ResearchHealth and historical levels of freedom

Background: The link between political freedom and health is unclear. We aimed to determine the association by exploring the relationship of historical and cumulative freedom levels with important health outcomes. Methods: We obtained countrywide health indicators for life expectancy, infant mortality, maternal mortality ratio, % low birth weight babies, Gini coefficient (a measure of wealth inequality) and various markers of freedom based on political rights and civil liberties. We applied multivariable logistic regression to examine the association between health indicators and within-country years of freedom as determined by Freedom House rankings. Results: The total proportion of free years from 1972-2005, the duration of current freedom level, and the Gini coefficient show independent positive associations with health indicators, which remain after the adjustment for national wealth, total government expenditure, and spending on health. Countries identified as having high total proportion of free years demonstrated significantly better health outcomes than countries with low levels of freedom (life expectancy, Odds Ratio [OR] 7.2, 95% Confidence Interval [CI], 2.3-22.6, infant mortality OR 19.6, 95% CI, 5.6-67.7, maternal mortality ratio, OR 24.3, 95% CI, 6.2-94.9, and % low birth weight babies OR 3.8, 95% CI, 1.4-10.8). This was also the case for infant mortality (OR 3.4, 95% CI, 1.0-8.4), maternal mortality ratio (OR 4.0, 95% CI, 1.2-12.8), and % low birth weight babies (OR 2.6, 95% CI, 1.0-6.6) among countries considered as having medium levels of freedom. Interpretation: We found strong associations between country-level freedom and important health outcomes. The cumulative level of freedom over time shows stronger associations with all health indicators than the duration of current freedom level. Background Although the link between politics and health is often discussed,[1,2] few studies have determined the specific influence of national political rights and civil liberties on the health of individuals and populations living under them. With various organizations now providing data on political and health indicators from all over the world, large-scale global comparisons are now possible[3]. In this analysis, we explore the relationship of historical and cumulative freedom levels, based on political rights and civil liberties, using various health indicators. Previously, Franco and colleagues used freedom rankings from the Freedom House as a proxy for democracy and explored the relationship between democracy and health[4]. The authors found that higher levels of democracy were associated with better health outcomes. However, since they compared Freedom House ratings to health indicators for one year, the cross-sectional nature of the study limited the analysis of their results. For example, a country that transitioned from 'not free' to 'free' in 1998 was rated the same as a country that had been free for decades. Thus, their results did not account for the effects of recent political transition or the cumulative effect of political systems on health over time. We expand on this issue by examining the recency of political transition and the cumulative level of freedom. In this way, we aim to determine how a country's historical level of freedom since 1972, over roughly two human generations, influences its present day health status. Methods We created a database of 181 countries (in existence as of 2005) for life expectancy, infant mortality, maternal mortality ratio, % low birth weight babies, Gini coefficient (a measure of wealth distribution), total government expenditure (USD), Gross National Income/capita, % total GDP spent on health, and historical level of freedom. Data are * Correspondence: emills@sfu.ca 3 Ottawa Hospital Research Institute, Ottawa, Canada Full list of author information is available at the end of the article © 2010 Biggs et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Biggs et al. Globalization and Health 2010, 6:11 http://www.globalizationandhealth.com/content/6/1/11 Page 2 of 8 collected from various sources,[5-13] identified in Table 1, and relate to the year 2007. Where data for 2007 was unavailable, the last observation was carried forward. Historical levels of freedom are measured using Freedom House ratings from 1972-2005, based on the availability of Freedom House data. Freedom House uses key metrics, that relate to political rights and civil liberties to rate countries as free, partially free, or not free for each year[3]. Political rights considered in the rating are electoral process, political pluralism and participation, and functioning of government. Civil liberties include freedom of expression and belief, associational and organizational rights, rule of law, and personal autonomy and individual rights. Methods for rating countries are described in detail elsewhere[14]. The freedom ratings are manipulated to determine the total proportion of free years from 1972-2005 (TPFY), calculated as: Here, the years of missing data are omitted from the calculation altogether, whereas 'Not Free' years are included in the denonimator. Thus, 'Not Free' years reduce the overall value of the TPFY and years of missing data do not change the TPFY. The duration of free status (DFS), calculated for all countries with 'Free' status as of 2005. Countries that achieved 'Free' status in 2005 were assigned a score of 1, in 2004 were assigned a score of 2, etc. Countries scored a maximum of 34 if they were 'Free' from 1972-2005; The duration of partially free status (DPFS), calculated for all countries with 'Partially Free' status as of 2005. Countries that achieved 'Partially Free' status in 2005 were assigned a score of 1, in 2004 were assigned a score of 2, etc. Countries scored a maximum of 34 if they were 'Partially Free' from 1972-2005; The duration of not free status (DNFS), calculated for all countries with 'Not Free' status as of 2005. Countries that achieved 'Not Free' status in 2005 were assigned a score of 1, in 2004 were assigned a score of 2, etc. Countries scored a maximum of 34 if they were 'Not Free' from 1972-2005. The TPFY effectively measures the cumulative level of freedom over time. The DFS, DPFS, and DNFS measure the recency of political transition. Each country has only one of DFS, DPFS, or DNFS based on their Freedom House ranking as of 2005. Once the normality of all variables was confirmed, descriptive statistics were calculated for each variable. All health and freedom variables were split into tertiles, where the lower 33% of observations are labelled as 'low', the middle 33% as 'medium', and the upper 33% as 'high'. The Gini coefficient is also split into tertiles. Despite a lower Gini signifying greater equality, for simplicity, the tertiles are labelled such that the upper 33% of observations are labelled as 'low', the middle 33% as 'medium', and the lower 33% as 'high.' Analysis For health and freedom variables, we performed unadjusted and adjusted multivariable logistic regression to control for the effects of wealth (measured as per capita gross national product), level of inequality (measured with the Gini coefficient), and size of the public sector (measured as total government expenditure and percentage of GDP spent on health). For the Gini coefficient, we performed multivariable logistic regression to control for the effects of wealth (measured as per capita gross national product), size of the public sector (measured as total government expenditure and percentage of GDP spent on health), and the total proportion of free years. In the multivariate analysis, countries missing data for health indicators or control variables were excluded. Of all 181 countries considered, the unadjusted life expectancy analysis included 176 (adjusted, n = 129), unadjusted infant mortality analysis included 181 (adjusted, n = 133), unadjusted maternal mortality ratio analysis included 166 (adjusted, n = 128), and unadjusted low birth weight analysis included 173 (adjusted, n = 130) based on the availability of data. Data are presented as Odds Ratios [ORs] with 95% Confidence Intervals [CI]. All p-values are two-sided and exact. We considered a pvalue of < 0.05 as statistically significant. Results Our sample of 181 countries represents 98.5% of the world's population and includes 94.3% of the states recognized by the United Nations. As of 2005, 44% of the countries were considered free, 31.5% partially free, and 24% not free. Within populations exposed to lack of freedom, 17.1% and 36.2% live in a partially free and not free counTPFY [( free yrs 2) ( partially free yrs 1) ( not fre = # + # ...