30th November 2018, Volume 131 Number 1486

Nick Wilson, Osman D Mansoor, Michael G Baker

The international evidence is mixed on there being socio-economic mortality gradients from the 1918 influenza pandemic,1 but in New Zealand no clear socio-economic gradient in mortality rates has been reported for the overall European population.2 For example, Rice3 described the distribution of mortality by occupational groupings and socio-economic status but there was no obvious gradient (albeit without statistical analysis being performed). Also there was no socio-economic gradient in two studies of military populations when analysed using an occupational class measure.4,5 However, military populations in 1918 were of relatively younger men subjected to selection processes for health status as part of recruitment. An unpublished study of mortality in Dunedin identified the socio-economic characteristics of the deceased, but also did not calculate rates.6 Another local study7 did calculate rates, but reported no apparent variation in mortality for different housing districts in Auckland City. This latter work referred to an unpublished history thesis,8 which contained data that we have now analysed further using modern biostatistical methods.

The data in this unpublished thesis8 was at the suburb level with suburbs being categorised as ‘working-class suburbs’ and ‘well-to-do suburbs’ (albeit this being a sample that excluded 15 ‘outlying suburbs and districts’ and 11 other suburbs/areas in Auckland). When we analysed the data in these two groupings we calculated that the former collectively had a mortality rate of 9.1 per 1,000 and the latter of 6.4 per 1,000 population (Table 1). This was a statistically significant difference with a rate ratio of 1.42 (95% CI: 1.10–1.82, p=0.008, mid-p exact two-tailed test). Furthermore, for pandemic-related deaths of the ‘head of the household’ with occupational class coded on a six point scale (ie, from 1 = higher professional and administrative; vs 6 = unskilled),8 the mean score we calculated was higher in working-class suburbs than in ‘well-to-do’ suburbs (4.3 vs 3.7, p=0.0031, Table 1). That is, those dying in working-class suburbs also had lower (more deprived) individual-level occupational class.

Table 1: Analysis of pandemic-related mortality data for Auckland City for 1918 (data for individual suburbs extracted from Bryder 19808).

Type of suburb in 1918*

Pandemic-
related deaths in 1918

Population in the 1916 Census

Crude mortality rate per 1,000 population

(95% CI)

Average occupational class of the ‘head of the household’: 6 point scale; 6 is lowest (SD)

‘Well-to-do’ suburbs

Mt Eden

90

12,555

7.2 (5.8–8.8)

3.7 (1.5)

Takapuna

20

2,756

7.3 (4.7–11.2)

4.0 (1.6)

Birkenhead

15

2,116

7.1 (4.3–11.7)

3.4 (1.7)

Northcote

10

1,651

6.1 (3.3–11.1)

5.0 (0.9)

Devonport

36

7,613

4.7 (3.4–6.5)

3.4 (1.4)

Total

171

26,691

6.4 (5.5–7.4)

3.7 (1.5)

‘Working-class’ suburbs

Onehunga, Te Papapa

59

5,913

10.0 (7.7–12.9)

4.3 (1.5)

Newmarket

24

2,863

8.4 (5.6–12.7)

4.4 (1.4)

Ellerslie

9

1,363

6.6 (3.5–12.5)

3.8 (1.6)

Total

92

10,139

9.1 (7.4–11.1)

4.3 (1.5)

*Despite some levels of gentrification in the working-class suburbs, these socio-economic patterns still partly apply in modern day New Zealand. For example, using an area deprivation measure “NZDep2006”, Birkenhead and Northcote contain census area units (CAUs) that are in the 2nd decile of deprivation (ie, 2nd least deprived decile); while Onehunga and Te Papapa have CAUs in the 8th decile (ie, near to the most deprived 10th decile). At this time in New Zealand’s history the Māori urban population was very low and so this analysis is essentially of the European New Zealand population. 

We also took this opportunity to consider an analysis of national data collected by Rice.3 But given the difficulties of assigning socio-economic status to the occupational groupings, we simply compared ‘professionals’ to all the other occupations which were harder to rank (Table 2). This analysis suggested a statistically significant lower mortality rate in the professionals group vs the other occupations.

Table 2: Analysis of professionals vs other occupation groups among both male and female “breadwinners” (numerator and denominator data from Rice;3 with the latter from the 1916 Census15).

Occupational group

Pandemic-
related deaths in 1918

Population in the 1916 Census

Crude mortality rate per 1000 population (95% CI)

Rate ratio (95% CI)

Professionals

189

29,970

6.3 (5.5–7.3)

0.87 (0.76–1.01);

(p = 0.035, 1-tailed test)

All other occupational groups*

2,971

412,190

7.2 (7.0–7.5)

Reference (1.0)

*Includes: Armed forces; Accommodation and domestic services; Commercial, financial and retail; Transport and communication; Industry and manufacture; Trades and construction; and Primary production. Note: due to the lack of relevant data the rates are not age/sex standardised. 

All these results could potentially be somewhat confounded by age, eg, if proportionately more young adults in their 20s and 30s, with elevated pandemic-related death rates,3 resided in the working-class suburbs. However, we have no data on suburban age structures or occupational age structures to evaluate these possible effects. Furthermore, analysis of ‘suburbs’ has various limitations, eg, a ‘working class suburb’ may contain pockets of wealthier population groups.

