during the period 1991 - 2000 for both the Gulf and Era cohorts.

Restriction to the period 1995 - 2000 for Lymphatic and Haematopoietic cancer shows reduced mortality for the Gulf and Era cohorts, but only in the Era cohort does this reach statistical significance at the 5% level.

Annex B To D/SG (Plans) 950/2 Dated 05 Feb 2001

Deaths and Discharges due to Lymphatic and Haematopoietic cancer 1990 - 2000

Deaths + Discharges 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Army 7 5 3 2 4 2 4 5 7 1 4
Navy 3 4 3 2 1 1 3 2 4 3 3
RAF 9 6 5 2 3 3 3 2 6 7 4
Total 19 15 11 6 8 6 10 9 17 11 11

Strength ('OOO's) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Army 152.8 147.6 145.4 134.6 123 111.7 108.8 108.8 109.8 109.7 110.1
Navy 63.2 62.1 62.1 59.4 55.8 50.9 48.3 45.1 44.5 43.7 42.8
RAF 89.7 88.4 86 80.9 75.7 70.8 64.7 56.9 55.8 55.2 54.7
Total 305.7 298.1 293.4 274.8 254.5 233.3 221.9 210.8 210.1 208.6 207.6

Rates / 1000 Strength 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Army 4.6 3.4 2.1 1.5 3.3 1.8 3.7 4.6 6.4 0.9 3.6
Navy 4.7 6.4 4.8 3.4 1.8 2 6.2 4.4 9 6.9 7
RAF 10 6.8 5.8 2.5 4 4.2 4.6 3.5 10.7 12.7 7.3
Total 6.2 5 3.7 2.2 3.1 2.6 4.5 4.3 8.1 5.3 5.3

Graphical plots were produced of the rates / 1000 strength. Across the whole period for the Tri-Service data there was little visual evidence of ny trend although a slight tendency for an increase over the latter years 1996 - 2000 above the period average.

Ordinary least squares regression was used for the Tri-Service and Army data, to examine evidence for a trend with time over the period but one was found.

On the basis that there is no evidence of a trend with time over the period, any pair wise comparison of rates is fairly arbitrary. Comparisons using the 1998 rate against the years 1992 - 1995 were significant at the 5% level for the Tri-Service data, but all other comparisons for the Tri-Service data, and in particular the 1998 rate against the period average were non significant.

Annex C To D/SG (Plans) 950/2 Dated 05 Feb 2001

The Professor of Military Medicine at the Royal Hospital Haslar has conducted an analysis of the annual frequency of renal biopsies carried out on British Army-personnel before and after the Gulf War in 1991. The mean annual frequency of biopsies in a 6-year prior to the conflict was 10.83 and 10.6 in the 5 years following the conflict. There is no statistical significance between these values.2

Significant exposure to uranium is associated with interstitial nephritis. Interstitial nephritis has been a rare diagnosis in this group of personnel who have undergone renal biopsy comprising only 3/118 of the biopsies, one case prior to and two cases post the Gulf War. The professor of military medicine considers that a more likely cause of the interstitial nephritis diagnosed in these cases is related to the consumption of non-steroidal anti-inflammatory drugs. He has no evidence that there has been a surge in the number of cases of putative nephritis in the 5 years following the Gulf War, and hence no evidence of uranium toxicity.

  1. Variances were significantly heteroscedastic therefore a Mann-Whitney U test was applied to determine statistical significance.