Skip to content

NCD Risk Factor Collaboration (NCD-RisC). Worldwide trends in hypertension prevalence and progress in treatment and control from 1990 to 2019: a pooled analysis of 1201 population-representative studies with 104 million participants. Lancet. 2021;398:957-980. PMID 34450083

One-paragraph summary

Pooling 1,201 population-representative studies covering 104 million participants aged 30–79 across 200 countries and territories, and applying a Bayesian hierarchical model allowing non-linear, age-varying trends, the collaboration estimated hypertension prevalence (≥140/90 mm Hg or taking medication) plus the proportions diagnosed, treated and controlled from 1990 to 2019. The number of people with hypertension doubled, from 331 million women (95% credible interval 306–359) and 317 million men (292–344) in 1990 to 626 million (584–668) and 652 million (604–698) in 2019, while global age-standardised prevalence stayed essentially flat — so the entire increase is population growth and ageing, not a worsening of underlying risk. In 2019, 59% (55–62) of women and 49% (46–52) of men with hypertension had ever been diagnosed, 47% (43–51) and 38% (35–41) were treated, and 23% (20–27) and 18% (16–21) were controlled to below 140/90 mm Hg.

Key findings

  • Absolute burden doubled 1990→2019 on flat age-standardised prevalence: 648 million to 1.278 billion adults aged 30–79.
  • The global control rate is 23% (women) / 18% (men) — the single most quoted number in the field.
  • Prevalence exceeded 50% for women in two countries and men in nine, concentrated in central and eastern Europe, central Asia, Oceania and Latin America.
  • Lowest age-standardised prevalence: Canada and Peru for both sexes; Taiwan, South Korea, Japan and parts of western Europe for women; Eritrea, Bangladesh, Ethiopia and Solomon Islands for men.
  • Best performers on treatment (>70%) and control (>50%): South Korea, Canada, Iceland, then USA, Costa Rica, Germany, Portugal, Taiwan.
  • Worst: treatment <25% (women) and <20% (men) in Nepal, Indonesia and parts of sub-Saharan Africa and Oceania, with control below 10%.
  • Improvement since 1990 has been largest in high-income countries, central Europe and several upper-middle-income countries (Costa Rica, Taiwan, Kazakhstan, South Africa, Brazil, Chile, Turkey, Iran), and negligible in most of sub-Saharan Africa and Oceania.

Limitations

  • Prevalence depends entirely on the 140/90 threshold; the number would be very different under 130/80 criteria (see [Muntner 2018, PMID 29133599] in the bibliography).
  • Contributing surveys used heterogeneous measurement protocols — typically two or three seated office readings — so white-coat and masked phenotypes are misclassified in both directions.
  • Diagnosis and treatment status are self-reported and subject to recall and social-desirability bias.
  • Country-level estimates for data-sparse settings are model-driven, with credible intervals that are wide relative to the differences being compared.
  • The analysis is restricted to ages 30–79, excluding both the young-adult failure documented elsewhere and the very old.

Why it matters

This is the paper that reframed hypertension from a prevalence problem to a delivery problem, and it supplies the denominator for almost every argument in the field. Its two structural findings — that growth is demographic rather than risk-driven, and that several middle-income countries now outperform most high-income ones on treatment and control — jointly refute the assumption that control follows automatically from economic development. It is the empirical basis for the framing used throughout this condition: the science is settled far ahead of the delivery.

Cited by wiki pages

  • overview
  • epidemiology and burden
  • adherence and implementation