Nevertheless, the suggested socio-economic gradients we observed in these analyses are consistent with other data for New Zealand in this historical period which indicates lifespan differences by male occupational class9 and for ethnic inequalities in mortality for three previous influenza pandemics in New Zealand (higher Māori vs European mortality).10 Also these observed gradients are consistent with at least some of the international literature for mortality patterns in the 1918 pandemic, eg, for the US,11 for Chicago (US),12 for Sweden1 and for Norway13 (albeit with variation by pandemic wave in one setting in Norway14). It is also very plausible that such a pandemic-related mortality gradient existed given that poverty is associated with household crowding and chronic conditions such as tuberculosis, which increased the risk of death in this pandemic. Even so, more research relating to such socio-economic gradients (and ethnic gradients) in this, and subsequent pandemics, would seem desirable, so as to achieve a better understanding of their true impact. Indeed, future studies in New Zealand could try to devise ways to grade localities in 1918 according to socio-economic status or deprivation (eg, by infant mortality rates) and analyse individual level data with adjustment for potential confounders.

Prudent policy-makers should, however, assume that such socio-economic gradients may arise with future influenza and other pandemics and so pandemic planning should aim to minimise health inequalities. That is if border control was not implemented or failed, more deprived communities could be more intensely supported with disease control activities that both minimised incidence rates and maximised healthcare support to those who become ill. The bigger implication is that pandemic preparedness is yet another reason to reduce health inequalities now, eg, by improving housing, reducing crowding and eliminating preventable causes of health burden among low-income New Zealanders (eg, tobacco and obesogenic food environments).

Author Information

Nick Wilson, Department of Public Health, University of Otago, Wellington; 
Osman D Mansoor, Public Health Physician, Wellington; 
Michael G Baker, Department of Public Health, University of Otago, Wellington.

Correspondence

Nick Wilson, Department of Public Health, University of Otago, Wellington.

Correspondence Email

nick.wilson@otago.ac.nz

Competing Interests

Nil.

References

  1. Bengtsson T, Dribe M, Eriksson B. Social Class and Excess Mortality in Sweden During the 1918 Influenza Pandemic. Am J Epidemiol 2018; (E-publication 13 July).
  2. Summers J, Baker M, Wilson N. New Zealand’s experience of the 1918-19 Influenza Pandemic: A systematic review after 100 years. N Z Med J (In press).
  3. Rice G. Black November: The 1918 influenza pandemic in New Zealand. Christchurch: Canterbury University Press, 2005.
  4. Summers JA, Wilson N, Baker MG, Shanks GD. Mortality risk factors for pandemic influenza on New Zealand troop ship, 1918. Emerg Infect Dis 2010; 16:1931–7.
  5. Summers JA, Stanley J, Baker MG, Wilson N. Risk factors for death from pandemic influenza in 1918–1919: a case-control study. Influenza Other Respir Viruses 2014; 8:329–38.
  6. Cuff M. The great scourge: Dunedin in the 1918 influenza epidemic [Long Essay, Postgraduate Diploma in History]. Dunedin: University of Otago, 1980.
  7. Bryder L. “Lessons” of the 1918 influenza epidemic in Auckland. N Z J History 1982; 16:97–121.
  8. Bryder L. The 1918 Influenza Epidemic in Auckland. [Unpublished MA Thesis, University of Auckland], 1980.
  9. Wilson N, Clement C, Boyd M, Teng A, Woodward A, Blakely T. The long history of health inequality in New Zealand: occupational class and lifespan in the late 1800s and early 1900s. Aust N Z J Public Health 2018;(E-publication 15 February).
  10. Wilson N, Telfar Barnard L, Summers J, Shanks G, Baker M. Differential mortality by ethnicity in 3 influenza pandemics over a century, New Zealand. Emerg Infect Dis 2012; 18:71–77.
  11. Sydenstriker E. The Incidence of Influenza among Persons of Different Economic Status during the Epidemic of 1918. Public Health Rep 1931; 46:154–70. http://www.jstor.org/stable/pdf/4579923.pdf?refreqid=excelsior%3A282dbb104c7b65941be798c97a3ce3ae
  12. Grantz KH, Rane MS, Salje H, Glass GE, Schachterle SE, Cummings DA. Disparities in influenza mortality and transmission related to sociodemographic factors within Chicago in the pandemic of 1918. Proc Natl Acad Sci U S A 2016; 113:13839–44.
  13. Mamelund SE. A socially neutral disease? Individual social class, household wealth and mortality from Spanish influenza in two socially contrasting parishes in Kristiania 1918–19. Soc Sci Med 2006; 62:923–40.
  14. Mamelund SE. 1918 pandemic morbidity: The first wave hits the poor, the second wave hits the rich. Influenza Other Respir Viruses 2018; 12:307–13.
  15. Statistics New Zealand. Report on the results of a census of the population of the Dominion of New Zealand taken for the night of the 15th October, 1916. http://www3.stats.govt.nz/historic_publications/1916-census/Report%20on%20Results%20of%20Census%201916/1916-report-results-census%20.html?_ga=2.6668146.1932235955.1541614054-1286078767.1541614054#idpreface_1_34